Searched refs:slots (Results 1 - 25 of 55) sorted by relevance

123

/prebuilts/gdb/darwin-x86/lib/python2.7/
H A D	copy_reg.py	98 This needs to find slots defined by the class and its bases, so we 102 __slots__ attribute to misrepresent their slots after the class is 114 # This class has no slots 120 slots = c.__dict__['__slots__'] 122 if isinstance(slots, basestring): 123 slots = (slots,) 124 for name in slots:
/prebuilts/gdb/linux-x86/lib/python2.7/
H A D	copy_reg.py	98 This needs to find slots defined by the class and its bases, so we 102 __slots__ attribute to misrepresent their slots after the class is 114 # This class has no slots 120 slots = c.__dict__['__slots__'] 122 if isinstance(slots, basestring): 123 slots = (slots,) 124 for name in slots:
/prebuilts/python/darwin-x86/2.7.5/lib/python2.7/
H A D	copy_reg.py	98 This needs to find slots defined by the class and its bases, so we 102 __slots__ attribute to misrepresent their slots after the class is 114 # This class has no slots 120 slots = c.__dict__['__slots__'] 122 if isinstance(slots, basestring): 123 slots = (slots,) 124 for name in slots:
/prebuilts/python/linux-x86/2.7.5/lib/python2.7/
H A D	copy_reg.py	98 This needs to find slots defined by the class and its bases, so we 102 __slots__ attribute to misrepresent their slots after the class is 114 # This class has no slots 120 slots = c.__dict__['__slots__'] 122 if isinstance(slots, basestring): 123 slots = (slots,) 124 for name in slots:
/prebuilts/ndk/current/sources/third_party/shaderc/third_party/glslang/glslang/MachineIndependent/
H A D	iomapper.cpp	366 TSlotSetMap slots; member in struct:glslang::TDefaultIoResolverBase 370 return std::lower_bound(slots[set].begin(), slots[set].end(), slot); 376 return !(at != slots[set].end() && *at == slot); 385 if (at == slots[set].end() || *at != slot) 386 slots[set].insert(at, slot); 394 if (at == slots[set].end()) 398 for (; at != slots[set].end(); ++at, ++base)
/prebuilts/ndk/r16/sources/third_party/shaderc/third_party/glslang/glslang/MachineIndependent/
H A D	iomapper.cpp	366 TSlotSetMap slots; member in struct:glslang::TDefaultIoResolverBase 370 return std::lower_bound(slots[set].begin(), slots[set].end(), slot); 376 return !(at != slots[set].end() && *at == slot); 385 if (at == slots[set].end() || *at != slot) 386 slots[set].insert(at, slot); 394 if (at == slots[set].end()) 398 for (; at != slots[set].end(); ++at, ++base)
/prebuilts/misc/common/swig/include/2.0.11/python/
H A D	pyhead.swg	200 PyObject *name, *slots;
/prebuilts/misc/common/jython/
H A D	jython.jar	META-INF/MANIFEST.MF com/xhaus/modjy/ModjyJServlet.class ModjyJServlet.java package com.xhaus ...
/prebuilts/devtools/tools/lib/
H A D	jython-standalone-2.5.3.jar	META-INF/MANIFEST.MF META-INF/LICENSE META-INF/NOTICE META-INF/maven/jline/ ...
/prebuilts/tools/common/m2/repository/org/python/jython/2.5.3/
H A D	jython-2.5.3.jar	META-INF/MANIFEST.MF META-INF/services/javax.script.ScriptEngineFactory com/xhaus/modjy/ModjyJServlet ...
/prebuilts/tools/common/m2/repository/org/python/jython-standalone/2.5.3/
H A D	jython-standalone-2.5.3.jar	META-INF/MANIFEST.MF META-INF/LICENSE META-INF/NOTICE META-INF/maven/jline/ ...
/prebuilts/gdb/darwin-x86/lib/python2.7/test/
H A D	test_descr.py	1013 self.fail("[None] slots not caught") 1020 self.fail("['foo bar'] slots not caught") 1027 self.fail("['foo\\0bar'] slots not caught") 1034 self.fail("['1'] slots not caught") 1041 self.fail("[''] slots not caught") 1067 # _unicode_to_string used to modify slots in certain circumstances 1068 slots = (unicode("foo"), unicode("bar")) variable in class:.test_slots.C 1070 __slots__ = slots 1074 self.assertEqual(type(slots[0]), unicode) 1082 self.fail("[unichr(128)] slots no [all...]
/prebuilts/gdb/linux-x86/lib/python2.7/test/
H A D	test_descr.py	1013 self.fail("[None] slots not caught") 1020 self.fail("['foo bar'] slots not caught") 1027 self.fail("['foo\\0bar'] slots not caught") 1034 self.fail("['1'] slots not caught") 1041 self.fail("[''] slots not caught") 1067 # _unicode_to_string used to modify slots in certain circumstances 1068 slots = (unicode("foo"), unicode("bar")) variable in class:.test_slots.C 1070 __slots__ = slots 1074 self.assertEqual(type(slots[0]), unicode) 1082 self.fail("[unichr(128)] slots no [all...]
/prebuilts/python/darwin-x86/2.7.5/lib/python2.7/test/
H A D	test_descr.py	1013 self.fail("[None] slots not caught") 1020 self.fail("['foo bar'] slots not caught") 1027 self.fail("['foo\\0bar'] slots not caught") 1034 self.fail("['1'] slots not caught") 1041 self.fail("[''] slots not caught") 1067 # _unicode_to_string used to modify slots in certain circumstances 1068 slots = (unicode("foo"), unicode("bar")) variable in class:.test_slots.C 1070 __slots__ = slots 1074 self.assertEqual(type(slots[0]), unicode) 1082 self.fail("[unichr(128)] slots no [all...]
/prebuilts/python/linux-x86/2.7.5/lib/python2.7/test/
H A D	test_descr.py	1013 self.fail("[None] slots not caught") 1020 self.fail("['foo bar'] slots not caught") 1027 self.fail("['foo\\0bar'] slots not caught") 1034 self.fail("['1'] slots not caught") 1041 self.fail("[''] slots not caught") 1067 # _unicode_to_string used to modify slots in certain circumstances 1068 slots = (unicode("foo"), unicode("bar")) variable in class:.test_slots.C 1070 __slots__ = slots 1074 self.assertEqual(type(slots[0]), unicode) 1082 self.fail("[unichr(128)] slots no [all...]
/prebuilts/gcc/darwin-x86/x86/x86_64-linux-android-4.9/bin/
H A D	x86_64-linux-android-gcc-4.9.x	3286 %ssection anchors must be disabled when unit-at-a-time is disabledtoplevel reorder must be disabled when unit-at-a-time is disabledtransactional memory is not supported with non-call exceptionssection anchors must be disabled when toplevel reorder is disabled-freorder-blocks-and-partition does not work with exceptions on this architecture-freorder-blocks-and-partition does not support unwind info on this architecture-freorder-blocks-and-partition does not work on this architecture-fno-fat-lto-objects are supported only with linker pluginonly one -flto-partition value can be specified%<-fsplit-stack%> is not supported by this compiler configurationDebug generation via -g option disabled under -fripa -fprofile-generate (use -fripa-allow-debug to override)-fsanitize=address is incompatible with -fsanitize=kernel-address-fsanitize=address and -fsanitize=kernel-address are incompatible with -fsanitize=threadargument to %<-O%> should be a non-negative integer, %<g%>, %<s%> or %<fast%>argument %qs to %<-femit-struct-debug-detailed%> unknown%<-femit-struct-debug-detailed=dir:...%> must allow at least as much as %<-femit-struct-debug-detailed=ind:...%>argument %qs to %<-femit-struct-debug-detailed%> not recognized%s: --param arguments should be of the form NAME=VALUEinvalid --param value %qs--help argument %q.*s is ambiguous, please be more specificunrecognized argument to --help= option: %q.*s-Werror=%s: no option -%sgetting core file size maximum limit: %msetting core file size limit to maximum: %munrecognized gcc debugging option: %cstructure alignment must be a small power of two, not %dspecificunknown stack check parameter "%s"%<-gdwarf%s%> is ambiguous; use %<-gdwarf-%s%> for DWARF version or %<-gdwarf -g%s%> for debug leveldwarf version %d is not supported1unrecognized argument to -fsanitize= option: %q.*sstabscoffdwarf-2xcoffvmsthreadshiftinteger-divide-by-zerounreachablevla-boundreturnnullsigned-integer-overflowboolenumoptimizerstargetwarningsundocumentedparamsjoinedseparatecommoncommon_handle_optionprint_specific_helpThis switch lacks documentationset_Wstrict_aliasingmaybe_default_optiontarget_handle_optionsysbaseanynonegen:ord:ind:dir:dfn:command line option %qs is not supported by this configurationmissing argument to %qsargument to %qs should be a non-negative integerunrecognized argument in option %qsvalid arguments to %qs are: %s/Volumes/Android/buildbot/src/android/gcc/toolchain/build/../gcc/gcc-4.9/gcc/opts-common.cread_cmdline_optiongenerate_optiongenerate_canonical_option-Wno--fno--f-mno---debug=-g--machine---machine-no---machine=--machine=no---machineno---optimize=-O--std=-std=--std--warn---warn-no---no-decode_cmdline_optionprune_optionsopt_enum_arg_to_valueoptions.cKnown assembler dialects (for use with the -masm-dialect= option):Known ABIs (for use with the -mabi= option):Known code models (for use with the -mcmodel= option):unknown excess precision style %qsunknown floating point contraction style %qsValid arguments to -mfpmath=:unrecognized function reorder value %qsKnown indirect branch choices (for use with the -mindirect-branch=/-mfunction-return= options):unknown IRA algorithm %qsunknown IRA region %qsKnown vectorization library ABIs (for use with the -mveclibabi= option):Known address mode (for use with the -maddress-mode= option):Known stack protector guard (for use with the -mstack-protector-guard= option):unknown Stack Reuse Level %qsValid arguments to -mstringop-strategy=:unrecognized visibility value %qsKnown TLS dialects (for use with the -mtls-dialect= option):unknown TLS model %qsunknown vectorizer cost model %qsunknown vtable verify initialization priority %qsAdaAdaSCILAdaWhyCC++FortranGoJavaLTOObjCObjC++-###--CLASSPATH--all-warnings--ansi--assemble--assertassertion missing after %qs--assert=--bootclasspath--canonical-prefixes--classpath--comments--comments-in-macros--compile--coverage--debug--define-macromacro name missing after %qs--define-macro=--dependencies--dump--dump=--dumpbase--dumpdir--encoding--entry--entry=--extdirs--extra-warnings--for-assembler--for-assembler=--for-linker--for-linker=--force-link--force-link=Display this information--help=--help=<class> Display descriptions of a specific class of options. <class> is one or more of optimizers, target, warnings, undocumented, params--imacrosmissing filename after %qs--imacros=--include--include-barrier--include-directorymissing path after %qs--include-directory-after--include-directory-after=--include-directory=--include-prefix--include-prefix=--include-with-prefix--include-with-prefix-after--include-with-prefix-after=--include-with-prefix-before--include-with-prefix-before=--include-with-prefix=--include=--language--language=--library-directory--library-directory=--no-canonical-prefixes--no-integrated-cpp--no-line-commands--no-pie--no-standard-includes--no-standard-libraries--no-sysroot-suffix--no-warnings--optimize--output--output-class-directory--output-class-directory=--output-pch=--output=--param--param <param>=<value> Set parameter <param> to value. See below for a complete list of parameters--param=--pass-exit-codes--pedantic--pedantic-errors--pie--pipe--prefix--prefix=--preprocess--print-file-name--print-file-name=--print-libgcc-file-name--print-missing-file-dependencies--print-multi-directory--print-multi-lib--print-multi-os-directory--print-multiarch--print-prog-name--print-prog-name=--print-search-dirs--print-sysroot--print-sysroot-headers-suffix--profile--resource--resource=--save-temps--shared--specs--specs=--static--symbolic--sysroot--sysroot=--target-helpAlias for --help=target--time--trace-includes--traditional--traditional-cpp--trigraphs--undefine-macro--undefine-macro=--user-dependencies--verbose--write-dependencies--write-user-dependencies-A-A<question>=<answer> Assert the <answer> to <question>. Putting '-' before <question> disables the <answer> to <question>-B-CDo not discard comments-CCDo not discard comments in macro expansions-CLASSPATH-D-D<macro>[=<val>] Define a <macro> with <val> as its value. If just <macro> is given, <val> is taken to be 1-E-F <dir> Add <dir> to the end of the main framework include path-HPrint the name of header files as they are used-I-I <dir> Add <dir> to the end of the main include path-J-J<directory> Put MODULE files in 'directory'-L-MGenerate make dependencies-MDGenerate make dependencies and compile-MD_-MF-MF <file> Write dependency output to the given file-MGTreat missing header files as generated files-MMLike -M but ignore system header files-MMDLike -MD but ignore system header files-MMD_-MPGenerate phony targets for all headers-MQ-MQ <target> Add a MAKE-quoted targetmissing makefile target after %qs-MT-MT <target> Add an unquoted target-N-O<number> Set optimization level to <number>-OfastOptimize for speed disregarding exact standards compliance-OgOptimize for debugging experience rather than speed or size-OsOptimize for space rather than speed-PDo not generate #line directives-Q-Qn-Qy-R-S-T-Tbss-Tbss=-Tdata-Tdata=-Ttext-Ttext=-U-U<macro> Undefine <macro>-WThis switch is deprecated; use -Wextra instead-Wa,-WabiWarn about things that will change when compiling with an ABI-compliant compiler-Wabi-tagWarn if a subobject has an abi_tag attribute that the complete object type does not have-WaddressWarn about suspicious uses of memory addresses-Waggregate-returnWarn about returning structures, unions or arrays-Waggressive-loop-optimizationsWarn if a loop with constant number of iterations triggers undefined behavior-WaliasingWarn about possible aliasing of dummy arguments-Walign-commonsWarn about alignment of COMMON blocks-WallEnable most warning messages-Wall-deprecation-Wall-javadoc-WampersandWarn about missing ampersand in continued character constants-Warray-boundsWarn if an array is accessed out of bounds-Warray-temporariesWarn about creation of array temporaries-Wassert-identifier-Wassign-interceptWarn whenever an Objective-C assignment is being intercepted by the garbage collector-WattributesWarn about inappropriate attribute usage-Wbad-function-castWarn about casting functions to incompatible types-Wboxing-Wbuiltin-macro-redefinedWarn when a built-in preprocessor macro is undefined or redefined-Wc++-compatWarn about C constructs that are not in the common subset of C and C++-Wc++0x-compatDeprecated in favor of -Wc++11-compat-Wc++11-compatWarn about C++ constructs whose meaning differs between ISO C++ 1998 and ISO C++ 2011-Wc-binding-typeWarn if the type of a variable might be not interoperable with C-Wcast-alignWarn about pointer casts which increase alignment-Wcast-qualWarn about casts which discard qualifiers-Wchar-concat-Wchar-subscriptsWarn about subscripts whose type is "char"-Wcharacter-truncationWarn about truncated character expressions-WclobberedWarn about variables that might be changed by "longjmp" or "vfork"-WcommentWarn about possibly nested block comments, and C++ comments spanning more than one physical line-WcommentsSynonym for -Wcomment-Wcompare-realsWarn about equality comparisons involving REAL or COMPLEX expressions-Wcondition-assign-Wconditionally-supportedWarn for conditionally-supported constructs-Wconstructor-name-WconversionWarn for implicit type conversions that may change a value-Wconversion-extraWarn about most implicit conversions-Wconversion-nullWarn for converting NULL from/to a non-pointer type-Wcoverage-mismatchWarn in case profiles in -fprofile-use do not match-WcppWarn when a #warning directive is encountered-Wctor-dtor-privacyWarn when all constructors and destructors are private-Wdate-timeWarn about __TIME__, __DATE__ and __TIMESTAMP__ usage-Wdeclaration-after-statementWarn when a declaration is found after a statement-Wdelete-incompleteWarn when deleting a pointer to incomplete type-Wdelete-non-virtual-dtorWarn about deleting polymorphic objects with non-virtual destructors-Wdep-ann-WdeprecatedWarn if a deprecated compiler feature, class, method, or field is used-Wdeprecated-declarationsWarn about uses of __attribute__((deprecated)) declarations-Wdisabled-optimizationWarn when an optimization pass is disabled-Wdiscouraged-Wdiv-by-zeroWarn about compile-time integer division by zero-Wdouble-promotionWarn about implicit conversions from "float" to "double"-Weffc++Warn about violations of Effective C++ style rules-Wempty-block-Wempty-bodyWarn about an empty body in an if or else statement-Wendif-labelsWarn about stray tokens after #elif and #endif-Wenum-compareWarn about comparison of different enum types-Wenum-identifier-Wenum-switch-WerrorTreat all warnings as errors-Werror-implicit-function-declarationThis switch is deprecated; use -Werror=implicit-function-declaration insteadimplicit-function-declaration-Werror=Treat specified warning as error-WextraPrint extra (possibly unwanted) warnings-Wextraneous-semicolonWarn if deprecated empty statements are found-Wfallthrough-Wfatal-errorsExit on the first error occurred-Wfield-hiding-Wfinal-bound-Wfinally-Wfloat-conversionWarn for implicit type conversions that cause loss of floating point precision-Wfloat-equalWarn if testing floating point numbers for equality-Wforbidden-Wforce-warningsDisable promoting warnings to errors-WformatWarn about printf/scanf/strftime/strfmon format string anomalies-Wformat-contains-nulWarn about format strings that contain NUL bytes-Wformat-extra-argsWarn if passing too many arguments to a function for its format string-Wformat-nonliteralWarn about format strings that are not literals-Wformat-securityWarn about possible security problems with format functions-Wformat-y2kWarn about strftime formats yielding 2-digit years-Wformat-zero-lengthWarn about zero-length formats-Wformat=-Wframe-larger-than=-Wframe-larger-than=<number> Warn if a function's stack frame requires more than <number> bytes-Wfree-nonheap-objectWarn when attempting to free a non-heap object-Wfunction-eliminationWarn about function call elimination-Whiding-Wignored-qualifiersWarn whenever type qualifiers are ignored.-WimplicitWarn about implicit declarations-Wimplicit-function-declarationWarn about implicit function declarations-Wimplicit-intWarn when a declaration does not specify a type-Wimplicit-interfaceWarn about calls with implicit interface-Wimplicit-procedureWarn about called procedures not explicitly declared-Wimport-Windirect-static-Winherited-variadic-ctorWarn about C++11 inheriting constructors when the base has a variadic constructor-Winit-selfWarn about variables which are initialized to themselves-WinlineWarn when an inlined function cannot be inlined-Wint-to-pointer-castWarn when there is a cast to a pointer from an integer of a different size-Wintf-annotation-Wintf-non-inherited-Wintrinsic-shadowWarn if a user-procedure has the same name as an intrinsic-Wintrinsics-stdWarn on intrinsics not part of the selected standard-Winvalid-memory-modelWarn when an atomic memory model parameter is known to be outside the valid range.-Winvalid-offsetofWarn about invalid uses of the "offsetof" macro-Winvalid-pchWarn about PCH files that are found but not used-Wjavadoc-Wjump-misses-initWarn when a jump misses a variable initialization-Wl,-Wlarger-than--Wlarger-than=-Wlarger-than=<number> Warn if an object is larger than <number> bytes-Wline-truncationWarn about truncated source lines-Wliteral-suffixWarn when a string or character literal is followed by a ud-suffix which does not begin with an underscore.-Wlocal-hiding-Wlogical-opWarn when a logical operator is suspiciously always evaluating to true or false-Wlong-longDo not warn about using "long long" when -pedantic-WmainWarn about suspicious declarations of "main"-Wmasked-catch-block-Wmaybe-uninitializedWarn about maybe uninitialized automatic variables-Wmissing-bracesWarn about possibly missing braces around initializers-Wmissing-declarationsWarn about global functions without previous declarations-Wmissing-field-initializersWarn about missing fields in struct initializers-Wmissing-format-attribute-Wmissing-include-dirsWarn about user-specified include directories that do not exist-Wmissing-noreturn-Wmissing-parameter-typeWarn about function parameters declared without a type specifier in K&R-style functions-Wmissing-prototypesWarn about global functions without prototypes-Wmudflapswitch %qs is no longer supported-WmulticharWarn about use of multi-character character constants-WnarrowingWarn about narrowing conversions within { } that are ill-formed in C++11-Wnested-externsWarn about "extern" declarations not at file scope-Wnls-Wno-effect-assign-WnoexceptWarn when a noexcept expression evaluates to false even though the expression can't actually throw-Wnon-template-friendWarn when non-templatized friend functions are declared within a template-Wnon-virtual-dtorWarn about non-virtual destructors-WnonnullWarn about NULL being passed to argument slots marked as requiring non-NULL-Wnormalized=-Wnormalized=<id|nfc|nfkc> Warn about non-normalised Unicode strings-Wnull-Wold-style-castWarn if a C-style cast is used in a program-Wold-style-declarationWarn for obsolescent usage in a declaration-Wold-style-definitionWarn if an old-style parameter definition is used-Wopenmp-simdWarn if a simd directive is overridden by the vectorizer cost model-Wout-of-dateWarn if .class files are out of date-Wover-ann-WoverflowWarn about overflow in arithmetic expressions-Woverlength-stringsWarn if a string is longer than the maximum portable length specified by the standard-Woverloaded-virtualWarn about overloaded virtual function names-Woverride-initWarn about overriding initializers without side effects-Wp,-WpackedWarn when the packed attribute has no effect on struct layout-Wpacked-bitfield-compatWarn about packed bit-fields whose offset changed in GCC 4.4-WpaddedWarn when padding is required to align structure members-Wparam-assign-WparenthesesWarn about possibly missing parentheses-WpedanticIssue warnings needed for strict compliance to the standard-Wpkg-default-method-Wpmf-conversionsWarn when converting the type of pointers to member functions-Wpointer-arithWarn about function pointer arithmetic-Wpointer-signWarn when a pointer differs in signedness in an assignment-Wpointer-to-int-castWarn when a pointer is cast to an integer of a different size-Wpoison-system-directoriesWarn for -I and -L options using system directories if cross compiling-WpragmasWarn about misuses of pragmas-Wproperty-assign-defaultWarn if a property for an Objective-C object has no assign semantics specified-WprotocolWarn if inherited methods are unimplemented-Wpsabi-Wraw-Wreal-q-constantWarn about real-literal-constants with 'q' exponent-letter-Wrealloc-lhsWarn when a left-hand-side array variable is reallocated-Wrealloc-lhs-allWarn when a left-hand-side variable is reallocated-Wredundant-declsWarn about multiple declarations of the same object-Wredundant-modifiersWarn if modifiers are specified when not necessary-WreorderWarn when the compiler reorders code-Wreturn-local-addrWarn about returning a pointer/reference to a local or temporary variable.-Wreturn-typeWarn whenever a function's return type defaults to "int" (C), or about inconsistent return types (C++)-Wripa-opt-mismatchWarn if primary and auxiliary modules have mismatched command line options-WselectorWarn if a selector has multiple methods-Wself-assignWarn when a variable is assigned to itself-Wself-assign-non-podWarn when a variable of a non-POD type is assigned to itself-Wsequence-pointWarn about possible violations of sequence point rules-Wserial-WshadowWarn when one local variable shadows another-Wshadow-compatible-localWarn when one local variable shadows another local variable or parameter of compatible type-Wshadow-localWarn when one local variable shadows another local variable or parameter-Wsign-compareWarn about signed-unsigned comparisons-Wsign-conversionWarn for implicit type conversions between signed and unsigned integers-Wsign-promoWarn when overload promotes from unsigned to signed-Wsizeof-pointer-memaccess-Wspecial-param-hiding-Wstack-protectorWarn when not issuing stack smashing protection for some reason-Wstack-usage=Warn if stack usage might be larger than specified amount-Wstatic-access-Wstatic-receiver-Wstrict-aliasingWarn about code which might break strict aliasing rules-Wstrict-aliasing=-Wstrict-null-sentinelWarn about uncasted NULL used as sentinel-Wstrict-overflowWarn about optimizations that assume that signed overflow is undefined-Wstrict-overflow=-Wstrict-prototypesWarn about unprototyped function declarations-Wstrict-selector-matchWarn if type signatures of candidate methods do not match exactly-Wsuggest-attribute=constWarn about functions which might be candidates for __attribute__((const))-Wsuggest-attribute=formatWarn about functions which might be candidates for format attributes-Wsuggest-attribute=noreturnWarn about functions which might be candidates for __attribute__((noreturn))-Wsuggest-attribute=pureWarn about functions which might be candidates for __attribute__((pure))-Wsuppress-WsurprisingWarn about "suspicious" constructs-WswitchWarn about enumerated switches, with no default, missing a case-Wswitch-defaultWarn about enumerated switches missing a "default:" statement-Wswitch-enumWarn about all enumerated switches missing a specific case-Wsync-nandWarn when __sync_fetch_and_nand and __sync_nand_and_fetch built-in functions are used-WsynthDeprecated. This switch has no effect-Wsynthetic-access-Wsystem-headersDo not suppress warnings from system headers-WtabsPermit nonconforming uses of the tab character-Wtarget-lifetimeWarn if the pointer in a pointer assignment might outlive its target-Wtasks-Wthread-mismatched-lock-acq-relWarn about mismatched lock acquisition and release-Wthread-mismatched-lock-orderWarn about lock acquisition order inconsistent with what specified in the attributes-Wthread-reentrant-lockWarn about a lock being acquired recursively-Wthread-safetyWarn about potential thread safety issues when the code is annotated with thread safety attributes-Wthread-safety-analysisDoes nothing. For compatibility with clang thread safety analysis.-Wthread-safety-attributes-Wthread-safety-precise-Wthread-unguarded-funcWarn about function calls not properly protected by locks specified in the attributes-Wthread-unguarded-varWarn about shared variables not properly protected by locks specified in the attributes-Wthread-unsupported-lock-nameWarn about uses of unsupported lock names in attributes-WtraditionalWarn about features not present in traditional C-Wtraditional-conversionWarn of prototypes causing type conversions different from what would happen in the absence of prototype-WtrampolinesWarn whenever a trampoline is generated-WtrigraphsWarn if trigraphs are encountered that might affect the meaning of the program-Wtype-hiding-Wtype-limitsWarn if a comparison is always true or always false due to the limited range of the data type-Wuncheck-Wundeclared-selectorWarn about @selector()s without previously declared methods-WundefWarn if an undefined macro is used in an #if directive-WunderflowWarn about underflow of numerical constant expressions-WuninitializedWarn about uninitialized automatic variables-Wunknown-pragmasWarn about unrecognized pragmas-Wunnecessary-else-Wunqualified-field-Wunreachable-codeDoes nothing. Preserved for backward compatibility.-Wunsafe-loop-optimizationsWarn if the loop cannot be optimized due to nontrivial assumptions.-Wunsuffixed-float-constantsWarn about unsuffixed float constants-WunusedEnable all -Wunused- warnings-Wunused-argument-Wunused-but-set-parameterWarn when a function parameter is only set, otherwise unused-Wunused-but-set-variableWarn when a variable is only set, otherwise unused-Wunused-dummy-argumentWarn about unused dummy arguments.-Wunused-functionWarn when a function is unused-Wunused-import-Wunused-labelWarn when a label is unused-Wunused-local-Wunused-local-typedefsWarn when typedefs locally defined in a function are not used-Wunused-macrosWarn about macros defined in the main file that are not used-Wunused-parameterWarn when a function parameter is unused-Wunused-private-Wunused-resultWarn if a caller of a function, marked with attribute warn_unused_result, does not use its return value-Wunused-thrown-Wunused-valueWarn when an expression value is unused-Wunused-variableWarn when a variable is unused-Wuseless-castWarn about useless casts-Wuseless-type-check-WvarargsWarn about questionable usage of the macros used to retrieve variable arguments-Wvarargs-cast-Wvariadic-macrosWarn about using variadic macros-Wvector-operation-performanceWarn when a vector operation is compiled outside the SIMD-Wvirtual-move-assignWarn if a virtual base has a non-trivial move assignment operator-WvlaWarn if a variable length array is used-Wvolatile-register-varWarn when a register variable is declared volatile-Wwarning-token-Wwrite-stringsIn C++, nonzero means warn about deprecated conversion from string literals to 'char *'. In C, similar warning, except that the conversion is of course not deprecated by the ISO C standard.-Wzero-as-null-pointer-constantWarn when a literal '0' is used as null pointer-WzerotripWarn about zero-trip DO loops-Xassembler-Xclang-only=-Xlinker-Xpreprocessor-Z-ansiA synonym for -std=c89 (for C) or -std=c++98 (for C++)-aux-info-aux-info <file> Emit declaration information into <file>-aux-info=-auxbase-auxbase-strip-bootclasspath-c-canonical-prefixes-classpath-coverage-cppEnable preprocessing-cpp=-d-d<letters> Enable dumps from specific passes of the compiler-dumpbase-dumpbase <file> Set the file basename to be used for dumps-dumpdir-dumpdir <dir> Set the directory name to be used for dumps-dumpmachine-dumpspecs-dumpversion-e-encoding-export-dynamic-extdirs-fCLASSPATH=--CLASSPATH Deprecated; use --classpath instead-fPICGenerate position-independent code if possible (large mode)-fPIEGenerate position-independent code for executables if possible (large mode)-fRTS=Select the runtime-fabi-version=-faccess-controlEnforce class member access control semantics-fada-spec-parent=-fada-spec-parent=unit Dump Ada specs as child units of given parent-faggressive-function-eliminationEliminate multiple function invokations also for impure functions-faggressive-loop-optimizationsAggressively optimize loops using language constraints-falign-commonsEnable alignment of COMMON blocks-falign-functionsAlign the start of functions-falign-functions=-falign-jumpsAlign labels which are only reached by jumping-falign-jumps=-falign-labelsAlign all labels-falign-labels=-falign-loopsAlign the start of loops-falign-loops=-fall-intrinsicsAll intrinsics procedures are available regardless of selected standard-fall-virtual-fallow-leading-underscore-fallow-parameterless-variadic-functionsAllow variadic functions without named parameter-falt-external-templatesNo longer supported-fargument-alias-fargument-noalias-fargument-noalias-anything-fargument-noalias-global-fasmRecognize the "asm" keyword-fassertPermit the use of the assert keyword-fassociative-mathAllow optimization for floating-point arithmetic which may change the result of the operation due to rounding.-fassume-compiled-fassume-compiled=-fasynchronous-unwind-tablesGenerate unwind tables that are exact at each instruction boundary-fauto-inc-decGenerate auto-inc/dec instructions-fauto-profileUse sample profile information for call graph node weights. The default profile file is fbdata.afdo in 'pwd'.-fauto-profile-accurateWhether to assume the sample profile is accurate.-fauto-profile-record-coverage-in-elfWhether to record annotation coverage info in elf.-fauto-profile=Use sample profile information for call graph node weights. The profile file is specified in the argument.-fautomaticDo not treat local variables and COMMON blocks as if they were named in SAVE statements-faux-classpath-fbackslashSpecify that backslash in string introduces an escape character-fbacktraceProduce a backtrace when a runtime error is encountered-fblas-matmul-limit=-fblas-matmul-limit=<n> Size of the smallest matrix for which matmul will use BLAS-fbootclasspath=--bootclasspath=<path> Replace system path-fbootstrap-classesGenerated should be loaded by bootstrap loader-fbounds-checkGenerate code to check bounds before indexing arrays-fbranch-count-regReplace add, compare, branch with branch on count register-fbranch-probabilitiesUse profiling information for branch probabilities-fbranch-target-load-optimizePerform branch target load optimization before prologue / epilogue threading-fbranch-target-load-optimize2Perform branch target load optimization after prologue / epilogue threading-fbtr-bb-exclusiveRestrict target load migration not to re-use registers in any basic block-fbuilding-libgcc-fbuiltinRecognize built-in functions-fbuiltin--fcall-saved--fcall-saved-<register> Mark <register> as being preserved across functions-fcall-used--fcall-used-<register> Mark <register> as being corrupted by function calls-fcaller-savesSave registers around function calls-fcanonical-system-headersWhere shorter, use canonicalized paths to systems headers.-fcheck-array-temporariesProduce a warning at runtime if a array temporary has been created for a procedure argument-fcheck-branch-annotationCompare branch prediction result and autofdo profile information, store the result in a section in the generated elf file.-fcheck-branch-annotation-threshold=The number of executions a basic block needs to reach before GCC dumps its branch prediction information with -fcheck-branch-annotation.-fcheck-data-depsCompare the results of several data dependence analyzers.-fcheck-newCheck the return value of new in C++-fcheck-referencesGenerate checks for references to NULL-fcheck=-fcheck=[...] Specify which runtime checks are to be performed-fcilkplusEnable Cilk Plus-fclasspath=--classpath=<path> Set class path-fcoarray=-fcoarray=[...] Specify which coarray parallelization should be used-fcombine-stack-adjustmentsLooks for opportunities to reduce stack adjustments and stack references.-fcommonDo not put uninitialized globals in the common section-fcompare-debug-fcompare-debug-secondRun only the second compilation of -fcompare-debug-fcompare-debug[=<opts>] Compile with and without e.g. -gtoggle, and compare the final-insns dump-fcompare-elimPerform comparison elimination after register allocation has finished-fcompile-resource=-fcond-mismatchAllow the arguments of the '?' operator to have different types-fconserve-spaceDoes nothing. Preserved for backward compatibility.-fconserve-stackDo not perform optimizations increasing noticeably stack usage-fconstant-string-class=-fconst-string-class=<name> Use class <name> for constant stringsno class name specified with %qs-fconstexpr-depth=-fconstexpr-depth=<number> Specify maximum constexpr recursion depth-fconvert=big-endianUse big-endian format for unformatted files-fconvert=little-endianUse little-endian format for unformatted files-fconvert=nativeUse native format for unformatted files-fconvert=swapSwap endianness for unformatted files-fcprop-registersPerform a register copy-propagation optimization pass-fcray-pointerUse the Cray Pointer extension-fcrossjumpingPerform cross-jumping optimization-fcse-follow-jumpsWhen running CSE, follow jumps to their targets-fcse-skip-blocks-fcx-fortran-rulesComplex multiplication and division follow Fortran rules-fcx-limited-rangeOmit range reduction step when performing complex division-fd-lines-as-codeIgnore 'D' in column one in fixed form-fd-lines-as-commentsTreat lines with 'D' in column one as comments-fdata-sectionsPlace data items into their own section-fdbg-cnt-listList all available debugging counters with their limits and counts.-fdbg-cnt=-fdbg-cnt=<counter>:<limit>[,<counter>:<limit>,...] Set the debug counter limit. -fdceUse the RTL dead code elimination pass-fdebug-cppEmit debug annotations during preprocessing-fdebug-prefix-map=Map one directory name to another in debug information-fdebug-types-sectionOutput .debug_types section when using DWARF v4 debuginfo.-fdeclone-ctor-dtorFactor complex constructors and destructors to favor space over speed-fdeduce-init-list-fdeduce-init-list enable deduction of std::initializer_list for a template type parameter from a brace-enclosed initializer-list-fdefault-double-8Set the default double precision kind to an 8 byte wide type-fdefault-inline-fdefault-integer-8Set the default integer kind to an 8 byte wide type-fdefault-noexceptMake functions no-throw/noexcept by default-fdefault-real-8Set the default real kind to an 8 byte wide type-fdefer-popDefer popping functions args from stack until later-fdelayed-branchAttempt to fill delay slots of branch instructions-fdelete-dead-exceptionsDelete dead instructions that may throw exceptions-fdelete-null-pointer-checksDelete useless null pointer checks-fdevirtualizeTry to convert virtual calls to direct ones.-fdevirtualize-speculativelyPerform speculative devirtualization-fdiagnostics-coloralwaysnever-fdiagnostics-color=-fdiagnostics-color=[never|always|auto] Colorize diagnostics-fdiagnostics-show-caretShow the source line with a caret indicating the column-fdiagnostics-show-location=-fdiagnostics-show-location=[once|every-line] How often to emit source location at the beginning of line-wrapped diagnostics-fdiagnostics-show-optionAmend appropriate diagnostic messages with the command line option that controls them-fdirectives-onlyPreprocess directives only.-fdisable--fdisable-[tree|rtl|ipa]-<pass>=range1+range2 disables an optimization pass-fdisable-assertions-fdisable-assertions=-fdollar-okAllow dollar signs in entity names-fdollars-in-identifiersPermit '$' as an identifier character-fdseUse the RTL dead store elimination pass-fdump--fdump-<type> Dump various compiler internals to a file-fdump-ada-specWrite all declarations as Ada code transitively-fdump-ada-spec-slimWrite all declarations as Ada code for the given file only-fdump-core-fdump-final-insns-fdump-final-insns=-fdump-final-insns=filename Dump to filename the insns at the end of translation-fdump-fortran-optimizedDisplay the code tree after front end optimization-fdump-fortran-originalDisplay the code tree after parsing-fdump-go-spec=-fdump-go-spec=filename Write all declarations to file as Go code-fdump-noaddrSuppress output of addresses in debugging dumps-fdump-parse-treeDisplay the code tree after parsing; deprecated option-fdump-passesDump optimization passes-fdump-unnumberedSuppress output of instruction numbers, line number notes and addresses in debugging dumps-fdump-unnumbered-linksSuppress output of previous and next insn numbers in debugging dumps-fdwarf2-cfi-asmEnable CFI tables via GAS assembler directives.-fearly-inliningPerform early inlining-felide-constructors-feliminate-dwarf2-dupsPerform DWARF2 duplicate elimination-feliminate-unused-debug-symbolsPerform unused type elimination in debug info-feliminate-unused-debug-types-femit-class-debug-alwaysDo not suppress C++ class debug information.-femit-class-fileOutput a class file-femit-class-filesAlias for -femit-class-file-femit-function-namesPrint to stderr the mapping from module name and function id to assembler function name when -ftest-coverage, -fprofile-generate or -fprofile-use are active, for use in correlating function ids in gcda files with the function name.-femit-struct-debug-baseonly-femit-struct-debug-baseonly Aggressive reduced debug info for structs-femit-struct-debug-detailed=-femit-struct-debug-detailed=<spec-list> Detailed reduced debug info for structs-femit-struct-debug-reduced-femit-struct-debug-reduced Conservative reduced debug info for structs-fenable--fenable-[tree|rtl|ipa]-<pass>=range1+range2 enables an optimization pass-fenable-assertions-fenable-assertions=-fencoding=--encoding=<encoding> Choose input encoding (defaults from your locale)-fenforce-eh-specsGenerate code to check exception specifications-fenum-int-equiv-fexceptionsEnable exception handling-fexcess-precision=-fexcess-precision=[fast|standard] Specify handling of excess floating-point precision-fexec-charset=-fexec-charset=<cset> Convert all strings and character constants to character set <cset>-fexpensive-optimizationsPerform a number of minor, expensive optimizations-fext-numeric-literalsInterpret imaginary, fixed-point, or other gnu number suffix as the corresponding number literal rather than a user-defined number literal.-fextdirs=--extdirs=<path> Set the extension directory path-fextended-identifiersPermit universal character names (\u and \U) in identifiers-fextern-tls-initSupport dynamic initialization of thread-local variables in a different translation unit-fexternal-blasSpecify that an external BLAS library should be used for matmul calls on large-size arrays-fexternal-templates-ff2cUse f2c calling convention-ffast-math-ffat-lto-objectsOutput lto objects containing both the intermediate language and binary output.-ffilelist-fileInput file is a file with a list of filenames to compile-ffinite-math-onlyAssume no NaNs or infinities are generated-ffixed--ffixed-<register> Mark <register> as being unavailable to the compiler-ffixed-formAssume that the source file is fixed form-ffixed-line-length--ffixed-line-length-<n> Use n as character line width in fixed mode-ffixed-line-length-noneAllow arbitrary character line width in fixed mode-ffloat-storeDon't allocate floats and doubles in extended-precision registers-ffor-scopeScope of for-init-statement variables is local to the loop-fforce-addr-fforce-classes-archive-checkAlways check for non gcj generated classes archives-fforward-propagatePerform a forward propagation pass on RTL-ffp-contract=-ffp-contract=[off|on|fast] Perform floating-point expression contraction.-ffpe-summary=-ffpe-summary=[...] Print summary of floating point exceptions-ffpe-trap=-ffpe-trap=[...] Stop on following floating point exceptions-ffree-formAssume that the source file is free form-ffree-line-length--ffree-line-length-<n> Use n as character line width in free mode-ffree-line-length-noneAllow arbitrary character line width in free mode-ffreestandingDo not assume that standard C libraries and "main" exist-ffriend-injectionInject friend functions into enclosing namespace-ffrontend-optimizeEnable front end optimization-ffunction-attribute-list=-ffunction-attribute-list=attribute:name,... Add attribute to named functions-ffunction-cseAllow function addresses to be held in registers-ffunction-sectionsPlace each function into its own section-fgcsePerform global common subexpression elimination-fgcse-after-reloadPerform global common subexpression elimination after register allocation has finished-fgcse-lasPerform redundant load after store elimination in global common subexpression elimination-fgcse-lmPerform enhanced load motion during global common subexpression elimination-fgcse-smPerform store motion after global common subexpression elimination-fgnu-keywordsRecognize GNU-defined keywords-fgnu-runtimeGenerate code for GNU runtime environment-fgnu-tmEnable support for GNU transactional memory-fgnu-uniqueUse STB_GNU_UNIQUE if supported by the assembler-fgnu89-inlineUse traditional GNU semantics for inline functions-fgo-check-divide-overflowAdd explicit checks for division overflow in INT_MIN / -1-fgo-check-divide-zeroAdd explicit checks for division by zero-fgo-dump--fgo-dump-<type> Dump Go frontend internal information-fgo-optimize--fgo-optimize-<type> Turn on optimization passes in the frontend-fgo-pkgpath=-fgo-pkgpath=<string> Set Go package path-fgo-prefix=-fgo-prefix=<string> Set package-specific prefix for exported Go names-fgo-relative-import-path=-fgo-relative-import-path=<path> Treat a relative import as relative to path-fgraphiteEnable in and out of Graphite representation-fgraphite-identityEnable Graphite Identity transformation-fguess-branch-probabilityEnable guessing of branch probabilities-fguiding-decls-fhandle-exceptions-fhandle-exceptions has been renamed -fexceptions (and is now on by default)-fhash-synchronizationAssume the runtime uses a hash table to map an object to its synchronization structure-fhelp-fhelp=-fhoist-adjacent-loadsEnable hoisting adjacent loads to encourage generating conditional move instructions-fhonor-std-fhostedAssume normal C execution environment-fhuge-objects-fidentProcess #ident directives-fif-conversionPerform conversion of conditional jumps to branchless equivalents-fif-conversion2Perform conversion of conditional jumps to conditional execution-fimplement-inlinesExport functions even if they can be inlined-fimplicit-inline-templatesEmit implicit instantiations of inline templates-fimplicit-noneSpecify that no implicit typing is allowed, unless overridden by explicit IMPLICIT statements-fimplicit-templatesEmit implicit instantiations of templates-findirect-classesGenerate instances of Class at runtime-findirect-dispatchUse offset tables for virtual method calls-findirect-inliningPerform indirect inlining-finhibit-size-directiveDo not generate .size directives-finit-character=-finit-character=<n> Initialize local character variables to ASCII value n-finit-integer=-finit-integer=<n> Initialize local integer variables to n-finit-local-zeroInitialize local variables to zero (from g77)-finit-logical=-finit-logical=<true|false> Initialize local logical variables-finit-real=-finit-real=<zero|nan|inf|-inf> Initialize local real variables-finlineEnable inlining of function declared "inline", disabling disables all inlining-finline-atomicsInline __atomic operations when a lock free instruction sequence is available.-finline-functionsIntegrate functions not declared "inline" into their callers when profitable-finline-functions-called-onceIntegrate functions only required by their single caller-finline-limit--finline-limit=-finline-limit=<number> Limit the size of inlined functions to <number>-finline-small-functionsIntegrate functions into their callers when code size is known not to grow-finput-charset=-finput-charset=<cset> Specify the default character set for source files-finstrument-functionsInstrument function entry and exit with profiling calls-finstrument-functions-exclude-file-list=-finstrument-functions-exclude-file-list=filename,... Do not instrument functions listed in files-finstrument-functions-exclude-function-list=-finstrument-functions-exclude-function-list=name,... Do not instrument listed functions-finteger-4-integer-8Interpret any INTEGER(4) as an INTEGER(8)-fintrinsic-modules-pathSpecify where to find the compiled intrinsic modules-fintrinsic-modules-path=-fipa-cpPerform interprocedural constant propagation-fipa-cp-clonePerform cloning to make Interprocedural constant propagation stronger-fipa-matrix-reorg-fipa-profilePerform interprocedural profile propagation-fipa-ptaPerform interprocedural points-to analysis-fipa-pure-constDiscover pure and const functions-fipa-referenceDiscover readonly and non addressable static variables-fipa-sraPerform interprocedural reduction of aggregates-fipa-struct-reorg-fira-algorithm=-fira-algorithm=[CB|priority] Set the used IRA algorithm-fira-hoist-pressureUse IRA based register pressure calculation in RTL hoist optimizations.-fira-loop-pressureUse IRA based register pressure calculation in RTL loop optimizations.-fira-region=-fira-region=[one|all|mixed] Set regions for IRA-fira-share-save-slotsShare slots for saving different hard registers.-fira-share-spill-slotsShare stack slots for spilled pseudo-registers.-fira-verbose=-fira-verbose=<number> Control IRA's level of diagnostic messages.-fisolate-erroneous-paths-attributeDetect paths which trigger erroneous or undefined behaviour due a NULL value being used in a way which is forbidden by a returns_nonnull or nonnull attribute. Isolate those paths from the main control flow and turn the statement with erroneous or undefined behaviour into a trap. -fisolate-erroneous-paths-dereferenceDetect paths which trigger erroneous or undefined behaviour due to dereferencing a NULL pointer. Isolate those paths from the main control flow and turn the statement with erroneous or undefined behaviour into a trap.-fivoptsOptimize induction variables on trees-fjniAssume native functions are implemented using JNI-fjump-tablesUse jump tables for sufficiently large switch statements-fkeep-inline-dllexportDon't emit dllexported inline functions unless needed-fkeep-inline-functionsGenerate code for functions even if they are fully inlined-fkeep-static-constsEmit static const variables even if they are not used-flabels-ok-flax-vector-conversionsAllow implicit conversions between vectors with differing numbers of subparts and/or differing element types.-fleading-underscoreGive external symbols a leading underscore-flifetime-dseTell DSE that the storage for a C++ object is dead when the constructor starts and when the destructor finishes.-flive-range-shrinkageRelief of register pressure through live range shrinkage-floop-blockEnable Loop Blocking transformation-floop-flatten-floop-interchangeEnable Loop Interchange transformation-floop-nest-optimizeEnable the ISL based loop nest optimizer-floop-optimize-floop-parallelize-allMark all loops as parallel-floop-strip-mineEnable Loop Strip Mining transformation-fltoEnable link-time optimization.-flto-compression-level=-flto-compression-level=<number> Use zlib compression level <number> for IL-flto-partition=1to1Partition symbols and vars at linktime based on object files they originate from-flto-partition=balancedPartition functions and vars at linktime into approximately same sized buckets-flto-partition=maxPut every symbol into separate partition-flto-partition=noneDisable partioning and streaming-flto-reportReport various link-time optimization statistics-flto-report-wpaReport various link-time optimization statistics for WPA only-flto=Link-time optimization with number of parallel jobs or jobserver.-fltransRun the link-time optimizer in local transformation (LTRANS) mode.-fltrans-output-list=Specify a file to which a list of files output by LTRANS is written.-fmain=-fmath-errnoSet errno after built-in math functions-fmax-array-constructor=-fmax-array-constructor=<n> Maximum number of objects in an array constructor-fmax-errors=-fmax-errors=<number> Maximum number of errors to report-fmax-identifier-length=-fmax-identifier-length=<n> Maximum identifier length-fmax-stack-var-size=-fmax-stack-var-size=<n> Size in bytes of the largest array that will be put on the stack-fmax-subrecord-length=-fmax-subrecord-length=<n> Maximum length for subrecords-fmem-reportReport on permanent memory allocation-fmem-report-wpaReport on permanent memory allocation in WPA only-fmerge-all-constantsAttempt to merge identical constants and constant variables-fmerge-constantsAttempt to merge identical constants across compilation units-fmerge-debug-stringsAttempt to merge identical debug strings across compilation units-fmessage-length=-fmessage-length=<number> Limit diagnostics to <number> characters per line. 0 suppresses line-wrapping-fmodule-privateSet default accessibility of module entities to PRIVATE.-fmodulo-schedPerform SMS based modulo scheduling before the first scheduling pass-fmodulo-sched-allow-regmovesPerform SMS based modulo scheduling with register moves allowed-fmove-loop-invariantsMove loop invariant computations out of loops-fms-extensionsDon't warn about uses of Microsoft extensions-fmudflap-fmudflapir-fmudflapth-fname-mangling-version--fnew-abi-fnext-runtimeGenerate code for NeXT (Apple Mac OS X) runtime environment-fnil-receiversAssume that receivers of Objective-C messages may be nil-fno-vect-cost-modelEnables the unlimited vectorizer cost model. Preserved for backward compatibility.unlimited-fnon-call-exceptionsSupport synchronous non-call exceptions-fnonansi-builtins-fnonnull-objects-fnothrow-optTreat a throw() exception specification as noexcept to improve code size-fobjc-abi-version=Specify which ABI to use for Objective-C family code and meta-data generation.-fobjc-call-cxx-cdtorsGenerate special Objective-C methods to initialize/destroy non-POD C++ ivars, if needed-fobjc-direct-dispatchAllow fast jumps to the message dispatcher-fobjc-exceptionsEnable Objective-C exception and synchronization syntax-fobjc-gcEnable garbage collection (GC) in Objective-C/Objective-C++ programs-fobjc-nilcheckEnable inline checks for nil receivers with the NeXT runtime and ABI version 2.-fobjc-sjlj-exceptionsEnable Objective-C setjmp exception handling runtime-fobjc-std=objc1Conform to the Objective-C 1.0 language as implemented in GCC 4.0-fomit-frame-pointerWhen possible do not generate stack frames-fopenmpEnable OpenMP (implies -frecursive in Fortran)-fopenmp-simdEnable OpenMP's SIMD directives-foperator-namesRecognize C++ keywords like "compl" and "xor"-fopt-infoEnable all optimization info dumps on stderr-fopt-info--fopt-info[-<type>=filename] Dump compiler optimization details-foptimize-register-move-foptimize-sibling-callsOptimize sibling and tail recursive calls-foptimize-static-class-initializationEnable optimization of static class initialization code-foptimize-strlenEnable string length optimizations on trees-foptional-diags-foutput-class-dir=-fpack-derivedTry to lay out derived types as compactly as possible-fpack-structPack structure members together without holes-fpack-struct=-fpack-struct=<number> Set initial maximum structure member alignment-fpartial-inliningPerform partial inlining-fpcc-struct-returnReturn small aggregates in memory, not registers-fpch-deps-fpch-preprocessLook for and use PCH files even when preprocessing-fpeel-codesize-limitLimit non-const non-FP loop peeling under profile estimates of large code footprint-fpeel-loopsPerform loop peeling-fpeepholeEnable machine specific peephole optimizations-fpeephole2Enable an RTL peephole pass before sched2-fpermissiveDowngrade conformance errors to warnings-fpicGenerate position-independent code if possible (small mode)-fpieGenerate position-independent code for executables if possible (small mode)-fplan9-extensionsEnable Plan 9 language extensions-fpltUse PLT for PIC calls (-fno-plt: load the address from GOT at call site)-fplugin-arg--fplugin-arg-<name>-<key>[=<value>] Specify argument <key>=<value> for plugin <name>-fplugin=Specify a plugin to load-fpost-ipa-mem-reportReport on memory allocation before interprocedural optimization-fpre-ipa-mem-report-fpredictive-commoningRun predictive commoning optimization.-fprefetch-loop-arraysGenerate prefetch instructions, if available, for arrays in loops-fpreprocessedTreat the input file as already preprocessed-fpretty-templates-fno-pretty-templates Do not pretty-print template specializations as the template signature followed by the arguments-fprofileEnable basic program profiling code-fprofile-arcsInsert arc-based program profiling code-fprofile-correctionEnable correction of flow inconsistent profile data input-fprofile-dir=Set the top-level directory for storing the profile data. The default is 'pwd'.-fprofile-dumpDump CFG profile for comparison.-fprofile-generateEnable common options for generating profile info for profile feedback directed optimizations-fprofile-generate-atomic=fprofile-generate-atomic=[0..3] Atomically increments for profile counters.-fprofile-generate-buildinfo=-fprofile-generate-buildinfo=filename Read build info to include in gcda file from filename-fprofile-generate-samplingTurn on instrumentation sampling with -fprofile-generate with rate set by --param profile-generate-sampling-rate or environment variable GCOV_SAMPLING_RATE-fprofile-generate=Enable common options for generating profile info for profile feedback directed optimizations, and set -fprofile-dir=-fprofile-reorder-functionsEnable function reordering that improves code placement-fprofile-reportReport on consistency of profile-fprofile-strip=Specify a substring to be stripped from the profile base file name-fprofile-useEnable common options for performing profile feedback directed optimizations-fprofile-use=Enable common options for performing profile feedback directed optimizations, and set -fprofile-dir=-fprofile-valuesInsert code to profile values of expressions-fprotect-parensProtect parentheses in expressions-frandom-seed-frandom-seed=-frandom-seed=<string> Make compile reproducible using <string>-frange-checkEnable range checking during compilation-freal-4-real-10Interpret any REAL(4) as a REAL(10)-freal-4-real-16Interpret any REAL(4) as a REAL(16)-freal-4-real-8Interpret any REAL(4) as a REAL(8)-freal-8-real-10Interpret any REAL(8) as a REAL(10)-freal-8-real-16Interpret any REAL(8) as a REAL(16)-freal-8-real-4Interpret any REAL(8) as a REAL(4)-frealloc-lhsReallocate the LHS in assignments-freciprocal-mathSame as -fassociative-math for expressions which include division.-frecord-compilation-info-in-elfRecord the compiler optimizations in a .gnu.switches.text section.-frecord-gcc-switchesRecord gcc command line switches in the object file.-frecord-marker=4Use a 4-byte record marker for unformatted files-frecord-marker=8Use an 8-byte record marker for unformatted files-frecursiveAllocate local variables on the stack to allow indirect recursion-freduced-reflectionReduce the amount of reflection meta-data generated-freeTurn on Redundant Extensions Elimination pass.-freg-struct-returnReturn small aggregates in registers-fregmove-frename-registersPerform a register renaming optimization pass-freorder-blocksReorder basic blocks to improve code placement-freorder-blocks-and-partitionReorder basic blocks and partition into hot and cold sections-freorder-functionsReorder functions to improve code placement-freorder-functions=-freorder-functions=[callgraph] Select the scheme for function reordering. This invokes a linker plugin. Generate .gnu.callgraph.text sections listing callees and edge counts.-frepack-arraysCopy array sections into a contiguous block on procedure entry-freplace-objc-classesUsed in Fix-and-Continue mode to indicate that object files may be swapped in at runtime-frepoEnable automatic template instantiation-frequire-return-statementFunctions which return values must end with return statements-frerun-cse-after-loopAdd a common subexpression elimination pass after loop optimizations-frerun-loop-opt-freschedule-modulo-scheduled-loopsEnable/Disable the traditional scheduling in loops that already passed modulo scheduling-fresolution=The resolution file-fripaPerform Dynamic Inter-Procedural Analysis.-fripa-allow-debugAllow -g enablement for -fripa -fprofile-generate compiles.-fripa-disallow-asm-modulesDon't import an auxiliary module if it contains asm statements-fripa-disallow-opt-mismatchDon't import an auxiliary module if the command line options mismatch with the primary module-fripa-inc-path-sub=Substitute substring in include paths with a new string to allow reuse profile data-fripa-no-promote-always-inline-funcDon't promote always inline static functions assuming they will be inlined and no copy is needed.-frounding-mathDisable optimizations that assume default FP rounding behavior-frttiGenerate run time type descriptor information-fsanitize-coverage=trace-pcEnable coverage-guided fuzzing code instrumentation. Inserts call to __sanitizer_cov_trace_pc into every basic block.-fsanitize=Select what to sanitize-fsaw-java-file-fsched-critical-path-heuristicEnable the critical path heuristic in the scheduler-fsched-dep-count-heuristicEnable the dependent count heuristic in the scheduler-fsched-group-heuristicEnable the group heuristic in the scheduler-fsched-interblockEnable scheduling across basic blocks-fsched-last-insn-heuristicEnable the last instruction heuristic in the scheduler-fsched-pressureEnable register pressure sensitive insn scheduling-fsched-rank-heuristicEnable the rank heuristic in the scheduler-fsched-specAllow speculative motion of non-loads-fsched-spec-insn-heuristicEnable the speculative instruction heuristic in the scheduler-fsched-spec-loadAllow speculative motion of some loads-fsched-spec-load-dangerousAllow speculative motion of more loads-fsched-stalled-insnsAllow premature scheduling of queued insns-fsched-stalled-insns-depSet dependence distance checking in premature scheduling of queued insns-fsched-stalled-insns-dep=-fsched-stalled-insns-dep=<number> Set dependence distance checking in premature scheduling of queued insns-fsched-stalled-insns=-fsched-stalled-insns=<number> Set number of queued insns that can be prematurely scheduled-fsched-verbose=-fsched-verbose=<number> Set the verbosity level of the scheduler-fsched2-use-superblocksIf scheduling post reload, do superblock scheduling-fsched2-use-traces-fschedule-insnsReschedule instructions before register allocation-fschedule-insns2Reschedule instructions after register allocation-fsecond-underscoreAppend a second underscore if the name already contains an underscore-fsection-anchorsAccess data in the same section from shared anchor points-fsee-fsel-sched-pipeliningPerform software pipelining of inner loops during selective scheduling-fsel-sched-pipelining-outer-loopsPerform software pipelining of outer loops during selective scheduling-fsel-sched-reschedule-pipelinedReschedule pipelined regions without pipelining-fselective-schedulingSchedule instructions using selective scheduling algorithm-fselective-scheduling2Run selective scheduling after reload-fshort-doubleUse the same size for double as for float-fshort-enumsUse the narrowest integer type possible for enumeration types-fshort-wcharForce the underlying type for "wchar_t" to be "unsigned short"-fshow-columnShow column numbers in diagnostics, when available. Default on-fshrink-wrapEmit function prologues only before parts of the function that need it, rather than at the top of the function.-fshrink-wrap-frame-pointerFramepointer shrinkwrapping optimization.-fsign-zeroApply negative sign to zero values-fsignaling-nansDisable optimizations observable by IEEE signaling NaNs-fsigned-bitfieldsWhen "signed" or "unsigned" is not given make the bitfield signed-fsigned-charMake "char" signed by default-fsigned-zerosDisable floating point optimizations that ignore the IEEE signedness of zero-fsimd-cost-model=Specifies the vectorization cost model for code marked with a simd directive-fsingle-precision-constantConvert floating point constants to single precision constants-fsized-deleteSupport delete operator with objetc's size as the second parameter.-fsource-filename=-fsource=Set the source language version-fsplit-ivs-in-unrollerSplit lifetimes of induction variables when loops are unrolled-fsplit-stackGenerate discontiguous stack frames-fsplit-wide-typesSplit wide types into independent registers-fsquangle-fstack-arraysPut all local arrays on stack.-fstack-checkInsert stack checking code into the program. Same as -fstack-check=specific-fstack-check=-fstack-check=[no|generic|specific] Insert stack checking code into the program-fstack-limit-fstack-limit-register=-fstack-limit-register=<register> Trap if the stack goes past <register>-fstack-limit-symbol=-fstack-limit-symbol=<name> Trap if the stack goes past symbol <name>-fstack-protectorUse propolice as a stack protection method-fstack-protector-allUse a stack protection method for every function-fstack-protector-strongUse a smart stack protection method for certain functions-fstack-reuse=-fstack-reuse=[all|named_vars|none] Set stack reuse level for local variables.-fstack-usageOutput stack usage information on a per-function basis-fstatsDisplay statistics accumulated during compilation-fstore-checkEnable assignability checks for stores into object arrays-fstrength-reduce-fstrict-aliasingAssume strict aliasing rules apply-fstrict-enum-precisionPerform transformations based on enum precision-fstrict-enumsAssume that values of enumeration type are always within the minimum range of that type-fstrict-overflowTreat signed overflow as undefined-fstrict-prototype-fstrict-volatile-bitfieldsForce bitfield accesses to match their type width-fsync-libcallsImplement __atomic operations via libcalls to legacy __sync functions-fsyntax-onlyCheck for syntax errors, then stop-ftabstop=-ftabstop=<number> Distance between tab stops for column reporting-ftarget-help-ftarget=Set the target VM version-ftemplate-backtrace-limit=Set the maximum number of template instantiation notes for a single warning or error-ftemplate-depth--ftemplate-depth=-ftemplate-depth=<number> Specify maximum template instantiation depth-ftest-coverageCreate data files needed by "gcov"-fthis-is-variable-fthread-jumpsPerform jump threading optimizations-fthreadsafe-statics-fno-threadsafe-statics Do not generate thread-safe code for initializing local statics-ftime-reportReport the time taken by each compiler pass-ftls-model=-ftls-model=[global-dynamic|local-dynamic|initial-exec|local-exec] Set the default thread-local storage code generation model-ftoplevel-reorderReorder top level functions, variables, and asms-ftracerPerform superblock formation via tail duplication-ftrack-macro-expansion-ftrack-macro-expansion=-ftrack-macro-expansion=<0|1|2> Track locations of tokens coming from macro expansion and display them in error messages-ftrapping-mathAssume floating-point operations can trap-ftrapvTrap for signed overflow in addition, subtraction and multiplication-ftree-bit-ccpEnable SSA-BIT-CCP optimization on trees-ftree-builtin-call-dceEnable conditional dead code elimination for builtin calls-ftree-ccpEnable SSA-CCP optimization on trees-ftree-chEnable loop header copying on trees-ftree-coalesce-inlined-varsEnable coalescing of copy-related user variables that are inlined-ftree-coalesce-varsEnable coalescing of all copy-related user variables-ftree-copy-propEnable copy propagation on trees-ftree-copyrenameReplace SSA temporaries with better names in copies-ftree-cselimTransform condition stores into unconditional ones-ftree-dceEnable SSA dead code elimination optimization on trees-ftree-dominator-optsEnable dominator optimizations-ftree-dseEnable dead store elimination-ftree-forwpropEnable forward propagation on trees-ftree-freEnable Full Redundancy Elimination (FRE) on trees-ftree-loop-distribute-patternsEnable loop distribution for patterns transformed into a library call-ftree-loop-distributionEnable loop distribution on trees-ftree-loop-if-convertConvert conditional jumps in innermost loops to branchless equivalents-ftree-loop-if-convert-storesAlso if-convert conditional jumps containing memory writes-ftree-loop-imEnable loop invariant motion on trees-ftree-loop-ivcanonCreate canonical induction variables in loops-ftree-loop-linearEnable loop interchange transforms. Same as -floop-interchange-ftree-loop-optimizeEnable loop optimizations on tree level-ftree-loop-vectorizeEnable loop vectorization on trees-ftree-lrsPerform live range splitting during the SSA->normal pass-ftree-parallelize-loops=Enable automatic parallelization of loops-ftree-partial-preIn SSA-PRE optimization on trees, enable partial-partial redundancy elimination-ftree-phipropEnable hoisting loads from conditional pointers.-ftree-preEnable SSA-PRE optimization on trees-ftree-ptaPerform function-local points-to analysis on trees.-ftree-reassocEnable reassociation on tree level-ftree-salias-ftree-scev-cpropEnable copy propagation of scalar-evolution information.-ftree-sinkEnable SSA code sinking on trees-ftree-slp-vectorizeEnable basic block vectorization (SLP) on trees-ftree-slsrPerform straight-line strength reduction-ftree-sraPerform scalar replacement of aggregates-ftree-store-ccp-ftree-store-copy-prop-ftree-switch-conversionPerform conversions of switch initializations.-ftree-tail-mergeEnable tail merging on trees-ftree-terReplace temporary expressions in the SSA->normal pass-ftree-vect-loop-version-ftree-vectorizeEnable vectorization on trees-ftree-vectorizer-verbose=-ftree-vrpPerform Value Range Propagation on trees-ftwo-level-all-subprogsWhen generating two-level line tables in DWARF (experimental), add linkage names for all functions (not just inlined functions).-ftwo-level-line-tablesUse two-level line tables in DWARF (experimental).-funderscoringAppend underscores to externally visible names-funit-at-a-timeCompile whole compilation unit at a time-funroll-all-loopsPerform loop unrolling for all loops-funroll-codesize-limitLimit non-const non-FP loop unrolling under profile estimates of large code footprint-funroll-loopsPerform loop unrolling when iteration count is known-funsafe-loop-optimizationsAllow loop optimizations to assume that the loops behave in normal way-funsafe-math-optimizationsAllow math optimizations that may violate IEEE or ISO standards-funsigned-bitfieldsWhen "signed" or "unsigned" is not given make the bitfield unsigned-funsigned-charMake "char" unsigned by default-funswitch-loopsPerform loop unswitching-funwind-tablesJust generate unwind tables for exception handling-fuse-atomic-builtinsGenerate code for built-in atomic operations-fuse-boehm-gcGenerate code for the Boehm GC-fuse-cxa-atexitUse __cxa_atexit to register destructors-fuse-cxa-get-exception-ptrUse __cxa_get_exception_ptr in exception handling-fuse-divide-subroutineCall a library routine to do integer divisions-fuse-ld=bfdUse the bfd linker instead of the default linker-fuse-ld=goldUse the gold linker instead of the default linker-fuse-ld=mcldUse the mcld linker instead of the default linker-fuse-linker-plugin-fvar-trackingPerform variable tracking-fvar-tracking-assignmentsPerform variable tracking by annotating assignments-fvar-tracking-assignments-toggleToggle -fvar-tracking-assignments-fvar-tracking-uninitPerform variable tracking and also tag variables that are uninitialized-fvariable-expansion-in-unrollerApply variable expansion when loops are unrolled-fvect-cost-modelEnables the dynamic vectorizer cost model. Preserved for backward compatibility.dynamic-fvect-cost-model=Specifies the cost model for vectorization-fverbose-asmAdd extra commentary to assembler output-fversion-fvisibility-inlines-hiddenMarks all inlined functions and methods as having hidden visibility-fvisibility-ms-compatChanges visibility to match Microsoft Visual Studio by default-fvisibility=-fvisibility=[default|internal|hidden|protected] Set the default symbol visibility-fvptUse expression value profiles in optimizations-fvtable-gc-fvtable-thunks-fvtable-verify=Validate vtable pointers before using them.-fvtv-countsOutput vtable verification counters.-fvtv-debugOutput vtable verification pointer sets information.-fweakEmit common-like symbols as weak symbols-fwebConstruct webs and split unrelated uses of single variable-fwhole-file-fwhole-programPerform whole program optimizations-fwide-exec-charset=-fwide-exec-charset=<cset> Convert all wide strings and character constants to character set <cset>-fworking-directoryGenerate a #line directive pointing at the current working directory-fwpaRun the link-time optimizer in whole program analysis (WPA) mode.-fwpa=Whole program analysis (WPA) mode with number of parallel jobs specified.-fwrapvAssume signed arithmetic overflow wraps around-fxref-fzee-fzero-initialized-in-bssPut zero initialized data in the bss section-fzero-linkGenerate lazy class lookup (via objc_getClass()) for use in Zero-Link modeGenerate debug information in default format-gantCatch typos-gcoffGenerate debug information in COFF format-gdwarfGenerate debug information in default version of DWARF format-gdwarf-Generate debug information in DWARF v2 (or later) format-gen-declsDump declarations to a .decl file-gfission-ggdbGenerate debug information in default extended format-ggnu-pubnamesGenerate DWARF pubnames and pubtypes sections with GNU extensions.-gline-tables-onlyGenerate DWARF line number tables and no other debug sections-gmltGenerate debug information at level 1 with minimal line table-gnat-gnat<options> Specify options to GNAT-gnatOSet name of output ALI file (internal switch)-gno-fission-gno-pubnamesDon't generate DWARF pubnames and pubtypes sections.-gno-record-gcc-switchesDon't record gcc command line switches in DWARF DW_AT_producer.-gno-split-dwarfDon't generate debug information in separate .dwo files-gno-strict-dwarfEmit DWARF additions beyond selected version-gpubnamesGenerate DWARF pubnames and pubtypes sections.-grecord-gcc-switchesRecord gcc command line switches in DWARF DW_AT_producer.-gsplit-dwarfGenerate debug information in separate .dwo files-gstabsGenerate debug information in STABS format-gstabs+Generate debug information in extended STABS format-gstrict-dwarfDon't emit DWARF additions beyond selected versionToggle debug information generation-gvmsGenerate debug information in VMS format-gxcoffGenerate debug information in XCOFF format-gxcoff+Generate debug information in extended XCOFF format-h-idirafter-idirafter <dir> Add <dir> to the end of the system include path-imacros-imacros <file> Accept definition of macros in <file>-imultiarch-imultiarch <dir> Set <dir> to be the multiarch include subdirectory-imultilib <dir> Set <dir> to be the multilib include subdirectory-include-include <file> Include the contents of <file> before other files-iplugindir=<dir> Set <dir> to be the default plugin directory-iprefix <path> Specify <path> as a prefix for next two options-iquote-iquote <dir> Add <dir> to the end of the quote include path-isysroot <dir> Set <dir> to be the system root directory-isystem <dir> Add <dir> to the start of the system include path-iwithprefix-iwithprefix <dir> Add <dir> to the end of the system include path-iwithprefixbefore-iwithprefixbefore <dir> Add <dir> to the end of the main include path-k8Synonym of -gnatk8-lang-asm-m128bit-long-doublesizeof(long double) is 16-m16Generate 16bit i386 code-m32Generate 32bit i386 code-m3dnowSupport 3DNow! built-in functions-m3dnowaSupport Athlon 3Dnow! built-in functions-m64Generate 64bit x86-64 code-m80387Use hardware fp-m8bit-idivExpand 32bit/64bit integer divide into 8bit unsigned integer divide with run-time check-m96bit-long-doublesizeof(long double) is 12-mabi=Generate code that conforms to the given ABI-mabmSupport code generation of Advanced Bit Manipulation (ABM) instructions.-maccumulate-outgoing-argsReserve space for outgoing arguments in the function prologue-maddress-mode=Use given address mode-madxSupport flag-preserving add-carry instructions-maesSupport AES built-in functions and code generation-malign-doubleAlign some doubles on dword boundary-malign-functions=Function starts are aligned to this power of 2-malign-jumps=Jump targets are aligned to this power of 2-malign-loops=Loop code aligned to this power of 2-malign-stringopsAlign destination of the string operations-mandroidGenerate code for the Android platform.-march=Generate code for given CPU-masm=Use given assembler dialect-mavxSupport MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2 and AVX built-in functions and code generation-mavx2Support MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, AVX and AVX2 built-in functions and code generation-mavx256-split-unaligned-loadSplit 32-byte AVX unaligned load-mavx256-split-unaligned-storeSplit 32-byte AVX unaligned store-mavx512cdSupport MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, AVX, AVX2 and AVX512F and AVX512CD built-in functions and code generation-mavx512erSupport MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, AVX, AVX2 and AVX512F and AVX512ER built-in functions and code generation-mavx512fSupport MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, AVX, AVX2 and AVX512F built-in functions and code generation-mavx512pfSupport MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, AVX, AVX2 and AVX512F and AVX512PF built-in functions and code generation-mbionicUse Bionic C library-mbmiSupport BMI built-in functions and code generation-mbmi2Support BMI2 built-in functions and code generation-mbranch-cost=Branches are this expensive (1-5, arbitrary units)-mcldGenerate cld instruction in the function prologue.-mcmodel=Use given x86-64 code model-mcpu=%<-mcpu=%> is deprecated; use %<-mtune=%> or %<-march=%> instead-mcrc32Support code generation of crc32 instruction.-mcx16Support code generation of cmpxchg16b instruction.-mdispatch-schedulerDo dispatch scheduling if processor is bdver1 or bdver2 or bdver3 or bdver4 and Haifa scheduling is selected.-mdump-tune-features-mf16cSupport F16C built-in functions and code generation-mfancy-math-387Generate sin, cos, sqrt for FPU-mfentryEmit profiling counter call at function entry before prologue.-mfmaSupport MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, AVX and FMA built-in functions and code generation-mfma4Support FMA4 built-in functions and code generation -mforce-drapAlways use Dynamic Realigned Argument Pointer (DRAP) to realign stack-mfp-ret-in-387Return values of functions in FPU registers-mfpmath=Generate floating point mathematics using given instruction set-mfsgsbaseSupport FSGSBASE built-in functions and code generation-mfunction-return=Convert function return to call and return thunk.-mfused-madd%<-mfused-madd%> is deprecated; use %<-ffp-contract=%> insteadfastoff-mfxsrSupport FXSAVE and FXRSTOR instructions-mglibcUse GNU C library-mhard-float-mhleSupport Hardware Lock Elision prefixes-mieee-fpUse IEEE math for fp comparisons-mincoming-stack-boundary=Assume incoming stack aligned to this power of 2-mindirect-branch-registerForce indirect call and jump via register.-mindirect-branch=Convert indirect call and jump to call and return thunks.-minline-all-stringopsInline all known string operations-minline-stringops-dynamicallyInline memset/memcpy string operations, but perform inline version only for small blocks-mintel-syntax%<-mintel-syntax%> and %<-mno-intel-syntax%> are deprecated; use %<-masm=intel%> and %<-masm=att%> insteadatt-mlarge-data-threshold=Data greater than given threshold will go into .ldata section in x86-64 medium model-mlong-double-128Use 128-bit long double-mlong-double-64Use 64-bit long double-mlong-double-80Use 80-bit long double-mlwpSupport LWP built-in functions and code generation -mlzcntSupport LZCNT built-in function and code generation-mmemcpy-strategy=Specify memcpy expansion strategy when expected size is known-mmemset-strategy=Specify memset expansion strategy when expected size is known-mmmxSupport MMX built-in functions-mmovbeSupport code generation of movbe instruction.-mms-bitfieldsUse native (MS) bitfield layout-mno-align-stringops-mno-defaultClear all tune features-mno-fancy-math-387-mno-push-args-mno-red-zone-mno-sse4Do not support SSE4.1 and SSE4.2 built-in functions and code generation-momit-leaf-frame-pointerOmit the frame pointer in leaf functions-mpc32Set 80387 floating-point precision to 32-bit-mpc64Set 80387 floating-point precision to 64-bit-mpc80Set 80387 floating-point precision to 80-bit-mpclmulSupport PCLMUL built-in functions and code generation-mpopcntSupport code generation of popcnt instruction.-mprefer-avx128Use 128-bit AVX instructions instead of 256-bit AVX instructions in the auto-vectorizer.-mpreferred-stack-boundary=Attempt to keep stack aligned to this power of 2-mprefetchwt1Support PREFETCHWT1 built-in functions and code generation-mprfchwSupport PREFETCHW instruction-mpush-argsUse push instructions to save outgoing arguments-mrdrndSupport RDRND built-in functions and code generation-mrdseedSupport RDSEED instruction-mrecipGenerate reciprocals instead of divss and sqrtss.-mrecip=Control generation of reciprocal estimates.-mred-zoneUse red-zone in the x86-64 code-mregparm=Number of registers used to pass integer arguments-mrtdAlternate calling convention-mrtmSupport RTM built-in functions and code generation-msahfSupport code generation of sahf instruction in 64bit x86-64 code.-mshaSupport SHA1 and SHA256 built-in functions and code generation-msoft-floatDo not use hardware fp-msseSupport MMX and SSE built-in functions and code generation-msse2Support MMX, SSE and SSE2 built-in functions and code generation-msse2avxEncode SSE instructions with VEX prefix-msse3Support MMX, SSE, SSE2 and SSE3 built-in functions and code generation-msse4Support MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1 and SSE4.2 built-in functions and code generation-msse4.1Support MMX, SSE, SSE2, SSE3, SSSE3 and SSE4.1 built-in functions and code generation-msse4.2-msse4aSupport MMX, SSE, SSE2, SSE3 and SSE4A built-in functions and code generation-msse5%<-msse5%> was removed-msseregparmUse SSE register passing conventions for SF and DF mode-mssse3Support MMX, SSE, SSE2, SSE3 and SSSE3 built-in functions and code generation-mstack-arg-probeEnable stack probing-mstack-protector-guard=Use given stack-protector guard-mstackrealignRealign stack in prologue-mstringop-strategy=Chose strategy to generate stringop using-mtbmSupport TBM built-in functions and code generation-mtls-dialect=Use given thread-local storage dialect-mtls-direct-seg-refsUse direct references against %gs when accessing tls data-mtune-ctrl=Fine grain control of tune features-mtune=Schedule code for given CPU-muclibcUse uClibc C library-mveclibabi=Vector library ABI to use-mvect8-ret-in-memReturn 8-byte vectors in memory-mvzeroupperGenerate vzeroupper instruction before a transfer of control flow out of the function.-mx32Generate 32bit x86-64 code-mxopSupport XOP built-in functions and code generation -mxsaveSupport XSAVE and XRSTOR instructions-mxsaveoptSupport XSAVEOPT instruction-n-no-canonical-prefixes-no-integrated-cpp-no-pieCreate a position dependent executable-nocppDisable preprocessing-nodefaultlibs-nostartfiles-nostdincDo not search standard system include directories (those specified with -isystem will still be used)-nostdinc++Do not search standard system include directories for C++-nostdlibDo not look for object files in standard path-o <file> Place output into <file>-pEnable function profiling-pass-exit-codes-pedantic-pedantic-errorsLike -pedantic but issue them as errors-pg-pieCreate a position independent executable-pipe-posix-print-file-name=-print-libgcc-file-name-print-multi-directory-print-multi-lib-print-multi-os-directory-print-multiarch-print-objc-runtime-infoGenerate C header of platform-specific features-print-prog-name=-print-search-dirs-print-sysroot-print-sysroot-headers-suffix-profile-pthread-quietDo not display functions compiled or elapsed time-r-rdynamic-remapRemap file names when including files-s-s-bc-abi-save-temps-save-temps=-sharedCreate a shared library-shared-libgcc-specs-specs=-static-static-libasan-static-libgcc-static-libgcj-static-libgfortranStatically link the GNU Fortran helper library (libgfortran)-static-libgo-static-liblsan-static-libstdc++-static-libtsan-static-libubsan-std=c++03Conform to the ISO 1998 C++ standard revised by the 2003 technical corrigendum-std=c++0xDeprecated in favor of -std=c++11-std=c++11Conform to the ISO 2011 C++ standard-std=c++14-std=c++1yConform to the ISO 2014(?) C++ draft standard (experimental and incomplete support)-std=c++98-std=c11Conform to the ISO 2011 C standard (experimental and incomplete support)-std=c1xDeprecated in favor of -std=c11-std=c89Conform to the ISO 1990 C standard-std=c90-std=c99Conform to the ISO 1999 C standard-std=c9xDeprecated in favor of -std=c99-std=f2003Conform to the ISO Fortran 2003 standard-std=f2008Conform to the ISO Fortran 2008 standard-std=f2008tsConform to the ISO Fortran 2008 standard including TS 29113-std=f95Conform to the ISO Fortran 95 standard-std=gnuConform to nothing in particular-std=gnu++03Conform to the ISO 1998 C++ standard revised by the 2003 technical corrigendum with GNU extensions-std=gnu++0xDeprecated in favor of -std=gnu++11-std=gnu++11Conform to the ISO 2011 C++ standard with GNU extensions (experimental and incomplete support)-std=gnu++14-std=gnu++1yConform to the ISO 201y(7?) C++ draft standard with GNU extensions (experimental and incomplete support)-std=gnu++98-std=gnu11Conform to the ISO 2011 C standard with GNU extensions (experimental and incomplete support)-std=gnu1xDeprecated in favor of -std=gnu11-std=gnu89Conform to the ISO 1990 C standard with GNU extensions-std=gnu90-std=gnu99Conform to the ISO 1999 C standard with GNU extensions-std=gnu9xDeprecated in favor of -std=gnu99-std=iso9899:1990-std=iso9899:199409Conform to the ISO 1990 C standard as amended in 1994-std=iso9899:1999-std=iso9899:199xDeprecated in favor of -std=iso9899:1999-std=iso9899:2011-std=legacyAccept extensions to support legacy code-symbolic-t-time-time=-tno-android-cc-tno-android-ld-traditional-traditional-cppEnable traditional preprocessing-trigraphs-trigraphs Support ISO C trigraphs-u-undefDo not predefine system-specific and GCC-specific macros-vEnable verbose output-versionDisplay the compiler's version-wSuppress warnings-wrapper-x-zcommon_handle_option_automssysv32kernellargemediumsmallautoevery-lineoncestandardon387387+sse387,ssebothssesse+387sse,387callgraphkeepthunkthunk-externthunk-inlineCBpriorityallmixedoneacmlsvmlshortglobaltlsnamed_varsnonebyte_looplibcalllooprep_4byterep_8byterep_byteunrolled_loopvector_loopdefaulthiddeninternalprotectedgnugnu2global-dynamicinitial-execlocal-dynamiclocal-execcheappreinitstd/Volumes/Android/buildbot/src/android/gcc/toolchain/build/../gcc/gcc-4.9/gcc/vec.ccalculate_allocation_1register_overhead/Volumes/Android/buildbot/src/android/gcc/toolchain/build/../gcc/gcc-4.9/gcc/hooks.chook_int_rtx_unreachablediagnostic.c0x%lx %s 3340 slots %lu
/prebuilts/tools/common/m2/repository/net/sf/saxon/Saxon-HE/9.8.0-5/
H A D	Saxon-HE-9.8.0-5.jar	META-INF/MANIFEST.MF META-INF/SAXONICA.SF META-INF/SAXONICA.RSA META ...
/prebuilts/tools/common/m2/repository/net/sourceforge/saxon/saxon/9.1.0.8/
H A D	saxon-9.1.0.8.jar	META-INF/ META-INF/MANIFEST.MF net/ net/sf/ net/sf/saxon/ net/sf/saxon/AugmentedSource ...
/prebuilts/gcc/linux-x86/aarch64/aarch64-linux-android-4.9/bin/
H A D	aarch64-linux-android-gcc-4.9.x	3804 optimizersinvalid --param value %qs-Werror=%s: no option -%snogenericspecificshiftinteger-divide-by-zerounreachablevla-boundreturnnullsigned-integer-overflowboolenumtargetwarningsundocumentedparamsjoinedseparatecommonstabscoffdwarf-2xcoffvmsmissing argument to %qs-Wno--W-fno--f-mno--m--debug=-g--machine---machine-no---machine=--machine=no---machine--optimize=-O--std=-std=--std--warn---warn-no-----no-options.cpreinitstdcheapdynamicunlimiteddesctradglobal-dynamicinitial-execlocal-dynamiclocal-execdefaulthiddeninternalprotectedallnamed_varsmixedoneCBprioritycallgraphfastoffonstandardevery-lineoncealwaysautoneverlargesmalltinyilp32lp64-###--CLASSPATH--all-warnings--ansi--assemble--assertassertion missing after %qs--assert=--bootclasspath--canonical-prefixes--classpath--comments--comments-in-macros--compile--coverage--debug--define-macromacro name missing after %qs--define-macro=--dependencies--dump--dump=--dumpbase--dumpdir--encoding--entry--entry=--extdirs--extra-warnings--for-assembler--for-assembler=--for-linker--for-linker=--force-link--force-link=Display this information--help=--imacrosmissing filename after %qs--imacros=--include--include-barrier--include-directorymissing path after %qs--include-directory-after--include-directory-after=--include-directory=--include-prefix--include-prefix=--include-with-prefix--include-with-prefix-after--include-with-prefix-after=--include-with-prefix-before--include-with-prefix-before=--include-with-prefix=--include=--language--language=--library-directory--library-directory=--no-canonical-prefixes--no-integrated-cpp--no-line-commands--no-pie--no-standard-includes--no-standard-libraries--no-sysroot-suffix--no-warnings--optimize--output--output-class-directory--output-class-directory=--output-pch=--output=--param--param=--pass-exit-codes--pedantic--pedantic-errors--pie--pipe--prefix--prefix=--preprocess--print-file-name--print-file-name=--print-libgcc-file-name--print-multi-directory--print-multi-lib--print-multi-os-directory--print-multiarch--print-prog-name--print-prog-name=--print-search-dirs--print-sysroot--profile--resource--resource=--save-temps--shared--specs--specs=--static--symbolic--sysroot--sysroot=Alias for --help=target--time--trace-includes--traditional--traditional-cpp--trigraphs--undefine-macro--undefine-macro=--user-dependencies--verbose--write-dependencies--write-user-dependencies-A-B-CDo not discard comments-CC-CLASSPATH-D-E-H-I-J-MGenerate make dependencies-MD-MD_-MF-MG-MM-MMD-MMD_-MP-MQ-MT-N-Ofast-Og-Os-P-Q-Qn-Qy-R-S-T-Tbss-Tbss=-Tdata-Tdata=-Ttext-Ttext=-U-U<macro> Undefine <macro>-Wa,-Wabi-Wabi-tag-Waddress-Waggregate-return-Waliasing-Walign-commons-WallEnable most warning messages-Wall-deprecation-Wall-javadoc-Wampersand-Warray-bounds-Warray-temporaries-Wassert-identifier-Wassign-intercept-Wattributes-Wbad-function-cast-Wboxing-Wbuiltin-macro-redefined-Wc++-compat-Wc++0x-compat-Wc++11-compat-Wc-binding-type-Wcast-align-Wcast-qual-Wchar-concat-Wchar-subscripts-Wcharacter-truncation-Wclobbered-Wcomment-WcommentsSynonym for -Wcomment-Wcompare-reals-Wcondition-assign-Wconditionally-supported-Wconstructor-name-Wconversion-Wconversion-extra-Wconversion-null-Wcoverage-mismatch-Wcpp-Wctor-dtor-privacy-Wdate-time-Wdeclaration-after-statement-Wdelete-incomplete-Wdelete-non-virtual-dtor-Wdep-ann-Wdeprecated-Wdeprecated-declarations-Wdisabled-optimization-Wdiscouraged-Wdiv-by-zero-Wdouble-promotion-Weffc++-Wempty-block-Wempty-body-Wendif-labels-Wenum-compare-Wenum-identifier-Wenum-switch-WerrorTreat all warnings as errorsimplicit-function-declaration-Werror=-Wextra-Wextraneous-semicolon-Wfallthrough-Wfatal-errors-Wfield-hiding-Wfinal-bound-Wfinally-Wfloat-conversion-Wfloat-equal-Wforbidden-Wforce-warnings-Wformat-Wformat-contains-nul-Wformat-extra-args-Wformat-nonliteral-Wformat-security-Wformat-y2k-Wformat-zero-length-Wformat=-Wframe-larger-than=-Wfree-nonheap-object-Wfunction-elimination-Whiding-Wignored-qualifiers-Wimplicit-Wimplicit-int-Wimplicit-interface-Wimplicit-procedure-Wimport-Windirect-static-Winherited-variadic-ctor-Winit-self-Winline-Wint-to-pointer-cast-Wintf-annotation-Wintf-non-inherited-Wintrinsic-shadow-Wintrinsics-std-Winvalid-memory-model-Winvalid-offsetof-Winvalid-pch-Wjavadoc-Wjump-misses-init-Wl,-Wlarger-than--Wlarger-than=-Wline-truncation-Wliteral-suffix-Wlocal-hiding-Wlogical-op-Wlong-long-Wmain-Wmasked-catch-block-Wmaybe-uninitialized-Wmissing-braces-Wmissing-declarations-Wmissing-field-initializers-Wmissing-format-attribute-Wmissing-include-dirs-Wmissing-noreturn-Wmissing-parameter-type-Wmissing-prototypes-Wmudflap-Wmultichar-Wnarrowing-Wnested-externs-Wnls-Wno-effect-assign-Wnoexcept-Wnon-template-friend-Wnon-virtual-dtor-Wnonnull-Wnormalized=-Wnull-Wold-style-cast-Wold-style-declaration-Wold-style-definition-Wopenmp-simd-Wout-of-date-Wover-ann-Woverflow-Woverlength-strings-Woverloaded-virtual-Woverride-init-Wp,-Wpacked-Wpacked-bitfield-compat-Wpadded-Wparam-assign-Wparentheses-Wpedantic-Wpkg-default-method-Wpmf-conversions-Wpointer-arith-Wpointer-sign-Wpointer-to-int-cast-Wpoison-system-directories-WpragmasWarn about misuses of pragmas-Wproperty-assign-default-Wprotocol-Wpsabi-Wraw-Wreal-q-constant-Wrealloc-lhs-Wrealloc-lhs-all-Wredundant-decls-Wredundant-modifiers-Wreorder-Wreturn-local-addr-Wreturn-type-Wripa-opt-mismatch-Wselector-Wself-assign-Wself-assign-non-pod-Wsequence-point-Wserial-Wshadow-Wshadow-compatible-local-Wshadow-local-Wsign-compare-Wsign-conversion-Wsign-promo-Wsizeof-pointer-memaccess-Wspecial-param-hiding-Wstack-protector-Wstack-usage=-Wstatic-access-Wstatic-receiver-Wstrict-aliasing-Wstrict-aliasing=-Wstrict-null-sentinel-Wstrict-overflow-Wstrict-overflow=-Wstrict-prototypes-Wstrict-selector-match-Wsuggest-attribute=const-Wsuggest-attribute=format-Wsuggest-attribute=noreturn-Wsuggest-attribute=pure-Wsuppress-Wsurprising-Wswitch-Wswitch-default-Wswitch-enum-Wsync-nand-Wsynth-Wsynthetic-access-Wsystem-headers-Wtabs-Wtarget-lifetime-Wtasks-Wthread-reentrant-lock-Wthread-safety-Wthread-safety-analysis-Wthread-safety-attributes-Wthread-safety-precise-Wthread-unguarded-func-Wthread-unguarded-var-Wtraditional-Wtraditional-conversion-Wtrampolines-Wtrigraphs-Wtype-hiding-Wtype-limits-Wuncheck-Wundeclared-selector-Wundef-Wunderflow-Wuninitialized-Wunknown-pragmas-Wunnecessary-else-Wunqualified-field-Wunreachable-code-Wunsafe-loop-optimizations-Wunsuffixed-float-constants-WunusedEnable all -Wunused- warnings-Wunused-argument-Wunused-but-set-parameter-Wunused-but-set-variable-Wunused-dummy-argument-Wunused-function-Wunused-import-Wunused-labelWarn when a label is unused-Wunused-local-Wunused-local-typedefs-Wunused-macros-Wunused-parameter-Wunused-private-Wunused-result-Wunused-thrown-Wunused-value-Wunused-variable-Wuseless-castWarn about useless casts-Wuseless-type-check-Wvarargs-Wvarargs-cast-Wvariadic-macros-Wvirtual-move-assign-Wvla-Wvolatile-register-var-Wwarning-token-Wwrite-strings-WzerotripWarn about zero-trip DO loops-Xassembler-Xclang-only=-Xlinker-Xpreprocessor-Z-ansi-aux-info-aux-info=-auxbase-auxbase-strip-bootclasspath-c-canonical-prefixes-classpath-coverage-cppEnable preprocessing-cpp=-d-dumpbase-dumpdir-dumpmachine-dumpspecs-dumpversion-e-encoding-export-dynamic-extdirs-fCLASSPATH=-fPIC-fPIE-fRTS=Select the runtime-fabi-version=-faccess-control-fada-spec-parent=-falign-commons-falign-functionsAlign the start of functions-falign-functions=-falign-jumps-falign-jumps=-falign-labelsAlign all labels-falign-labels=-falign-loopsAlign the start of loops-falign-loops=-fall-intrinsics-fall-virtual-fallow-leading-underscore-falt-external-templatesNo longer supported-fargument-alias-fargument-noalias-fargument-noalias-anything-fargument-noalias-global-fasmRecognize the "asm" keyword-fassert-fassociative-math-fassume-compiled-fassume-compiled=-fasynchronous-unwind-tables-fauto-inc-dec-fauto-profile-fauto-profile-accurate-fauto-profile=-fautomatic-faux-classpath-fbackslash-fbacktrace-fblas-matmul-limit=-fbootclasspath=-fbootstrap-classes-fbounds-check-fbranch-count-reg-fbranch-probabilities-fbranch-target-load-optimize-fbtr-bb-exclusive-fbuilding-libgcc-fbuiltinRecognize built-in functions-fbuiltin--fcall-saved--fcall-used--fcaller-saves-fcanonical-system-headers-fcheck-array-temporaries-fcheck-branch-annotation-fcheck-data-deps-fcheck-new-fcheck-references-fcheck=-fcilkplusEnable Cilk Plus-fclasspath=-fcoarray=-fcombine-stack-adjustments-fcommon-fcompare-debug-fcompare-debug-second-fcompare-elim-fcompile-resource=-fcond-mismatch-fconserve-space-fconserve-stack-fconstant-string-class=-fconstexpr-depth=-fconvert=big-endian-fconvert=little-endian-fconvert=native-fconvert=swap-fcprop-registers-fcray-pointer-fcrossjumping-fcse-follow-jumps-fcse-skip-blocks-fcx-fortran-rules-fcx-limited-range-fd-lines-as-code-fd-lines-as-comments-fdata-sections-fdbg-cnt-list-fdbg-cnt=-fdce-fdebug-cpp-fdebug-prefix-map=-fdebug-types-section-fdeclone-ctor-dtor-fdeduce-init-list-fdefault-double-8-fdefault-inline-fdefault-integer-8-fdefault-noexcept-fdefault-real-8-fdefer-pop-fdelayed-branch-fdelete-dead-exceptions-fdelete-null-pointer-checks-fdevirtualize-fdevirtualize-speculatively-fdiagnostics-color-fdiagnostics-color=-fdiagnostics-show-caret-fdiagnostics-show-location=-fdiagnostics-show-option-fdirectives-onlyPreprocess directives only.-fdisable--fdisable-assertions-fdisable-assertions=-fdollar-ok-fdollars-in-identifiers-fdse-fdump--fdump-ada-spec-fdump-ada-spec-slim-fdump-core-fdump-final-insns-fdump-fortran-optimized-fdump-fortran-original-fdump-go-spec=-fdump-noaddr-fdump-parse-tree-fdump-passesDump optimization passes-fdump-unnumbered-fdump-unnumbered-links-fdwarf2-cfi-asm-fearly-inliningPerform early inlining-felide-constructors-feliminate-dwarf2-dups-femit-class-debug-always-femit-class-fileOutput a class file-femit-class-filesAlias for -femit-class-file-femit-function-names-femit-struct-debug-baseonly-femit-struct-debug-detailed=-femit-struct-debug-reduced-fenable--fenable-assertions-fenable-assertions=-fencoding=-fenforce-eh-specs-fenum-int-equiv-fexceptionsEnable exception handling-fexcess-precision=-fexec-charset=-fexpensive-optimizations-fext-numeric-literals-fextdirs=-fextended-identifiers-fextern-tls-init-fexternal-blas-fexternal-templates-ff2cUse f2c calling convention-ffast-math-ffat-lto-objects-ffilelist-file-ffinite-math-only-ffixed--ffixed-form-ffixed-line-length--ffixed-line-length-none-ffloat-store-ffor-scope-fforce-addr-fforce-classes-archive-check-fforward-propagate-ffp-contract=-ffpe-summary=-ffpe-trap=-ffree-form-ffree-line-length--ffree-line-length-none-ffreestanding-ffriend-injection-ffrontend-optimizeEnable front end optimization-ffunction-attribute-list=-ffunction-cse-ffunction-sections-fgcse-fgcse-after-reload-fgcse-las-fgcse-lm-fgcse-sm-fgnu-keywords-fgnu-runtime-fgnu-tm-fgnu-unique-fgnu89-inline-fgo-check-divide-overflow-fgo-check-divide-zero-fgo-dump--fgo-optimize--fgo-pkgpath=-fgo-prefix=-fgo-relative-import-path=-fgraphite-fgraphite-identity-fguess-branch-probability-fguiding-decls-fhandle-exceptions-fhash-synchronization-fhelp-fhelp=-fhoist-adjacent-loads-fhonor-std-fhosted-fhuge-objects-fidentProcess #ident directives-fif-conversion-fif-conversion2-fimplement-inlines-fimplicit-inline-templates-fimplicit-none-fimplicit-templates-findirect-classes-findirect-dispatch-findirect-inliningPerform indirect inlining-finhibit-size-directive-finit-character=-finit-integer=-finit-local-zero-finit-logical=-finit-real=-finline-finline-atomics-finline-functions-finline-limit--finline-limit=-finline-small-functions-finput-charset=-finstrument-functions-finteger-4-integer-8-fintrinsic-modules-path-fintrinsic-modules-path=-fipa-cp-fipa-cp-clone-fipa-matrix-reorg-fipa-profile-fipa-pta-fipa-pure-const-fipa-reference-fipa-sra-fipa-struct-reorg-fira-algorithm=-fira-hoist-pressure-fira-loop-pressure-fira-region=-fira-share-save-slots-fira-share-spill-slots-fira-verbose=-fivopts-fjni-fjump-tables-fkeep-inline-dllexport-fkeep-inline-functions-fkeep-static-consts-flabels-ok-flax-vector-conversions-fleading-underscore-flifetime-dse-flive-range-shrinkage-floop-block-floop-flatten-floop-interchange-floop-nest-optimize-floop-optimize-floop-parallelize-allMark all loops as parallel-floop-strip-mine-flto-flto-compression-level=-flto-partition=1to1-flto-partition=balanced-flto-partition=max-flto-partition=none-flto-report-flto-report-wpa-flto=-fltrans-fltrans-output-list=-fmain=-fmath-errno-fmax-array-constructor=-fmax-errors=-fmax-identifier-length=-fmax-stack-var-size=-fmax-subrecord-length=-fmem-report-fmem-report-wpa-fmerge-all-constants-fmerge-constants-fmerge-debug-strings-fmessage-length=-fmodule-private-fmodulo-sched-fmodulo-sched-allow-regmoves-fmove-loop-invariants-fms-extensions-fmudflap-fmudflapir-fmudflapth-fname-mangling-version--fnew-abi-fnext-runtime-fnil-receivers-fno-vect-cost-model-fnon-call-exceptions-fnonansi-builtins-fnonnull-objects-fnothrow-opt-fobjc-abi-version=-fobjc-call-cxx-cdtors-fobjc-direct-dispatch-fobjc-exceptions-fobjc-gc-fobjc-nilcheck-fobjc-sjlj-exceptions-fobjc-std=objc1-fomit-frame-pointer-fopenmp-fopenmp-simd-foperator-names-fopt-info-fopt-info--foptimize-register-move-foptimize-sibling-calls-foptimize-strlen-foptional-diags-foutput-class-dir=-fpack-derived-fpack-struct-fpack-struct=-fpartial-inliningPerform partial inlining-fpcc-struct-return-fpch-deps-fpch-preprocess-fpeel-codesize-limit-fpeel-loopsPerform loop peeling-fpeephole-fpeephole2-fpermissive-fpic-fpie-fplan9-extensions-fplt-fplugin-arg--fplugin=Specify a plugin to load-fpost-ipa-mem-report-fpre-ipa-mem-report-fpredictive-commoning-fprefetch-loop-arrays-fpreprocessed-fpretty-templates-fprofile-fprofile-arcs-fprofile-correction-fprofile-dir=-fprofile-dump-fprofile-generate-fprofile-generate-atomic=-fprofile-generate-buildinfo=-fprofile-generate-sampling-fprofile-generate=-fprofile-reorder-functions-fprofile-report-fprofile-strip=-fprofile-use-fprofile-use=-fprofile-values-fprotect-parens-frandom-seed-frandom-seed=-frange-check-freal-4-real-10-freal-4-real-16-freal-4-real-8-freal-8-real-10-freal-8-real-16-freal-8-real-4-frealloc-lhs-freciprocal-math-frecord-gcc-switches-frecord-marker=4-frecord-marker=8-frecursive-freduced-reflection-free-freg-struct-return-fregmove-frename-registers-freorder-blocks-freorder-functions-freorder-functions=-frepack-arrays-freplace-objc-classes-frepo-frequire-return-statement-frerun-cse-after-loop-frerun-loop-opt-fresolution=The resolution file-fripa-fripa-allow-debug-fripa-disallow-asm-modules-fripa-disallow-opt-mismatch-fripa-inc-path-sub=-frounding-math-frtti-fsanitize-coverage=trace-pc-fsanitize=Select what to sanitize-fsaw-java-file-fsched-dep-count-heuristic-fsched-group-heuristic-fsched-interblock-fsched-last-insn-heuristic-fsched-pressure-fsched-rank-heuristic-fsched-spec-fsched-spec-insn-heuristic-fsched-spec-load-fsched-spec-load-dangerous-fsched-stalled-insns-fsched-stalled-insns-dep-fsched-stalled-insns-dep=-fsched-stalled-insns=-fsched-verbose=-fsched2-use-superblocks-fsched2-use-traces-fschedule-insns-fschedule-insns2-fsecond-underscore-fsection-anchors-fsee-fsel-sched-pipelining-fselective-scheduling-fselective-scheduling2-fshort-double-fshort-enums-fshort-wchar-fshow-column-fshrink-wrap-fshrink-wrap-frame-pointer-fsign-zero-fsignaling-nans-fsigned-bitfields-fsigned-charMake "char" signed by default-fsigned-zeros-fsimd-cost-model=-fsingle-precision-constant-fsized-delete-fsource-filename=-fsource=-fsplit-ivs-in-unroller-fsplit-stack-fsplit-wide-types-fsquangle-fstack-arrays-fstack-check-fstack-check=-fstack-limit-fstack-limit-register=-fstack-limit-symbol=-fstack-protector-fstack-protector-all-fstack-protector-strong-fstack-reuse=-fstack-usage-fstats-fstore-check-fstrength-reduce-fstrict-aliasing-fstrict-enum-precision-fstrict-enums-fstrict-overflow-fstrict-prototype-fstrict-volatile-bitfields-fsync-libcalls-fsyntax-only-ftabstop=-ftarget-help-ftarget=Set the target VM version-ftemplate-backtrace-limit=-ftemplate-depth--ftemplate-depth=-ftest-coverage-fthis-is-variable-fthread-jumps-fthreadsafe-statics-ftime-report-ftls-model=-ftoplevel-reorder-ftracer-ftrack-macro-expansion-ftrack-macro-expansion=-ftrapping-math-ftrapv-ftree-bit-ccp-ftree-builtin-call-dce-ftree-ccp-ftree-ch-ftree-coalesce-inlined-vars-ftree-coalesce-vars-ftree-copy-prop-ftree-copyrename-ftree-cselim-ftree-dce-ftree-dominator-opts-ftree-dseEnable dead store elimination-ftree-forwprop-ftree-fre-ftree-loop-distribution-ftree-loop-if-convert-ftree-loop-if-convert-stores-ftree-loop-im-ftree-loop-ivcanon-ftree-loop-linear-ftree-loop-optimize-ftree-loop-vectorize-ftree-lrs-ftree-parallelize-loops=-ftree-partial-pre-ftree-phiprop-ftree-pre-ftree-pta-ftree-reassoc-ftree-salias-ftree-scev-cprop-ftree-sink-ftree-slp-vectorize-ftree-slsr-ftree-sra-ftree-store-ccp-ftree-store-copy-prop-ftree-switch-conversion-ftree-tail-mergeEnable tail merging on trees-ftree-ter-ftree-vect-loop-version-ftree-vectorizeEnable vectorization on trees-ftree-vectorizer-verbose=-ftree-vrp-ftwo-level-all-subprogs-ftwo-level-line-tables-funderscoring-funit-at-a-time-funroll-all-loops-funroll-codesize-limit-funroll-loops-funsafe-loop-optimizations-funsafe-math-optimizations-funsigned-bitfields-funsigned-char-funswitch-loopsPerform loop unswitching-funwind-tables-fuse-atomic-builtins-fuse-boehm-gc-fuse-cxa-atexit-fuse-cxa-get-exception-ptr-fuse-divide-subroutine-fuse-ld=bfd-fuse-ld=gold-fuse-ld=mcld-fuse-linker-plugin-fvar-trackingPerform variable tracking-fvar-tracking-assignments-fvar-tracking-uninit-fvect-cost-model-fvect-cost-model=-fverbose-asm-fversion-fvisibility-inlines-hidden-fvisibility-ms-compat-fvisibility=-fvpt-fvtable-gc-fvtable-thunks-fvtable-verify=-fvtv-counts-fvtv-debug-fweak-fweb-fwhole-file-fwhole-program-fwide-exec-charset=-fworking-directory-fwpa-fwpa=-fwrapv-fxref-fzee-fzero-initialized-in-bss-fzero-link-gantCatch typos-gcoff-gdwarf-gdwarf--gen-decls-gfission-ggdb-ggnu-pubnames-gline-tables-only-gmlt-gnat-gnatO-gno-fission-gno-pubnames-gno-record-gcc-switches-gno-split-dwarf-gno-strict-dwarf-gpubnames-grecord-gcc-switches-gsplit-dwarf-gstabs-gstabs+-gstrict-dwarf-gvms-gxcoff-gxcoff+-h-idirafter-imacros-include-iquote-iwithprefix-iwithprefixbefore-k8Synonym of -gnatk8-lang-asm-mabi=-mandroid-march=-mbig-endian-mbionicUse Bionic C library-mcmodel=Specify the code model-mcpu=-mfix-cortex-a53-835769-mfix-cortex-a53-843419-mgeneral-regs-only-mglibcUse GNU C library-mlittle-endian-mlra-momit-leaf-frame-pointer-mstrict-align-mtls-dialect=Specify TLS dialect-mtune=-mtune=CPU Optimize for CPU-muclibcUse uClibc C library-n-no-canonical-prefixes-no-integrated-cpp-no-pie-nocppDisable preprocessing-nodefaultlibs-nostartfiles-nostdinc-nostdinc++-nostdlib-pEnable function profiling-pass-exit-codes-pedantic-pedantic-errors-pg-pie-pipe-posix-print-file-name=-print-libgcc-file-name-print-multi-directory-print-multi-lib-print-multi-os-directory-print-multiarch-print-objc-runtime-info-print-prog-name=-print-search-dirs-print-sysroot-print-sysroot-headers-suffix-profile-pthread-quiet-r-rdynamic-remap-s-s-bc-abi-save-temps-save-temps=-sharedCreate a shared library-shared-libgcc-specs-specs=-static-static-libasan-static-libgcc-static-libgcj-static-libgfortran-static-libgo-static-liblsan-static-libstdc++-static-libtsan-static-libubsan-std=c++03-std=c++0x-std=c++11-std=c++14-std=c++1y-std=c++98-std=c11-std=c1x-std=c89-std=c90-std=c99-std=c9x-std=f2003-std=f2008-std=f2008ts-std=f95-std=gnu-std=gnu++03-std=gnu++0x-std=gnu++11-std=gnu++14-std=gnu++1y-std=gnu++98-std=gnu11-std=gnu1x-std=gnu89-std=gnu90-std=gnu99-std=gnu9x-std=iso9899:1990-std=iso9899:199409-std=iso9899:1999-std=iso9899:199x-std=iso9899:2011-std=legacy-symbolic-t-time-time=-tno-android-cc-tno-android-ld-traditional-traditional-cpp-trigraphs-u-undef-vEnable verbose output-version-wSuppress warnings-wrapper-x-zAdaAdaSCILAdaWhyCC++FortranGoJavaLTOObjCObjC++unknown IRA algorithm %qsunknown IRA region %qsunknown Stack Reuse Level %qsunknown TLS model %qsThe possible TLS dialects:: %s%s%s 3829 slots %lu 3919 All options with the desired characteristics have already been displayedThe following options are not documenteddebug format "%s" conflicts with prior selectionunrecognised debug output level "%s"debug output level %s is too highargument %qs to %<-femit-struct-debug-detailed%> unknown%<-femit-struct-debug-detailed=dir:...%> must allow at least as much as %<-femit-struct-debug-detailed=ind:...%>argument %qs to %<-femit-struct-debug-detailed%> not recognizedargument to %<-O%> should be a non-negative integer, %<g%>, %<s%> or %<fast%>section anchors must be disabled when unit-at-a-time is disabledtoplevel reorder must be disabled when unit-at-a-time is disabledtransactional memory is not supported with non-call exceptionssection anchors must be disabled when toplevel reorder is disabled-freorder-blocks-and-partition does not work with exceptions on this architecture-freorder-blocks-and-partition does not support unwind info on this architecture-freorder-blocks-and-partition does not work on this architecture-fno-fat-lto-objects are supported only with linker pluginonly one -flto-partition value can be specified%<-fsplit-stack%> is not supported by this compiler configurationDebug generation via -g option disabled under -fripa -fprofile-generate (use -fripa-allow-debug to override)-fsanitize=address is incompatible with -fsanitize=kernel-address-fsanitize=address and -fsanitize=kernel-address are incompatible with -fsanitize=thread%s: --param arguments should be of the form NAME=VALUE--help argument %q.*s is ambiguous, please be more specificunrecognized argument to --help= option: %q.*sgetting core file size maximum limit: %msetting core file size limit to maximum: %munrecognized gcc debugging option: %cstructure alignment must be a small power of two, not %dunknown stack check parameter "%s"%<-gdwarf%s%> is ambiguous; use %<-gdwarf-%s%> for DWARF version or %<-gdwarf -g%s%> for debug leveldwarf version %d is not supportedunrecognized argument to -fsanitize= option: %q.*s/buildbot/src/android/gcc/toolchain/build/../gcc/gcc-4.9/gcc/opts-common.ccommand line option %qs is not supported by this configurationargument to %qs should be a non-negative integerunrecognized argument in option %qsvalid arguments to %qs are: %s--help=<class> Display descriptions of a specific class of options. <class> is one or more of optimizers, target, warnings, undocumented, params--param <param>=<value> Set parameter <param> to value. See below for a complete list of parameters--print-missing-file-dependencies--print-sysroot-headers-suffix-A<question>=<answer> Assert the <answer> to <question>. Putting '-' before <question> disables the <answer> to <question>Do not discard comments in macro expansions-D<macro>[=<val>] Define a <macro> with <val> as its value. If just <macro> is given, <val> is taken to be 1-F <dir> Add <dir> to the end of the main framework include pathPrint the name of header files as they are used-I <dir> Add <dir> to the end of the main include path-J<directory> Put MODULE files in 'directory'Generate make dependencies and compile-MF <file> Write dependency output to the given fileTreat missing header files as generated filesLike -M but ignore system header filesLike -MD but ignore system header filesGenerate phony targets for all headers-MQ <target> Add a MAKE-quoted targetmissing makefile target after %qs-MT <target> Add an unquoted target-O<number> Set optimization level to <number>Optimize for speed disregarding exact standards complianceOptimize for debugging experience rather than speed or sizeOptimize for space rather than speedDo not generate #line directivesThis switch is deprecated; use -Wextra insteadWarn about things that will change when compiling with an ABI-compliant compilerWarn if a subobject has an abi_tag attribute that the complete object type does not haveWarn about suspicious uses of memory addressesWarn about returning structures, unions or arrays-Waggressive-loop-optimizationsWarn if a loop with constant number of iterations triggers undefined behaviorWarn about possible aliasing of dummy argumentsWarn about alignment of COMMON blocksWarn about missing ampersand in continued character constantsWarn if an array is accessed out of boundsWarn about creation of array temporariesWarn whenever an Objective-C assignment is being intercepted by the garbage collectorWarn about inappropriate attribute usageWarn about casting functions to incompatible typesWarn when a built-in preprocessor macro is undefined or redefinedWarn about C constructs that are not in the common subset of C and C++Deprecated in favor of -Wc++11-compatWarn about C++ constructs whose meaning differs between ISO C++ 1998 and ISO C++ 2011Warn if the type of a variable might be not interoperable with CWarn about pointer casts which increase alignmentWarn about casts which discard qualifiersWarn about subscripts whose type is "char"Warn about truncated character expressionsWarn about variables that might be changed by "longjmp" or "vfork"Warn about possibly nested block comments, and C++ comments spanning more than one physical lineWarn about equality comparisons involving REAL or COMPLEX expressionsWarn for conditionally-supported constructsWarn for implicit type conversions that may change a valueWarn about most implicit conversionsWarn for converting NULL from/to a non-pointer typeWarn in case profiles in -fprofile-use do not matchWarn when a #warning directive is encounteredWarn when all constructors and destructors are privateWarn about __TIME__, __DATE__ and __TIMESTAMP__ usageWarn when a declaration is found after a statementWarn when deleting a pointer to incomplete typeWarn about deleting polymorphic objects with non-virtual destructorsWarn if a deprecated compiler feature, class, method, or field is usedWarn about uses of __attribute__((deprecated)) declarationsWarn when an optimization pass is disabledWarn about compile-time integer division by zeroWarn about implicit conversions from "float" to "double"Warn about violations of Effective C++ style rulesWarn about an empty body in an if or else statementWarn about stray tokens after #elif and #endifWarn about comparison of different enum types-Werror-implicit-function-declarationThis switch is deprecated; use -Werror=implicit-function-declaration insteadTreat specified warning as errorPrint extra (possibly unwanted) warningsWarn if deprecated empty statements are foundExit on the first error occurredWarn for implicit type conversions that cause loss of floating point precisionWarn if testing floating point numbers for equalityDisable promoting warnings to errorsWarn about printf/scanf/strftime/strfmon format string anomaliesWarn about format strings that contain NUL bytesWarn if passing too many arguments to a function for its format stringWarn about format strings that are not literalsWarn about possible security problems with format functionsWarn about strftime formats yielding 2-digit yearsWarn about zero-length formats-Wframe-larger-than=<number> Warn if a function's stack frame requires more than <number> bytesWarn when attempting to free a non-heap objectWarn about function call eliminationWarn whenever type qualifiers are ignored.Warn about implicit declarations-Wimplicit-function-declarationWarn about implicit function declarationsWarn when a declaration does not specify a typeWarn about calls with implicit interfaceWarn about called procedures not explicitly declaredWarn about C++11 inheriting constructors when the base has a variadic constructorWarn about variables which are initialized to themselvesWarn when an inlined function cannot be inlinedWarn when there is a cast to a pointer from an integer of a different sizeWarn if a user-procedure has the same name as an intrinsicWarn on intrinsics not part of the selected standardWarn when an atomic memory model parameter is known to be outside the valid range.Warn about invalid uses of the "offsetof" macroWarn about PCH files that are found but not usedWarn when a jump misses a variable initialization-Wlarger-than=<number> Warn if an object is larger than <number> bytesWarn about truncated source linesWarn when a string or character literal is followed by a ud-suffix which does not begin with an underscore.Warn when a logical operator is suspiciously always evaluating to true or falseDo not warn about using "long long" when -pedanticWarn about suspicious declarations of "main"Warn about maybe uninitialized automatic variablesWarn about possibly missing braces around initializersWarn about global functions without previous declarationsWarn about missing fields in struct initializersWarn about user-specified include directories that do not existWarn about function parameters declared without a type specifier in K&R-style functionsWarn about global functions without prototypesswitch %qs is no longer supportedWarn about use of multi-character character constantsWarn about narrowing conversions within { } that are ill-formed in C++11Warn about "extern" declarations not at file scopeWarn when a noexcept expression evaluates to false even though the expression can't actually throwWarn when non-templatized friend functions are declared within a templateWarn about non-virtual destructorsWarn about NULL being passed to argument slots marked as requiring non-NULL-Wnormalized=<id|nfc|nfkc> Warn about non-normalised Unicode stringsWarn if a C-style cast is used in a programWarn for obsolescent usage in a declarationWarn if an old-style parameter definition is usedWarn if a simd directive is overridden by the vectorizer cost modelWarn if .class files are out of dateWarn about overflow in arithmetic expressionsWarn if a string is longer than the maximum portable length specified by the standardWarn about overloaded virtual function namesWarn about overriding initializers without side effectsWarn when the packed attribute has no effect on struct layoutWarn about packed bit-fields whose offset changed in GCC 4.4Warn when padding is required to align structure membersWarn about possibly missing parenthesesIssue warnings needed for strict compliance to the standardWarn when converting the type of pointers to member functionsWarn about function pointer arithmeticWarn when a pointer differs in signedness in an assignmentWarn when a pointer is cast to an integer of a different sizeWarn for -I and -L options using system directories if cross compilingWarn if a property for an Objective-C object has no assign semantics specifiedWarn if inherited methods are unimplementedWarn about real-literal-constants with 'q' exponent-letterWarn when a left-hand-side array variable is reallocatedWarn when a left-hand-side variable is reallocatedWarn about multiple declarations of the same objectWarn if modifiers are specified when not necessaryWarn when the compiler reorders codeWarn about returning a pointer/reference to a local or temporary variable.Warn whenever a function's return type defaults to "int" (C), or about inconsistent return types (C++)Warn if primary and auxiliary modules have mismatched command line optionsWarn if a selector has multiple methodsWarn when a variable is assigned to itselfWarn when a variable of a non-POD type is assigned to itselfWarn about possible violations of sequence point rulesWarn when one local variable shadows anotherWarn when one local variable shadows another local variable or parameter of compatible typeWarn when one local variable shadows another local variable or parameterWarn about signed-unsigned comparisonsWarn for implicit type conversions between signed and unsigned integersWarn when overload promotes from unsigned to signedWarn when not issuing stack smashing protection for some reasonWarn if stack usage might be larger than specified amountWarn about code which might break strict aliasing rulesWarn about uncasted NULL used as sentinelWarn about optimizations that assume that signed overflow is undefinedWarn about unprototyped function declarationsWarn if type signatures of candidate methods do not match exactlyWarn about functions which might be candidates for __attribute__((const))Warn about functions which might be candidates for format attributesWarn about functions which might be candidates for __attribute__((noreturn))Warn about functions which might be candidates for __attribute__((pure))Warn about "suspicious" constructsWarn about enumerated switches, with no default, missing a caseWarn about enumerated switches missing a "default:" statementWarn about all enumerated switches missing a specific caseWarn when __sync_fetch_and_nand and __sync_nand_and_fetch built-in functions are usedDeprecated. This switch has no effectDo not suppress warnings from system headersPermit nonconforming uses of the tab characterWarn if the pointer in a pointer assignment might outlive its target-Wthread-mismatched-lock-acq-relWarn about mismatched lock acquisition and release-Wthread-mismatched-lock-orderWarn about lock acquisition order inconsistent with what specified in the attributesWarn about a lock being acquired recursivelyWarn about potential thread safety issues when the code is annotated with thread safety attributesDoes nothing. For compatibility with clang thread safety analysis.Warn about function calls not properly protected by locks specified in the attributesWarn about shared variables not properly protected by locks specified in the attributes-Wthread-unsupported-lock-nameWarn about uses of unsupported lock names in attributesWarn about features not present in traditional CWarn of prototypes causing type conversions different from what would happen in the absence of prototypeWarn whenever a trampoline is generatedWarn if trigraphs are encountered that might affect the meaning of the programWarn if a comparison is always true or always false due to the limited range of the data typeWarn about @selector()s without previously declared methodsWarn if an undefined macro is used in an #if directiveWarn about underflow of numerical constant expressionsWarn about uninitialized automatic variablesWarn about unrecognized pragmasDoes nothing. Preserved for backward compatibility.Warn if the loop cannot be optimized due to nontrivial assumptions.Warn about unsuffixed float constantsWarn when a function parameter is only set, otherwise unusedWarn when a variable is only set, otherwise unusedWarn about unused dummy arguments.Warn when a function is unusedWarn when typedefs locally defined in a function are not usedWarn about macros defined in the main file that are not usedWarn when a function parameter is unusedWarn if a caller of a function, marked with attribute warn_unused_result, does not use its return valueWarn when an expression value is unusedWarn when a variable is unusedWarn about questionable usage of the macros used to retrieve variable argumentsWarn about using variadic macros-Wvector-operation-performanceWarn when a vector operation is compiled outside the SIMDWarn if a virtual base has a non-trivial move assignment operatorWarn if a variable length array is usedWarn when a register variable is declared volatileIn C++, nonzero means warn about deprecated conversion from string literals to 'char *'. In C, similar warning, except that the conversion is of course not deprecated by the ISO C standard.-Wzero-as-null-pointer-constantWarn when a literal '0' is used as null pointerA synonym for -std=c89 (for C) or -std=c++98 (for C++)-aux-info <file> Emit declaration information into <file>-d<letters> Enable dumps from specific passes of the compiler-dumpbase <file> Set the file basename to be used for dumps-dumpdir <dir> Set the directory name to be used for dumps--CLASSPATH Deprecated; use --classpath insteadGenerate position-independent code if possible (large mode)Generate position-independent code for executables if possible (large mode)Enforce class member access control semantics-fada-spec-parent=unit Dump Ada specs as child units of given parent-faggressive-function-eliminationEliminate multiple function invokations also for impure functions-faggressive-loop-optimizationsAggressively optimize loops using language constraintsEnable alignment of COMMON blocksAlign labels which are only reached by jumpingAll intrinsics procedures are available regardless of selected standard-fallow-parameterless-variadic-functionsAllow variadic functions without named parameterPermit the use of the assert keywordAllow optimization for floating-point arithmetic which may change the result of the operation due to rounding.Generate unwind tables that are exact at each instruction boundaryGenerate auto-inc/dec instructionsUse sample profile information for call graph node weights. The default profile file is fbdata.afdo in 'pwd'.Whether to assume the sample profile is accurate.-fauto-profile-record-coverage-in-elfWhether to record annotation coverage info in elf.Use sample profile information for call graph node weights. The profile file is specified in the argument.Do not treat local variables and COMMON blocks as if they were named in SAVE statementsSpecify that backslash in string introduces an escape characterProduce a backtrace when a runtime error is encountered-fblas-matmul-limit=<n> Size of the smallest matrix for which matmul will use BLAS--bootclasspath=<path> Replace system pathGenerated should be loaded by bootstrap loaderGenerate code to check bounds before indexing arraysReplace add, compare, branch with branch on count registerUse profiling information for branch probabilitiesPerform branch target load optimization before prologue / epilogue threading-fbranch-target-load-optimize2Perform branch target load optimization after prologue / epilogue threadingRestrict target load migration not to re-use registers in any basic block-fcall-saved-<register> Mark <register> as being preserved across functions-fcall-used-<register> Mark <register> as being corrupted by function callsSave registers around function callsWhere shorter, use canonicalized paths to systems headers.Produce a warning at runtime if a array temporary has been created for a procedure argumentCompare branch prediction result and autofdo profile information, store the result in a section in the generated elf file.-fcheck-branch-annotation-threshold=The number of executions a basic block needs to reach before GCC dumps its branch prediction information with -fcheck-branch-annotation.Compare the results of several data dependence analyzers.Check the return value of new in C++Generate checks for references to NULL-fcheck=[...] Specify which runtime checks are to be performed--classpath=<path> Set class path-fcoarray=[...] Specify which coarray parallelization should be usedLooks for opportunities to reduce stack adjustments and stack references.Do not put uninitialized globals in the common sectionRun only the second compilation of -fcompare-debug-fcompare-debug[=<opts>] Compile with and without e.g. -gtoggle, and compare the final-insns dumpPerform comparison elimination after register allocation has finishedAllow the arguments of the '?' operator to have different typesDoes nothing. Preserved for backward compatibility.Do not perform optimizations increasing noticeably stack usage-fconst-string-class=<name> Use class <name> for constant stringsno class name specified with %qs-fconstexpr-depth=<number> Specify maximum constexpr recursion depthUse big-endian format for unformatted filesUse little-endian format for unformatted filesUse native format for unformatted filesSwap endianness for unformatted filesPerform a register copy-propagation optimization passUse the Cray Pointer extensionPerform cross-jumping optimizationWhen running CSE, follow jumps to their targetsComplex multiplication and division follow Fortran rulesOmit range reduction step when performing complex divisionIgnore 'D' in column one in fixed formTreat lines with 'D' in column one as commentsPlace data items into their own sectionList all available debugging counters with their limits and counts.-fdbg-cnt=<counter>:<limit>[,<counter>:<limit>,...] Set the debug counter limit. Use the RTL dead code elimination passEmit debug annotations during preprocessingMap one directory name to another in debug informationOutput .debug_types section when using DWARF v4 debuginfo.Factor complex constructors and destructors to favor space over speed-fdeduce-init-list enable deduction of std::initializer_list for a template type parameter from a brace-enclosed initializer-listSet the default double precision kind to an 8 byte wide typeSet the default integer kind to an 8 byte wide typeMake functions no-throw/noexcept by defaultSet the default real kind to an 8 byte wide typeDefer popping functions args from stack until laterAttempt to fill delay slots of branch instructionsDelete dead instructions that may throw exceptionsDelete useless null pointer checksTry to convert virtual calls to direct ones.Perform speculative devirtualization-fdiagnostics-color=[never|always|auto] Colorize diagnosticsShow the source line with a caret indicating the column-fdiagnostics-show-location=[once|every-line] How often to emit source location at the beginning of line-wrapped diagnosticsAmend appropriate diagnostic messages with the command line option that controls them-fdisable-[tree|rtl|ipa]-<pass>=range1+range2 disables an optimization passAllow dollar signs in entity namesPermit '$' as an identifier characterUse the RTL dead store elimination pass-fdump-<type> Dump various compiler internals to a fileWrite all declarations as Ada code transitivelyWrite all declarations as Ada code for the given file only-fdump-final-insns=filename Dump to filename the insns at the end of translationDisplay the code tree after front end optimizationDisplay the code tree after parsing-fdump-go-spec=filename Write all declarations to file as Go codeSuppress output of addresses in debugging dumpsDisplay the code tree after parsing; deprecated optionSuppress output of instruction numbers, line number notes and addresses in debugging dumpsSuppress output of previous and next insn numbers in debugging dumpsEnable CFI tables via GAS assembler directives.Perform DWARF2 duplicate elimination-feliminate-unused-debug-symbolsPerform unused type elimination in debug info-feliminate-unused-debug-typesDo not suppress C++ class debug information.Print to stderr the mapping from module name and function id to assembler function name when -ftest-coverage, -fprofile-generate or -fprofile-use are active, for use in correlating function ids in gcda files with the function name.-femit-struct-debug-baseonly Aggressive reduced debug info for structs-femit-struct-debug-detailed=<spec-list> Detailed reduced debug info for structs-femit-struct-debug-reduced Conservative reduced debug info for structs-fenable-[tree|rtl|ipa]-<pass>=range1+range2 enables an optimization pass--encoding=<encoding> Choose input encoding (defaults from your locale)Generate code to check exception specifications-fexcess-precision=[fast|standard] Specify handling of excess floating-point precision-fexec-charset=<cset> Convert all strings and character constants to character set <cset>Perform a number of minor, expensive optimizationsInterpret imaginary, fixed-point, or other gnu number suffix as the corresponding number literal rather than a user-defined number literal.--extdirs=<path> Set the extension directory pathPermit universal character names (\u and \U) in identifiersSupport dynamic initialization of thread-local variables in a different translation unitSpecify that an external BLAS library should be used for matmul calls on large-size arraysOutput lto objects containing both the intermediate language and binary output.Input file is a file with a list of filenames to compileAssume no NaNs or infinities are generated-ffixed-<register> Mark <register> as being unavailable to the compilerAssume that the source file is fixed form-ffixed-line-length-<n> Use n as character line width in fixed modeAllow arbitrary character line width in fixed modeDon't allocate floats and doubles in extended-precision registersScope of for-init-statement variables is local to the loopAlways check for non gcj generated classes archivesPerform a forward propagation pass on RTL-ffp-contract=[off|on|fast] Perform floating-point expression contraction.-ffpe-summary=[...] Print summary of floating point exceptions-ffpe-trap=[...] Stop on following floating point exceptionsAssume that the source file is free form-ffree-line-length-<n> Use n as character line width in free modeAllow arbitrary character line width in free modeDo not assume that standard C libraries and "main" existInject friend functions into enclosing namespace-ffunction-attribute-list=attribute:name,... Add attribute to named functionsAllow function addresses to be held in registersPlace each function into its own sectionPerform global common subexpression eliminationPerform global common subexpression elimination after register allocation has finishedPerform redundant load after store elimination in global common subexpression eliminationPerform enhanced load motion during global common subexpression eliminationPerform store motion after global common subexpression eliminationRecognize GNU-defined keywordsGenerate code for GNU runtime environmentEnable support for GNU transactional memoryUse STB_GNU_UNIQUE if supported by the assemblerUse traditional GNU semantics for inline functionsAdd explicit checks for division overflow in INT_MIN / -1Add explicit checks for division by zero-fgo-dump-<type> Dump Go frontend internal information-fgo-optimize-<type> Turn on optimization passes in the frontend-fgo-pkgpath=<string> Set Go package path-fgo-prefix=<string> Set package-specific prefix for exported Go names-fgo-relative-import-path=<path> Treat a relative import as relative to pathEnable in and out of Graphite representationEnable Graphite Identity transformationEnable guessing of branch probabilities-fhandle-exceptions has been renamed -fexceptions (and is now on by default)Assume the runtime uses a hash table to map an object to its synchronization structureEnable hoisting adjacent loads to encourage generating conditional move instructionsAssume normal C execution environmentPerform conversion of conditional jumps to branchless equivalentsPerform conversion of conditional jumps to conditional executionExport functions even if they can be inlinedEmit implicit instantiations of inline templatesSpecify that no implicit typing is allowed, unless overridden by explicit IMPLICIT statementsEmit implicit instantiations of templatesGenerate instances of Class at runtimeUse offset tables for virtual method callsDo not generate .size directives-finit-character=<n> Initialize local character variables to ASCII value n-finit-integer=<n> Initialize local integer variables to nInitialize local variables to zero (from g77)-finit-logical=<true|false> Initialize local logical variables-finit-real=<zero|nan|inf|-inf> Initialize local real variablesEnable inlining of function declared "inline", disabling disables all inliningInline __atomic operations when a lock free instruction sequence is available.Integrate functions not declared "inline" into their callers when profitable-finline-functions-called-onceIntegrate functions only required by their single caller-finline-limit=<number> Limit the size of inlined functions to <number>Integrate functions into their callers when code size is known not to grow-finput-charset=<cset> Specify the default character set for source filesInstrument function entry and exit with profiling calls-finstrument-functions-exclude-file-list=-finstrument-functions-exclude-file-list=filename,... Do not instrument functions listed in files-finstrument-functions-exclude-function-list=-finstrument-functions-exclude-function-list=name,... Do not instrument listed functionsInterpret any INTEGER(4) as an INTEGER(8)Specify where to find the compiled intrinsic modulesPerform interprocedural constant propagationPerform cloning to make Interprocedural constant propagation strongerPerform interprocedural profile propagationPerform interprocedural points-to analysisDiscover pure and const functionsDiscover readonly and non addressable static variablesPerform interprocedural reduction of aggregates-fira-algorithm=[CB|priority] Set the used IRA algorithmUse IRA based register pressure calculation in RTL hoist optimizations.Use IRA based register pressure calculation in RTL loop optimizations.-fira-region=[one|all|mixed] Set regions for IRAShare slots for saving different hard registers.Share stack slots for spilled pseudo-registers.-fira-verbose=<number> Control IRA's level of diagnostic messages.-fisolate-erroneous-paths-attributeDetect paths which trigger erroneous or undefined behaviour due a NULL value being used in a way which is forbidden by a returns_nonnull or nonnull attribute. Isolate those paths from the main control flow and turn the statement with erroneous or undefined behaviour into a trap. -fisolate-erroneous-paths-dereferenceDetect paths which trigger erroneous or undefined behaviour due to dereferencing a NULL pointer. Isolate those paths from the main control flow and turn the statement with erroneous or undefined behaviour into a trap.Optimize induction variables on treesAssume native functions are implemented using JNIUse jump tables for sufficiently large switch statementsDon't emit dllexported inline functions unless neededGenerate code for functions even if they are fully inlinedEmit static const variables even if they are not usedAllow implicit conversions between vectors with differing numbers of subparts and/or differing element types.Give external symbols a leading underscoreTell DSE that the storage for a C++ object is dead when the constructor starts and when the destructor finishes.Relief of register pressure through live range shrinkageEnable Loop Blocking transformationEnable Loop Interchange transformationEnable the ISL based loop nest optimizerEnable Loop Strip Mining transformationEnable link-time optimization.-flto-compression-level=<number> Use zlib compression level <number> for ILPartition symbols and vars at linktime based on object files they originate fromPartition functions and vars at linktime into approximately same sized bucketsPut every symbol into separate partitionDisable partioning and streamingReport various link-time optimization statisticsReport various link-time optimization statistics for WPA onlyLink-time optimization with number of parallel jobs or jobserver.Run the link-time optimizer in local transformation (LTRANS) mode.Specify a file to which a list of files output by LTRANS is written.Set errno after built-in math functions-fmax-array-constructor=<n> Maximum number of objects in an array constructor-fmax-errors=<number> Maximum number of errors to report-fmax-identifier-length=<n> Maximum identifier length-fmax-stack-var-size=<n> Size in bytes of the largest array that will be put on the stack-fmax-subrecord-length=<n> Maximum length for subrecordsReport on permanent memory allocationReport on permanent memory allocation in WPA onlyAttempt to merge identical constants and constant variablesAttempt to merge identical constants across compilation unitsAttempt to merge identical debug strings across compilation units-fmessage-length=<number> Limit diagnostics to <number> characters per line. 0 suppresses line-wrappingSet default accessibility of module entities to PRIVATE.Perform SMS based modulo scheduling before the first scheduling passPerform SMS based modulo scheduling with register moves allowedMove loop invariant computations out of loopsDon't warn about uses of Microsoft extensionsGenerate code for NeXT (Apple Mac OS X) runtime environmentAssume that receivers of Objective-C messages may be nilEnables the unlimited vectorizer cost model. Preserved for backward compatibility.Support synchronous non-call exceptionsTreat a throw() exception specification as noexcept to improve code sizeSpecify which ABI to use for Objective-C family code and meta-data generation.Generate special Objective-C methods to initialize/destroy non-POD C++ ivars, if neededAllow fast jumps to the message dispatcherEnable Objective-C exception and synchronization syntaxEnable garbage collection (GC) in Objective-C/Objective-C++ programsEnable inline checks for nil receivers with the NeXT runtime and ABI version 2.Enable Objective-C setjmp exception handling runtimeConform to the Objective-C 1.0 language as implemented in GCC 4.0When possible do not generate stack framesEnable OpenMP (implies -frecursive in Fortran)Enable OpenMP's SIMD directivesRecognize C++ keywords like "compl" and "xor"Enable all optimization info dumps on stderr-fopt-info[-<type>=filename] Dump compiler optimization detailsOptimize sibling and tail recursive calls-foptimize-static-class-initializationEnable optimization of static class initialization codeEnable string length optimizations on treesTry to lay out derived types as compactly as possiblePack structure members together without holes-fpack-struct=<number> Set initial maximum structure member alignmentReturn small aggregates in memory, not registersLook for and use PCH files even when preprocessingLimit non-const non-FP loop peeling under profile estimates of large code footprintEnable machine specific peephole optimizationsEnable an RTL peephole pass before sched2Downgrade conformance errors to warningsGenerate position-independent code if possible (small mode)Generate position-independent code for executables if possible (small mode)Enable Plan 9 language extensionsUse PLT for PIC calls (-fno-plt: load the address from GOT at call site)-fplugin-arg-<name>-<key>[=<value>] Specify argument <key>=<value> for plugin <name>Report on memory allocation before interprocedural optimizationRun predictive commoning optimization.Generate prefetch instructions, if available, for arrays in loopsTreat the input file as already preprocessed-fno-pretty-templates Do not pretty-print template specializations as the template signature followed by the argumentsEnable basic program profiling codeInsert arc-based program profiling codeEnable correction of flow inconsistent profile data inputSet the top-level directory for storing the profile data. The default is 'pwd'.Dump CFG profile for comparison.Enable common options for generating profile info for profile feedback directed optimizationsfprofile-generate-atomic=[0..3] Atomically increments for profile counters.-fprofile-generate-buildinfo=filename Read build info to include in gcda file from filenameTurn on instrumentation sampling with -fprofile-generate with rate set by --param profile-generate-sampling-rate or environment variable GCOV_SAMPLING_RATEEnable common options for generating profile info for profile feedback directed optimizations, and set -fprofile-dir=Enable function reordering that improves code placementReport on consistency of profileSpecify a substring to be stripped from the profile base file nameEnable common options for performing profile feedback directed optimizationsEnable common options for performing profile feedback directed optimizations, and set -fprofile-dir=Insert code to profile values of expressionsProtect parentheses in expressions-frandom-seed=<string> Make compile reproducible using <string>Enable range checking during compilationInterpret any REAL(4) as a REAL(10)Interpret any REAL(4) as a REAL(16)Interpret any REAL(4) as a REAL(8)Interpret any REAL(8) as a REAL(10)Interpret any REAL(8) as a REAL(16)Interpret any REAL(8) as a REAL(4)Reallocate the LHS in assignmentsSame as -fassociative-math for expressions which include division.-frecord-compilation-info-in-elfRecord the compiler optimizations in a .gnu.switches.text section.Record gcc command line switches in the object file.Use a 4-byte record marker for unformatted filesUse an 8-byte record marker for unformatted filesAllocate local variables on the stack to allow indirect recursionReduce the amount of reflection meta-data generatedTurn on Redundant Extensions Elimination pass.Return small aggregates in registersPerform a register renaming optimization passReorder basic blocks to improve code placement-freorder-blocks-and-partitionReorder basic blocks and partition into hot and cold sectionsReorder functions to improve code placement-freorder-functions=[callgraph] Select the scheme for function reordering. This invokes a linker plugin. Generate .gnu.callgraph.text sections listing callees and edge counts.Copy array sections into a contiguous block on procedure entryUsed in Fix-and-Continue mode to indicate that object files may be swapped in at runtimeEnable automatic template instantiationFunctions which return values must end with return statementsAdd a common subexpression elimination pass after loop optimizations-freschedule-modulo-scheduled-loopsEnable/Disable the traditional scheduling in loops that already passed modulo schedulingPerform Dynamic Inter-Procedural Analysis.Allow -g enablement for -fripa -fprofile-generate compiles.Don't import an auxiliary module if it contains asm statementsDon't import an auxiliary module if the command line options mismatch with the primary moduleSubstitute substring in include paths with a new string to allow reuse profile data-fripa-no-promote-always-inline-funcDon't promote always inline static functions assuming they will be inlined and no copy is needed.Disable optimizations that assume default FP rounding behaviorGenerate run time type descriptor informationEnable coverage-guided fuzzing code instrumentation. Inserts call to __sanitizer_cov_trace_pc into every basic block.-fsched-critical-path-heuristicEnable the critical path heuristic in the schedulerEnable the dependent count heuristic in the schedulerEnable the group heuristic in the schedulerEnable scheduling across basic blocksEnable the last instruction heuristic in the schedulerEnable register pressure sensitive insn schedulingEnable the rank heuristic in the schedulerAllow speculative motion of non-loadsEnable the speculative instruction heuristic in the schedulerAllow speculative motion of some loadsAllow speculative motion of more loadsAllow premature scheduling of queued insnsSet dependence distance checking in premature scheduling of queued insns-fsched-stalled-insns-dep=<number> Set dependence distance checking in premature scheduling of queued insns-fsched-stalled-insns=<number> Set number of queued insns that can be prematurely scheduled-fsched-verbose=<number> Set the verbosity level of the schedulerIf scheduling post reload, do superblock schedulingReschedule instructions before register allocationReschedule instructions after register allocationAppend a second underscore if the name already contains an underscoreAccess data in the same section from shared anchor pointsPerform software pipelining of inner loops during selective scheduling-fsel-sched-pipelining-outer-loopsPerform software pipelining of outer loops during selective scheduling-fsel-sched-reschedule-pipelinedReschedule pipelined regions without pipeliningSchedule instructions using selective scheduling algorithmRun selective scheduling after reloadUse the same size for double as for floatUse the narrowest integer type possible for enumeration typesForce the underlying type for "wchar_t" to be "unsigned short"Show column numbers in diagnostics, when available. Default onEmit function prologues only before parts of the function that need it, rather than at the top of the function.Framepointer shrinkwrapping optimization.Apply negative sign to zero valuesDisable optimizations observable by IEEE signaling NaNsWhen "signed" or "unsigned" is not given make the bitfield signedDisable floating point optimizations that ignore the IEEE signedness of zeroSpecifies the vectorization cost model for code marked with a simd directiveConvert floating point constants to single precision constantsSupport delete operator with objetc's size as the second parameter.Set the source language versionSplit lifetimes of induction variables when loops are unrolledGenerate discontiguous stack framesSplit wide types into independent registersPut all local arrays on stack.Insert stack checking code into the program. Same as -fstack-check=specific-fstack-check=[no|generic|specific] Insert stack checking code into the program-fstack-limit-register=<register> Trap if the stack goes past <register>-fstack-limit-symbol=<name> Trap if the stack goes past symbol <name>Use propolice as a stack protection methodUse a stack protection method for every functionUse a smart stack protection method for certain functions-fstack-reuse=[all|named_vars|none] Set stack reuse level for local variables.Output stack usage information on a per-function basisDisplay statistics accumulated during compilationEnable assignability checks for stores into object arraysAssume strict aliasing rules applyPerform transformations based on enum precisionAssume that values of enumeration type are always within the minimum range of that typeTreat signed overflow as undefinedForce bitfield accesses to match their type widthImplement __atomic operations via libcalls to legacy __sync functionsCheck for syntax errors, then stop-ftabstop=<number> Distance between tab stops for column reportingSet the maximum number of template instantiation notes for a single warning or error-ftemplate-depth=<number> Specify maximum template instantiation depthCreate data files needed by "gcov"Perform jump threading optimizations-fno-threadsafe-statics Do not generate thread-safe code for initializing local staticsReport the time taken by each compiler pass-ftls-model=[global-dynamic|local-dynamic|initial-exec|local-exec] Set the default thread-local storage code generation modelReorder top level functions, variables, and asmsPerform superblock formation via tail duplication-ftrack-macro-expansion=<0|1|2> Track locations of tokens coming from macro expansion and display them in error messagesAssume floating-point operations can trapTrap for signed overflow in addition, subtraction and multiplicationEnable SSA-BIT-CCP optimization on treesEnable conditional dead code elimination for builtin callsEnable SSA-CCP optimization on treesEnable loop header copying on treesEnable coalescing of copy-related user variables that are inlinedEnable coalescing of all copy-related user variablesEnable copy propagation on treesReplace SSA temporaries with better names in copiesTransform condition stores into unconditional onesEnable SSA dead code elimination optimization on treesEnable dominator optimizationsEnable forward propagation on treesEnable Full Redundancy Elimination (FRE) on trees-ftree-loop-distribute-patternsEnable loop distribution for patterns transformed into a library callEnable loop distribution on treesConvert conditional jumps in innermost loops to branchless equivalentsAlso if-convert conditional jumps containing memory writesEnable loop invariant motion on treesCreate canonical induction variables in loopsEnable loop interchange transforms. Same as -floop-interchangeEnable loop optimizations on tree levelEnable loop vectorization on treesPerform live range splitting during the SSA->normal passEnable automatic parallelization of loopsIn SSA-PRE optimization on trees, enable partial-partial redundancy eliminationEnable hoisting loads from conditional pointers.Enable SSA-PRE optimization on treesPerform function-local points-to analysis on trees.Enable reassociation on tree levelEnable copy propagation of scalar-evolution information.Enable SSA code sinking on treesEnable basic block vectorization (SLP) on treesPerform straight-line strength reductionPerform scalar replacement of aggregatesPerform conversions of switch initializations.Replace temporary expressions in the SSA->normal passPerform Value Range Propagation on treesWhen generating two-level line tables in DWARF (experimental), add linkage names for all functions (not just inlined functions).Use two-level line tables in DWARF (experimental).Append underscores to externally visible namesCompile whole compilation unit at a timePerform loop unrolling for all loopsLimit non-const non-FP loop unrolling under profile estimates of large code footprintPerform loop unrolling when iteration count is knownAllow loop optimizations to assume that the loops behave in normal wayAllow math optimizations that may violate IEEE or ISO standardsWhen "signed" or "unsigned" is not given make the bitfield unsignedMake "char" unsigned by defaultJust generate unwind tables for exception handlingGenerate code for built-in atomic operationsGenerate code for the Boehm GCUse __cxa_atexit to register destructorsUse __cxa_get_exception_ptr in exception handlingCall a library routine to do integer divisionsUse the bfd linker instead of the default linkerUse the gold linker instead of the default linkerUse the mcld linker instead of the default linkerPerform variable tracking by annotating assignments-fvar-tracking-assignments-toggleToggle -fvar-tracking-assignmentsPerform variable tracking and also tag variables that are uninitialized-fvariable-expansion-in-unrollerApply variable expansion when loops are unrolledEnables the dynamic vectorizer cost model. Preserved for backward compatibility.Specifies the cost model for vectorizationAdd extra commentary to assembler outputMarks all inlined functions and methods as having hidden visibilityChanges visibility to match Microsoft Visual Studio by default-fvisibility=[default|internal|hidden|protected] Set the default symbol visibilityUse expression value profiles in optimizationsValidate vtable pointers before using them.Output vtable verification counters.Output vtable verification pointer sets information.Emit common-like symbols as weak symbolsConstruct webs and split unrelated uses of single variablePerform whole program optimizations-fwide-exec-charset=<cset> Convert all wide strings and character constants to character set <cset>Generate a #line directive pointing at the current working directoryRun the link-time optimizer in whole program analysis (WPA) mode.Whole program analysis (WPA) mode with number of parallel jobs specified.Assume signed arithmetic overflow wraps aroundPut zero initialized data in the bss sectionGenerate lazy class lookup (via objc_getClass()) for use in Zero-Link modeGenerate debug information in default formatGenerate debug information in COFF formatGenerate debug information in default version of DWARF formatGenerate debug information in DWARF v2 (or later) formatDump declarations to a .decl fileGenerate debug information in default extended formatGenerate DWARF pubnames and pubtypes sections with GNU extensions.Generate DWARF line number tables and no other debug sectionsGenerate debug information at level 1 with minimal line table-gnat<options> Specify options to GNATSet name of output ALI file (internal switch)Don't generate DWARF pubnames and pubtypes sections.Don't record gcc command line switches in DWARF DW_AT_producer.Don't generate debug information in separate .dwo filesEmit DWARF additions beyond selected versionGenerate DWARF pubnames and pubtypes sections.Record gcc command line switches in DWARF DW_AT_producer.Generate debug information in separate .dwo filesGenerate debug information in STABS formatGenerate debug information in extended STABS formatDon't emit DWARF additions beyond selected versionToggle debug information generationGenerate debug information in VMS formatGenerate debug information in XCOFF formatGenerate debug information in extended XCOFF format-idirafter <dir> Add <dir> to the end of the system include path-imacros <file> Accept definition of macros in <file>-imultiarch <dir> Set <dir> to be the multiarch include subdirectory-imultilib <dir> Set <dir> to be the multilib include subdirectory-include <file> Include the contents of <file> before other files-iplugindir=<dir> Set <dir> to be the default plugin directory-iprefix <path> Specify <path> as a prefix for next two options-iquote <dir> Add <dir> to the end of the quote include path-isysroot <dir> Set <dir> to be the system root directory-isystem <dir> Add <dir> to the start of the system include path-iwithprefix <dir> Add <dir> to the end of the system include path-iwithprefixbefore <dir> Add <dir> to the end of the main include path-mabi=ABI Generate code that conforms to the specified ABIGenerate code for the Android platform.-march=ARCH Use features of architecture ARCHAssume target CPU is configured as big endian-mcpu=CPU Use features of and optimize for CPUWorkaround for ARM Cortex-A53 Erratum number 835769Workaround for ARM Cortex-A53 Erratum number 843419Generate code which uses only the general registersAssume target CPU is configured as little endianUse LRA instead of reload (transitional)Omit the frame pointer in leaf functionsDon't assume that unaligned accesses are handled by the systemCreate a position dependent executableDo not search standard system include directories (those specified with -isystem will still be used)Do not search standard system include directories for C++Do not look for object files in standard path-o <file> Place output into <file>Like -pedantic but issue them as errorsCreate a position independent executableGenerate C header of platform-specific featuresDo not display functions compiled or elapsed timeRemap file names when including filesStatically link the GNU Fortran helper library (libgfortran)Conform to the ISO 1998 C++ standard revised by the 2003 technical corrigendumDeprecated in favor of -std=c++11Conform to the ISO 2011 C++ standardConform to the ISO 2014(?) C++ draft standard (experimental and incomplete support)Conform to the ISO 2011 C standard (experimental and incomplete support)Deprecated in favor of -std=c11Conform to the ISO 1990 C standardConform to the ISO 1999 C standardDeprecated in favor of -std=c99Conform to the ISO Fortran 2003 standardConform to the ISO Fortran 2008 standardConform to the ISO Fortran 2008 standard including TS 29113Conform to the ISO Fortran 95 standardConform to nothing in particularConform to the ISO 1998 C++ standard revised by the 2003 technical corrigendum with GNU extensionsDeprecated in favor of -std=gnu++11Conform to the ISO 2011 C++ standard with GNU extensions (experimental and incomplete support)Conform to the ISO 201y(7?) C++ draft standard with GNU extensions (experimental and incomplete support)Conform to the ISO 2011 C standard with GNU extensions (experimental and incomplete support)Deprecated in favor of -std=gnu11Conform to the ISO 1990 C standard with GNU extensionsConform to the ISO 1999 C standard with GNU extensionsDeprecated in favor of -std=gnu99Conform to the ISO 1990 C standard as amended in 1994Deprecated in favor of -std=iso9899:1999Accept extensions to support legacy codeEnable traditional preprocessing-trigraphs Support ISO C trigraphsDo not predefine system-specific and GCC-specific macrosDisplay the compiler's versionKnown AArch64 ABIs (for use with the -mabi= option):The code model option names for -mcmodel:unknown excess precision style %qsunknown floating point contraction style %qsunrecognized function reorder value %qsunrecognized visibility value %qsunknown vectorizer cost model %qsunknown vtable verify initialization priority %qs/buildbot/src/android/gcc/toolchain/build/../gcc/gcc-4.9/gcc/vec.c/buildbot/src/android/gcc/toolchain/build/../gcc/gcc-4.9/gcc/hooks.c%s: all warnings being treated as errors%s: some warnings being treated as errorsIn file included from %r%s:%d:%d%RIn file included from %r%s:%d%R,
/prebuilts/gcc/linux-x86/arm/arm-linux-androideabi-4.9/bin/
H A D	arm-linux-androideabi-gcc-4.9.x	3909 optimizersinvalid --param value %qs-Werror=%s: no option -%snogenericspecificshiftinteger-divide-by-zerounreachablevla-boundreturnnullsigned-integer-overflowboolenumtargetwarningsundocumentedparamsjoinedseparatecommonstabscoffdwarf-2xcoffvmsmissing argument to %qs-Wno--W-fno--f-mno--m--debug=-g--machine---machine-no---machine=--machine=no---machine--optimize=-O--std=-std=--std--warn---warn-no-----no-options.cpreinitstdcheapdynamicunlimitedgnu2global-dynamicinitial-execlocal-dynamiclocal-execdefaulthiddeninternalprotectedallnamed_varsarm1020earm1020tarm1022earm1026ej-sarm10earm10tdmiarm1136j-sarm1136jf-sarm1156t2-sarm1156t2f-sarm1176jz-sarm1176jzf-sarm2arm250arm3arm6arm60arm600arm610arm620arm7arm70arm700arm700iarm710arm7100arm710carm710tarm720arm720tarm740tarm7500arm7500fearm7darm7diarm7dmarm7dmiarm7marm7tdmiarm7tdmi-sarm8arm810arm9arm920arm920tarm922tarm926ej-sarm940tarm946e-sarm966e-sarm968e-sarm9earm9tdmicortex-a12cortex-a15cortex-a15.cortex-a7cortex-a5cortex-a53cortex-a57cortex-a57.cortex-a53cortex-a7cortex-a8cortex-a9cortex-m0cortex-m0pluscortex-m1cortex-m3cortex-m4cortex-r4cortex-r4fcortex-r5cortex-r7ep9312fa526fa606tefa626fa626tefa726tefmp626generic-armv7-aiwmmxtiwmmxt2marvell-pj4mpcorempcorenovfpnativestrongarmstrongarm110strongarm1100strongarm1110xscalemixedoneCBprioritycallgraphfastoffonhardsoftfpstandardevery-lineoncealwaysautonevercp15crypto-neon-fp-armv8fp-armv8fpv4-sp-d16neonneon-fp-armv8neon-fp16neon-vfpv4vfp3vfpv3vfpv3-d16vfpv3-d16-fp16vfpv3-fp16vfpv3xdvfpv3xd-fp16vfpv4vfpv4-d16alternativeieeearmv2armv2aarmv3armv3marmv4armv4tarmv5armv5earmv5tarmv6armv6-marmv6jarmv6karmv6s-marmv6t2armv6zarmv6zkarmv7armv7-aarmv7-marmv7-rarmv7e-marmv7vearmv8-aarmv8-a+crcaapcsaapcs-linuxapcs-gnuatpcs-###--CLASSPATH--all-warnings--ansi--assemble--assertassertion missing after %qs--assert=--bootclasspath--canonical-prefixes--classpath--comments--comments-in-macros--compile--coverage--debug--define-macromacro name missing after %qs--define-macro=--dependencies--dump--dump=--dumpbase--dumpdir--encoding--entry--entry=--extdirs--extra-warnings--for-assembler--for-assembler=--for-linker--for-linker=--force-link--force-link=Display this information--help=--imacrosmissing filename after %qs--imacros=--include--include-barrier--include-directorymissing path after %qs--include-directory-after--include-directory-after=--include-directory=--include-prefix--include-prefix=--include-with-prefix--include-with-prefix-after--include-with-prefix-after=--include-with-prefix-before--include-with-prefix-before=--include-with-prefix=--include=--language--language=--library-directory--library-directory=--no-canonical-prefixes--no-integrated-cpp--no-line-commands--no-pie--no-standard-includes--no-standard-libraries--no-sysroot-suffix--no-warnings--optimize--output--output-class-directory--output-class-directory=--output-pch=--output=--param--param=--pass-exit-codes--pedantic--pedantic-errors--pie--pipe--prefix--prefix=--preprocess--print-file-name--print-file-name=--print-libgcc-file-name--print-multi-directory--print-multi-lib--print-multi-os-directory--print-multiarch--print-prog-name--print-prog-name=--print-search-dirs--print-sysroot--profile--resource--resource=--save-temps--shared--specs--specs=--static--symbolic--sysroot--sysroot=Alias for --help=target--time--trace-includes--traditional--traditional-cpp--trigraphs--undefine-macro--undefine-macro=--user-dependencies--verbose--write-dependencies--write-user-dependencies-A-B-CDo not discard comments-CC-CLASSPATH-D-E-H-I-J-MGenerate make dependencies-MD-MD_-MF-MG-MM-MMD-MMD_-MP-MQ-MT-N-Ofast-Og-Os-P-Q-Qn-Qy-R-S-T-Tbss-Tbss=-Tdata-Tdata=-Ttext-Ttext=-U-U<macro> Undefine <macro>-Wa,-Wabi-Wabi-tag-Waddress-Waggregate-return-Waliasing-Walign-commons-WallEnable most warning messages-Wall-deprecation-Wall-javadoc-Wampersand-Warray-bounds-Warray-temporaries-Wassert-identifier-Wassign-intercept-Wattributes-Wbad-function-cast-Wboxing-Wbuiltin-macro-redefined-Wc++-compat-Wc++0x-compat-Wc++11-compat-Wc-binding-type-Wcast-align-Wcast-qual-Wchar-concat-Wchar-subscripts-Wcharacter-truncation-Wclobbered-Wcomment-WcommentsSynonym for -Wcomment-Wcompare-reals-Wcondition-assign-Wconditionally-supported-Wconstructor-name-Wconversion-Wconversion-extra-Wconversion-null-Wcoverage-mismatch-Wcpp-Wctor-dtor-privacy-Wdate-time-Wdeclaration-after-statement-Wdelete-incomplete-Wdelete-non-virtual-dtor-Wdep-ann-Wdeprecated-Wdeprecated-declarations-Wdisabled-optimization-Wdiscouraged-Wdiv-by-zero-Wdouble-promotion-Weffc++-Wempty-block-Wempty-body-Wendif-labels-Wenum-compare-Wenum-identifier-Wenum-switch-WerrorTreat all warnings as errorsimplicit-function-declaration-Werror=-Wextra-Wextraneous-semicolon-Wfallthrough-Wfatal-errors-Wfield-hiding-Wfinal-bound-Wfinally-Wfloat-conversion-Wfloat-equal-Wforbidden-Wforce-warnings-Wformat-Wformat-contains-nul-Wformat-extra-args-Wformat-nonliteral-Wformat-security-Wformat-y2k-Wformat-zero-length-Wformat=-Wframe-larger-than=-Wfree-nonheap-object-Wfunction-elimination-Whiding-Wignored-qualifiers-Wimplicit-Wimplicit-int-Wimplicit-interface-Wimplicit-procedure-Wimport-Windirect-static-Winherited-variadic-ctor-Winit-self-Winline-Wint-to-pointer-cast-Wintf-annotation-Wintf-non-inherited-Wintrinsic-shadow-Wintrinsics-std-Winvalid-memory-model-Winvalid-offsetof-Winvalid-pch-Wjavadoc-Wjump-misses-init-Wl,-Wlarger-than--Wlarger-than=-Wline-truncation-Wliteral-suffix-Wlocal-hiding-Wlogical-op-Wlong-long-Wmain-Wmasked-catch-block-Wmaybe-uninitialized-Wmissing-braces-Wmissing-declarations-Wmissing-field-initializers-Wmissing-format-attribute-Wmissing-include-dirs-Wmissing-noreturn-Wmissing-parameter-type-Wmissing-prototypes-Wmudflap-Wmultichar-Wnarrowing-Wnested-externs-Wnls-Wno-effect-assign-Wnoexcept-Wnon-template-friend-Wnon-virtual-dtor-Wnonnull-Wnormalized=-Wnull-Wold-style-cast-Wold-style-declaration-Wold-style-definition-Wopenmp-simd-Wout-of-date-Wover-ann-Woverflow-Woverlength-strings-Woverloaded-virtual-Woverride-init-Wp,-Wpacked-Wpacked-bitfield-compat-Wpadded-Wparam-assign-Wparentheses-Wpedantic-Wpkg-default-method-Wpmf-conversions-Wpointer-arith-Wpointer-sign-Wpointer-to-int-cast-Wpoison-system-directories-WpragmasWarn about misuses of pragmas-Wproperty-assign-default-Wprotocol-Wpsabi-Wraw-Wreal-q-constant-Wrealloc-lhs-Wrealloc-lhs-all-Wredundant-decls-Wredundant-modifiers-Wreorder-Wreturn-local-addr-Wreturn-type-Wripa-opt-mismatch-Wselector-Wself-assign-Wself-assign-non-pod-Wsequence-point-Wserial-Wshadow-Wshadow-compatible-local-Wshadow-local-Wsign-compare-Wsign-conversion-Wsign-promo-Wsizeof-pointer-memaccess-Wspecial-param-hiding-Wstack-protector-Wstack-usage=-Wstatic-access-Wstatic-receiver-Wstrict-aliasing-Wstrict-aliasing=-Wstrict-null-sentinel-Wstrict-overflow-Wstrict-overflow=-Wstrict-prototypes-Wstrict-selector-match-Wsuggest-attribute=const-Wsuggest-attribute=format-Wsuggest-attribute=noreturn-Wsuggest-attribute=pure-Wsuppress-Wsurprising-Wswitch-Wswitch-default-Wswitch-enum-Wsync-nand-Wsynth-Wsynthetic-access-Wsystem-headers-Wtabs-Wtarget-lifetime-Wtasks-Wthread-reentrant-lock-Wthread-safety-Wthread-safety-analysis-Wthread-safety-attributes-Wthread-safety-precise-Wthread-unguarded-func-Wthread-unguarded-var-Wtraditional-Wtraditional-conversion-Wtrampolines-Wtrigraphs-Wtype-hiding-Wtype-limits-Wuncheck-Wundeclared-selector-Wundef-Wunderflow-Wuninitialized-Wunknown-pragmas-Wunnecessary-else-Wunqualified-field-Wunreachable-code-Wunsafe-loop-optimizations-Wunsuffixed-float-constants-WunusedEnable all -Wunused- warnings-Wunused-argument-Wunused-but-set-parameter-Wunused-but-set-variable-Wunused-dummy-argument-Wunused-function-Wunused-import-Wunused-labelWarn when a label is unused-Wunused-local-Wunused-local-typedefs-Wunused-macros-Wunused-parameter-Wunused-private-Wunused-result-Wunused-thrown-Wunused-value-Wunused-variable-Wuseless-castWarn about useless casts-Wuseless-type-check-Wvarargs-Wvarargs-cast-Wvariadic-macros-Wvirtual-move-assign-Wvla-Wvolatile-register-var-Wwarning-token-Wwrite-strings-WzerotripWarn about zero-trip DO loops-Xassembler-Xclang-only=-Xlinker-Xpreprocessor-Z-ansi-aux-info-aux-info=-auxbase-auxbase-strip-bootclasspath-c-canonical-prefixes-classpath-coverage-cppEnable preprocessing-cpp=-d-dumpbase-dumpdir-dumpmachine-dumpspecs-dumpversion-e-encoding-export-dynamic-extdirs-fCLASSPATH=-fPIC-fPIE-fRTS=Select the runtime-fabi-version=-faccess-control-fada-spec-parent=-falign-commons-falign-functionsAlign the start of functions-falign-functions=-falign-jumps-falign-jumps=-falign-labelsAlign all labels-falign-labels=-falign-loopsAlign the start of loops-falign-loops=-fall-intrinsics-fall-virtual-fallow-leading-underscore-falt-external-templatesNo longer supported-fargument-alias-fargument-noalias-fargument-noalias-anything-fargument-noalias-global-fasmRecognize the "asm" keyword-fassert-fassociative-math-fassume-compiled-fassume-compiled=-fasynchronous-unwind-tables-fauto-inc-dec-fauto-profile-fauto-profile-accurate-fauto-profile=-fautomatic-faux-classpath-fbackslash-fbacktrace-fblas-matmul-limit=-fbootclasspath=-fbootstrap-classes-fbounds-check-fbranch-count-reg-fbranch-probabilities-fbranch-target-load-optimize-fbtr-bb-exclusive-fbuilding-libgcc-fbuiltinRecognize built-in functions-fbuiltin--fcall-saved--fcall-used--fcaller-saves-fcanonical-system-headers-fcheck-array-temporaries-fcheck-branch-annotation-fcheck-data-deps-fcheck-new-fcheck-references-fcheck=-fcilkplusEnable Cilk Plus-fclasspath=-fcoarray=-fcombine-stack-adjustments-fcommon-fcompare-debug-fcompare-debug-second-fcompare-elim-fcompile-resource=-fcond-mismatch-fconserve-space-fconserve-stack-fconstant-string-class=-fconstexpr-depth=-fconvert=big-endian-fconvert=little-endian-fconvert=native-fconvert=swap-fcprop-registers-fcray-pointer-fcrossjumping-fcse-follow-jumps-fcse-skip-blocks-fcx-fortran-rules-fcx-limited-range-fd-lines-as-code-fd-lines-as-comments-fdata-sections-fdbg-cnt-list-fdbg-cnt=-fdce-fdebug-cpp-fdebug-prefix-map=-fdebug-types-section-fdeclone-ctor-dtor-fdeduce-init-list-fdefault-double-8-fdefault-inline-fdefault-integer-8-fdefault-noexcept-fdefault-real-8-fdefer-pop-fdelayed-branch-fdelete-dead-exceptions-fdelete-null-pointer-checks-fdevirtualize-fdevirtualize-speculatively-fdiagnostics-color-fdiagnostics-color=-fdiagnostics-show-caret-fdiagnostics-show-location=-fdiagnostics-show-option-fdirectives-onlyPreprocess directives only.-fdisable--fdisable-assertions-fdisable-assertions=-fdollar-ok-fdollars-in-identifiers-fdse-fdump--fdump-ada-spec-fdump-ada-spec-slim-fdump-core-fdump-final-insns-fdump-fortran-optimized-fdump-fortran-original-fdump-go-spec=-fdump-noaddr-fdump-parse-tree-fdump-passesDump optimization passes-fdump-unnumbered-fdump-unnumbered-links-fdwarf2-cfi-asm-fearly-inliningPerform early inlining-felide-constructors-feliminate-dwarf2-dups-femit-class-debug-always-femit-class-fileOutput a class file-femit-class-filesAlias for -femit-class-file-femit-function-names-femit-struct-debug-baseonly-femit-struct-debug-detailed=-femit-struct-debug-reduced-fenable--fenable-assertions-fenable-assertions=-fencoding=-fenforce-eh-specs-fenum-int-equiv-fexceptionsEnable exception handling-fexcess-precision=-fexec-charset=-fexpensive-optimizations-fext-numeric-literals-fextdirs=-fextended-identifiers-fextern-tls-init-fexternal-blas-fexternal-templates-ff2cUse f2c calling convention-ffast-math-ffat-lto-objects-ffilelist-file-ffinite-math-only-ffixed--ffixed-form-ffixed-line-length--ffixed-line-length-none-ffloat-store-ffor-scope-fforce-addr-fforce-classes-archive-check-fforward-propagate-ffp-contract=-ffpe-summary=-ffpe-trap=-ffree-form-ffree-line-length--ffree-line-length-none-ffreestanding-ffriend-injection-ffrontend-optimizeEnable front end optimization-ffunction-attribute-list=-ffunction-cse-ffunction-sections-fgcse-fgcse-after-reload-fgcse-las-fgcse-lm-fgcse-sm-fgnu-keywords-fgnu-runtime-fgnu-tm-fgnu-unique-fgnu89-inline-fgo-check-divide-overflow-fgo-check-divide-zero-fgo-dump--fgo-optimize--fgo-pkgpath=-fgo-prefix=-fgo-relative-import-path=-fgraphite-fgraphite-identity-fguess-branch-probability-fguiding-decls-fhandle-exceptions-fhash-synchronization-fhelp-fhelp=-fhoist-adjacent-loads-fhonor-std-fhosted-fhuge-objects-fidentProcess #ident directives-fif-conversion-fif-conversion2-fimplement-inlines-fimplicit-inline-templates-fimplicit-none-fimplicit-templates-findirect-classes-findirect-dispatch-findirect-inliningPerform indirect inlining-finhibit-size-directive-finit-character=-finit-integer=-finit-local-zero-finit-logical=-finit-real=-finline-finline-atomics-finline-functions-finline-limit--finline-limit=-finline-small-functions-finput-charset=-finstrument-functions-finteger-4-integer-8-fintrinsic-modules-path-fintrinsic-modules-path=-fipa-cp-fipa-cp-clone-fipa-matrix-reorg-fipa-profile-fipa-pta-fipa-pure-const-fipa-reference-fipa-sra-fipa-struct-reorg-fira-algorithm=-fira-hoist-pressure-fira-loop-pressure-fira-region=-fira-share-save-slots-fira-share-spill-slots-fira-verbose=-fivopts-fjni-fjump-tables-fkeep-inline-dllexport-fkeep-inline-functions-fkeep-static-consts-flabels-ok-flax-vector-conversions-fleading-underscore-flifetime-dse-flive-range-shrinkage-floop-block-floop-flatten-floop-interchange-floop-nest-optimize-floop-optimize-floop-parallelize-allMark all loops as parallel-floop-strip-mine-flto-flto-compression-level=-flto-partition=1to1-flto-partition=balanced-flto-partition=max-flto-partition=none-flto-report-flto-report-wpa-flto=-fltrans-fltrans-output-list=-fmain=-fmath-errno-fmax-array-constructor=-fmax-errors=-fmax-identifier-length=-fmax-stack-var-size=-fmax-subrecord-length=-fmem-report-fmem-report-wpa-fmerge-all-constants-fmerge-constants-fmerge-debug-strings-fmessage-length=-fmodule-private-fmodulo-sched-fmodulo-sched-allow-regmoves-fmove-loop-invariants-fms-extensions-fmudflap-fmudflapir-fmudflapth-fname-mangling-version--fnew-abi-fnext-runtime-fnil-receivers-fno-vect-cost-model-fnon-call-exceptions-fnonansi-builtins-fnonnull-objects-fnothrow-opt-fobjc-abi-version=-fobjc-call-cxx-cdtors-fobjc-direct-dispatch-fobjc-exceptions-fobjc-gc-fobjc-nilcheck-fobjc-sjlj-exceptions-fobjc-std=objc1-fomit-frame-pointer-fopenmp-fopenmp-simd-foperator-names-fopt-info-fopt-info--foptimize-register-move-foptimize-sibling-calls-foptimize-strlen-foptional-diags-foutput-class-dir=-fpack-derived-fpack-struct-fpack-struct=-fpartial-inliningPerform partial inlining-fpcc-struct-return-fpch-deps-fpch-preprocess-fpeel-codesize-limit-fpeel-loopsPerform loop peeling-fpeephole-fpeephole2-fpermissive-fpic-fpie-fplan9-extensions-fplt-fplugin-arg--fplugin=Specify a plugin to load-fpost-ipa-mem-report-fpre-ipa-mem-report-fpredictive-commoning-fprefetch-loop-arrays-fpreprocessed-fpretty-templates-fprofile-fprofile-arcs-fprofile-correction-fprofile-dir=-fprofile-dump-fprofile-generate-fprofile-generate-atomic=-fprofile-generate-buildinfo=-fprofile-generate-sampling-fprofile-generate=-fprofile-reorder-functions-fprofile-report-fprofile-strip=-fprofile-use-fprofile-use=-fprofile-values-fprotect-parens-frandom-seed-frandom-seed=-frange-check-freal-4-real-10-freal-4-real-16-freal-4-real-8-freal-8-real-10-freal-8-real-16-freal-8-real-4-frealloc-lhs-freciprocal-math-frecord-gcc-switches-frecord-marker=4-frecord-marker=8-frecursive-freduced-reflection-free-freg-struct-return-fregmove-frename-registers-freorder-blocks-freorder-functions-freorder-functions=-frepack-arrays-freplace-objc-classes-frepo-frequire-return-statement-frerun-cse-after-loop-frerun-loop-opt-fresolution=The resolution file-fripa-fripa-allow-debug-fripa-disallow-asm-modules-fripa-disallow-opt-mismatch-fripa-inc-path-sub=-frounding-math-frtti-fsanitize-coverage=trace-pc-fsanitize=Select what to sanitize-fsaw-java-file-fsched-dep-count-heuristic-fsched-group-heuristic-fsched-interblock-fsched-last-insn-heuristic-fsched-pressure-fsched-rank-heuristic-fsched-spec-fsched-spec-insn-heuristic-fsched-spec-load-fsched-spec-load-dangerous-fsched-stalled-insns-fsched-stalled-insns-dep-fsched-stalled-insns-dep=-fsched-stalled-insns=-fsched-verbose=-fsched2-use-superblocks-fsched2-use-traces-fschedule-insns-fschedule-insns2-fsecond-underscore-fsection-anchors-fsee-fsel-sched-pipelining-fselective-scheduling-fselective-scheduling2-fshort-double-fshort-enums-fshort-wchar-fshow-column-fshrink-wrap-fshrink-wrap-frame-pointer-fsign-zero-fsignaling-nans-fsigned-bitfields-fsigned-charMake "char" signed by default-fsigned-zeros-fsimd-cost-model=-fsingle-precision-constant-fsized-delete-fsource-filename=-fsource=-fsplit-ivs-in-unroller-fsplit-stack-fsplit-wide-types-fsquangle-fstack-arrays-fstack-check-fstack-check=-fstack-limit-fstack-limit-register=-fstack-limit-symbol=-fstack-protector-fstack-protector-all-fstack-protector-strong-fstack-reuse=-fstack-usage-fstats-fstore-check-fstrength-reduce-fstrict-aliasing-fstrict-enum-precision-fstrict-enums-fstrict-overflow-fstrict-prototype-fstrict-volatile-bitfields-fsync-libcalls-fsyntax-only-ftabstop=-ftarget-help-ftarget=Set the target VM version-ftemplate-backtrace-limit=-ftemplate-depth--ftemplate-depth=-ftest-coverage-fthis-is-variable-fthread-jumps-fthreadsafe-statics-ftime-report-ftls-model=-ftoplevel-reorder-ftracer-ftrack-macro-expansion-ftrack-macro-expansion=-ftrapping-math-ftrapv-ftree-bit-ccp-ftree-builtin-call-dce-ftree-ccp-ftree-ch-ftree-coalesce-inlined-vars-ftree-coalesce-vars-ftree-copy-prop-ftree-copyrename-ftree-cselim-ftree-dce-ftree-dominator-opts-ftree-dseEnable dead store elimination-ftree-forwprop-ftree-fre-ftree-loop-distribution-ftree-loop-if-convert-ftree-loop-if-convert-stores-ftree-loop-im-ftree-loop-ivcanon-ftree-loop-linear-ftree-loop-optimize-ftree-loop-vectorize-ftree-lrs-ftree-parallelize-loops=-ftree-partial-pre-ftree-phiprop-ftree-pre-ftree-pta-ftree-reassoc-ftree-salias-ftree-scev-cprop-ftree-sink-ftree-slp-vectorize-ftree-slsr-ftree-sra-ftree-store-ccp-ftree-store-copy-prop-ftree-switch-conversion-ftree-tail-mergeEnable tail merging on trees-ftree-ter-ftree-vect-loop-version-ftree-vectorizeEnable vectorization on trees-ftree-vectorizer-verbose=-ftree-vrp-ftwo-level-all-subprogs-ftwo-level-line-tables-funderscoring-funit-at-a-time-funroll-all-loops-funroll-codesize-limit-funroll-loops-funsafe-loop-optimizations-funsafe-math-optimizations-funsigned-bitfields-funsigned-char-funswitch-loopsPerform loop unswitching-funwind-tables-fuse-atomic-builtins-fuse-boehm-gc-fuse-cxa-atexit-fuse-cxa-get-exception-ptr-fuse-divide-subroutine-fuse-ld=bfd-fuse-ld=gold-fuse-ld=mcld-fuse-linker-plugin-fvar-trackingPerform variable tracking-fvar-tracking-assignments-fvar-tracking-uninit-fvect-cost-model-fvect-cost-model=-fverbose-asm-fversion-fvisibility-inlines-hidden-fvisibility-ms-compat-fvisibility=-fvpt-fvtable-gc-fvtable-thunks-fvtable-verify=-fvtv-counts-fvtv-debug-fweak-fweb-fwhole-file-fwhole-program-fwide-exec-charset=-fworking-directory-fwpa-fwpa=-fwrapv-fxref-fzee-fzero-initialized-in-bss-fzero-link-gantCatch typos-gcoff-gdwarf-gdwarf--gen-decls-gfission-ggdb-ggnu-pubnames-gline-tables-only-gmlt-gnat-gnatO-gno-fission-gno-pubnames-gno-record-gcc-switches-gno-split-dwarf-gno-strict-dwarf-gpubnames-grecord-gcc-switches-gsplit-dwarf-gstabs-gstabs+-gstrict-dwarf-gvms-gxcoff-gxcoff+-h-idirafter-imacros-include-iquote-iwithprefix-iwithprefixbefore-k8Synonym of -gnatk8-lang-asm-mabi=Specify an ABI-mabort-on-noreturn-mandroid-mapcs-mapcs-float-mapcs-frame-mapcs-reentrantGenerate re-entrant, PIC code-mapcs-stack-check-march=-marm-mbig-endian-mbionicUse Bionic C library-mcallee-super-interworking-mcaller-super-interworking-mcpu=-mfix-cortex-m3-ldrd-mfloat-abi=-mfp16-format=-mfpu=-mglibcUse GNU C library-mhard-float-minline-thumb1-jumptableInline Thumb1 Jump table code-mlittle-endian-mlong-calls-mlra-mneon-for-64bits-mnew-generic-costs-mold-rtx-costs-mpic-data-is-text-relative-mpic-register=-mpoke-function-name-mrestrict-it-msched-prolog-msingle-pic-base-mslow-flash-data-msoft-float-mstructure-size-boundary=-mthumbGenerate code for Thumb state-mthumb-interwork-mtls-dialect=-mtp=-mtpcs-frame-mtpcs-leaf-frame-mtune=-muclibcUse uClibc C library-munaligned-access-mvectorize-with-neon-double-mvectorize-with-neon-quad-mword-relocations-mwords-little-endian-n-no-canonical-prefixes-no-integrated-cpp-no-pie-nocppDisable preprocessing-nodefaultlibs-nostartfiles-nostdinc-nostdinc++-nostdlib-pEnable function profiling-pass-exit-codes-pedantic-pedantic-errors-pg-pie-pipe-posix-print-file-name=-print-libgcc-file-name-print-multi-directory-print-multi-lib-print-multi-os-directory-print-multiarch-print-objc-runtime-info-print-prog-name=-print-search-dirs-print-sysroot-print-sysroot-headers-suffix-profile-pthread-quiet-r-rdynamic-remap-s-s-bc-abi-save-temps-save-temps=-sharedCreate a shared library-shared-libgcc-specs-specs=-static-static-libasan-static-libgcc-static-libgcj-static-libgfortran-static-libgo-static-liblsan-static-libstdc++-static-libtsan-static-libubsan-std=c++03-std=c++0x-std=c++11-std=c++14-std=c++1y-std=c++98-std=c11-std=c1x-std=c89-std=c90-std=c99-std=c9x-std=f2003-std=f2008-std=f2008ts-std=f95-std=gnu-std=gnu++03-std=gnu++0x-std=gnu++11-std=gnu++14-std=gnu++1y-std=gnu++98-std=gnu11-std=gnu1x-std=gnu89-std=gnu90-std=gnu99-std=gnu9x-std=iso9899:1990-std=iso9899:199409-std=iso9899:1999-std=iso9899:199x-std=iso9899:2011-std=legacy-symbolic-t-time-time=-tno-android-cc-tno-android-ld-traditional-traditional-cpp-trigraphs-u-undef-vEnable verbose output-version-wSuppress warnings-wrapper-x-zAdaAdaSCILAdaWhyCC++FortranGoJavaLTOObjCObjC++Valid arguments to -mtp=:unknown IRA algorithm %qsunknown IRA region %qsunknown Stack Reuse Level %qsunknown TLS model %qsTLS dialect to use:: %s%s%s 3934 slots %lu 4024 All options with the desired characteristics have already been displayedThe following options are not documenteddebug format "%s" conflicts with prior selectionunrecognised debug output level "%s"debug output level %s is too highargument %qs to %<-femit-struct-debug-detailed%> unknown%<-femit-struct-debug-detailed=dir:...%> must allow at least as much as %<-femit-struct-debug-detailed=ind:...%>argument %qs to %<-femit-struct-debug-detailed%> not recognizedargument to %<-O%> should be a non-negative integer, %<g%>, %<s%> or %<fast%>section anchors must be disabled when unit-at-a-time is disabledtoplevel reorder must be disabled when unit-at-a-time is disabledtransactional memory is not supported with non-call exceptionssection anchors must be disabled when toplevel reorder is disabled-freorder-blocks-and-partition does not work with exceptions on this architecture-freorder-blocks-and-partition does not support unwind info on this architecture-freorder-blocks-and-partition does not work on this architecture-fno-fat-lto-objects are supported only with linker pluginonly one -flto-partition value can be specified%<-fsplit-stack%> is not supported by this compiler configurationDebug generation via -g option disabled under -fripa -fprofile-generate (use -fripa-allow-debug to override)-fsanitize=address is incompatible with -fsanitize=kernel-address-fsanitize=address and -fsanitize=kernel-address are incompatible with -fsanitize=thread%s: --param arguments should be of the form NAME=VALUE--help argument %q.*s is ambiguous, please be more specificunrecognized argument to --help= option: %q.*sgetting core file size maximum limit: %msetting core file size limit to maximum: %munrecognized gcc debugging option: %cstructure alignment must be a small power of two, not %dunknown stack check parameter "%s"%<-gdwarf%s%> is ambiguous; use %<-gdwarf-%s%> for DWARF version or %<-gdwarf -g%s%> for debug leveldwarf version %d is not supportedunrecognized argument to -fsanitize= option: %q.*s/buildbot/src/android/gcc/toolchain/build/../gcc/gcc-4.9/gcc/opts-common.ccommand line option %qs is not supported by this configurationargument to %qs should be a non-negative integerunrecognized argument in option %qsvalid arguments to %qs are: %s--help=<class> Display descriptions of a specific class of options. <class> is one or more of optimizers, target, warnings, undocumented, params--param <param>=<value> Set parameter <param> to value. See below for a complete list of parameters--print-missing-file-dependencies--print-sysroot-headers-suffix-A<question>=<answer> Assert the <answer> to <question>. Putting '-' before <question> disables the <answer> to <question>Do not discard comments in macro expansions-D<macro>[=<val>] Define a <macro> with <val> as its value. If just <macro> is given, <val> is taken to be 1-F <dir> Add <dir> to the end of the main framework include pathPrint the name of header files as they are used-I <dir> Add <dir> to the end of the main include path-J<directory> Put MODULE files in 'directory'Generate make dependencies and compile-MF <file> Write dependency output to the given fileTreat missing header files as generated filesLike -M but ignore system header filesLike -MD but ignore system header filesGenerate phony targets for all headers-MQ <target> Add a MAKE-quoted targetmissing makefile target after %qs-MT <target> Add an unquoted target-O<number> Set optimization level to <number>Optimize for speed disregarding exact standards complianceOptimize for debugging experience rather than speed or sizeOptimize for space rather than speedDo not generate #line directivesThis switch is deprecated; use -Wextra insteadWarn about things that will change when compiling with an ABI-compliant compilerWarn if a subobject has an abi_tag attribute that the complete object type does not haveWarn about suspicious uses of memory addressesWarn about returning structures, unions or arrays-Waggressive-loop-optimizationsWarn if a loop with constant number of iterations triggers undefined behaviorWarn about possible aliasing of dummy argumentsWarn about alignment of COMMON blocksWarn about missing ampersand in continued character constantsWarn if an array is accessed out of boundsWarn about creation of array temporariesWarn whenever an Objective-C assignment is being intercepted by the garbage collectorWarn about inappropriate attribute usageWarn about casting functions to incompatible typesWarn when a built-in preprocessor macro is undefined or redefinedWarn about C constructs that are not in the common subset of C and C++Deprecated in favor of -Wc++11-compatWarn about C++ constructs whose meaning differs between ISO C++ 1998 and ISO C++ 2011Warn if the type of a variable might be not interoperable with CWarn about pointer casts which increase alignmentWarn about casts which discard qualifiersWarn about subscripts whose type is "char"Warn about truncated character expressionsWarn about variables that might be changed by "longjmp" or "vfork"Warn about possibly nested block comments, and C++ comments spanning more than one physical lineWarn about equality comparisons involving REAL or COMPLEX expressionsWarn for conditionally-supported constructsWarn for implicit type conversions that may change a valueWarn about most implicit conversionsWarn for converting NULL from/to a non-pointer typeWarn in case profiles in -fprofile-use do not matchWarn when a #warning directive is encounteredWarn when all constructors and destructors are privateWarn about __TIME__, __DATE__ and __TIMESTAMP__ usageWarn when a declaration is found after a statementWarn when deleting a pointer to incomplete typeWarn about deleting polymorphic objects with non-virtual destructorsWarn if a deprecated compiler feature, class, method, or field is usedWarn about uses of __attribute__((deprecated)) declarationsWarn when an optimization pass is disabledWarn about compile-time integer division by zeroWarn about implicit conversions from "float" to "double"Warn about violations of Effective C++ style rulesWarn about an empty body in an if or else statementWarn about stray tokens after #elif and #endifWarn about comparison of different enum types-Werror-implicit-function-declarationThis switch is deprecated; use -Werror=implicit-function-declaration insteadTreat specified warning as errorPrint extra (possibly unwanted) warningsWarn if deprecated empty statements are foundExit on the first error occurredWarn for implicit type conversions that cause loss of floating point precisionWarn if testing floating point numbers for equalityDisable promoting warnings to errorsWarn about printf/scanf/strftime/strfmon format string anomaliesWarn about format strings that contain NUL bytesWarn if passing too many arguments to a function for its format stringWarn about format strings that are not literalsWarn about possible security problems with format functionsWarn about strftime formats yielding 2-digit yearsWarn about zero-length formats-Wframe-larger-than=<number> Warn if a function's stack frame requires more than <number> bytesWarn when attempting to free a non-heap objectWarn about function call eliminationWarn whenever type qualifiers are ignored.Warn about implicit declarations-Wimplicit-function-declarationWarn about implicit function declarationsWarn when a declaration does not specify a typeWarn about calls with implicit interfaceWarn about called procedures not explicitly declaredWarn about C++11 inheriting constructors when the base has a variadic constructorWarn about variables which are initialized to themselvesWarn when an inlined function cannot be inlinedWarn when there is a cast to a pointer from an integer of a different sizeWarn if a user-procedure has the same name as an intrinsicWarn on intrinsics not part of the selected standardWarn when an atomic memory model parameter is known to be outside the valid range.Warn about invalid uses of the "offsetof" macroWarn about PCH files that are found but not usedWarn when a jump misses a variable initialization-Wlarger-than=<number> Warn if an object is larger than <number> bytesWarn about truncated source linesWarn when a string or character literal is followed by a ud-suffix which does not begin with an underscore.Warn when a logical operator is suspiciously always evaluating to true or falseDo not warn about using "long long" when -pedanticWarn about suspicious declarations of "main"Warn about maybe uninitialized automatic variablesWarn about possibly missing braces around initializersWarn about global functions without previous declarationsWarn about missing fields in struct initializersWarn about user-specified include directories that do not existWarn about function parameters declared without a type specifier in K&R-style functionsWarn about global functions without prototypesswitch %qs is no longer supportedWarn about use of multi-character character constantsWarn about narrowing conversions within { } that are ill-formed in C++11Warn about "extern" declarations not at file scopeWarn when a noexcept expression evaluates to false even though the expression can't actually throwWarn when non-templatized friend functions are declared within a templateWarn about non-virtual destructorsWarn about NULL being passed to argument slots marked as requiring non-NULL-Wnormalized=<id|nfc|nfkc> Warn about non-normalised Unicode stringsWarn if a C-style cast is used in a programWarn for obsolescent usage in a declarationWarn if an old-style parameter definition is usedWarn if a simd directive is overridden by the vectorizer cost modelWarn if .class files are out of dateWarn about overflow in arithmetic expressionsWarn if a string is longer than the maximum portable length specified by the standardWarn about overloaded virtual function namesWarn about overriding initializers without side effectsWarn when the packed attribute has no effect on struct layoutWarn about packed bit-fields whose offset changed in GCC 4.4Warn when padding is required to align structure membersWarn about possibly missing parenthesesIssue warnings needed for strict compliance to the standardWarn when converting the type of pointers to member functionsWarn about function pointer arithmeticWarn when a pointer differs in signedness in an assignmentWarn when a pointer is cast to an integer of a different sizeWarn for -I and -L options using system directories if cross compilingWarn if a property for an Objective-C object has no assign semantics specifiedWarn if inherited methods are unimplementedWarn about real-literal-constants with 'q' exponent-letterWarn when a left-hand-side array variable is reallocatedWarn when a left-hand-side variable is reallocatedWarn about multiple declarations of the same objectWarn if modifiers are specified when not necessaryWarn when the compiler reorders codeWarn about returning a pointer/reference to a local or temporary variable.Warn whenever a function's return type defaults to "int" (C), or about inconsistent return types (C++)Warn if primary and auxiliary modules have mismatched command line optionsWarn if a selector has multiple methodsWarn when a variable is assigned to itselfWarn when a variable of a non-POD type is assigned to itselfWarn about possible violations of sequence point rulesWarn when one local variable shadows anotherWarn when one local variable shadows another local variable or parameter of compatible typeWarn when one local variable shadows another local variable or parameterWarn about signed-unsigned comparisonsWarn for implicit type conversions between signed and unsigned integersWarn when overload promotes from unsigned to signedWarn when not issuing stack smashing protection for some reasonWarn if stack usage might be larger than specified amountWarn about code which might break strict aliasing rulesWarn about uncasted NULL used as sentinelWarn about optimizations that assume that signed overflow is undefinedWarn about unprototyped function declarationsWarn if type signatures of candidate methods do not match exactlyWarn about functions which might be candidates for __attribute__((const))Warn about functions which might be candidates for format attributesWarn about functions which might be candidates for __attribute__((noreturn))Warn about functions which might be candidates for __attribute__((pure))Warn about "suspicious" constructsWarn about enumerated switches, with no default, missing a caseWarn about enumerated switches missing a "default:" statementWarn about all enumerated switches missing a specific caseWarn when __sync_fetch_and_nand and __sync_nand_and_fetch built-in functions are usedDeprecated. This switch has no effectDo not suppress warnings from system headersPermit nonconforming uses of the tab characterWarn if the pointer in a pointer assignment might outlive its target-Wthread-mismatched-lock-acq-relWarn about mismatched lock acquisition and release-Wthread-mismatched-lock-orderWarn about lock acquisition order inconsistent with what specified in the attributesWarn about a lock being acquired recursivelyWarn about potential thread safety issues when the code is annotated with thread safety attributesDoes nothing. For compatibility with clang thread safety analysis.Warn about function calls not properly protected by locks specified in the attributesWarn about shared variables not properly protected by locks specified in the attributes-Wthread-unsupported-lock-nameWarn about uses of unsupported lock names in attributesWarn about features not present in traditional CWarn of prototypes causing type conversions different from what would happen in the absence of prototypeWarn whenever a trampoline is generatedWarn if trigraphs are encountered that might affect the meaning of the programWarn if a comparison is always true or always false due to the limited range of the data typeWarn about @selector()s without previously declared methodsWarn if an undefined macro is used in an #if directiveWarn about underflow of numerical constant expressionsWarn about uninitialized automatic variablesWarn about unrecognized pragmasDoes nothing. Preserved for backward compatibility.Warn if the loop cannot be optimized due to nontrivial assumptions.Warn about unsuffixed float constantsWarn when a function parameter is only set, otherwise unusedWarn when a variable is only set, otherwise unusedWarn about unused dummy arguments.Warn when a function is unusedWarn when typedefs locally defined in a function are not usedWarn about macros defined in the main file that are not usedWarn when a function parameter is unusedWarn if a caller of a function, marked with attribute warn_unused_result, does not use its return valueWarn when an expression value is unusedWarn when a variable is unusedWarn about questionable usage of the macros used to retrieve variable argumentsWarn about using variadic macros-Wvector-operation-performanceWarn when a vector operation is compiled outside the SIMDWarn if a virtual base has a non-trivial move assignment operatorWarn if a variable length array is usedWarn when a register variable is declared volatileIn C++, nonzero means warn about deprecated conversion from string literals to 'char *'. In C, similar warning, except that the conversion is of course not deprecated by the ISO C standard.-Wzero-as-null-pointer-constantWarn when a literal '0' is used as null pointerA synonym for -std=c89 (for C) or -std=c++98 (for C++)-aux-info <file> Emit declaration information into <file>-d<letters> Enable dumps from specific passes of the compiler-dumpbase <file> Set the file basename to be used for dumps-dumpdir <dir> Set the directory name to be used for dumps--CLASSPATH Deprecated; use --classpath insteadGenerate position-independent code if possible (large mode)Generate position-independent code for executables if possible (large mode)Enforce class member access control semantics-fada-spec-parent=unit Dump Ada specs as child units of given parent-faggressive-function-eliminationEliminate multiple function invokations also for impure functions-faggressive-loop-optimizationsAggressively optimize loops using language constraintsEnable alignment of COMMON blocksAlign labels which are only reached by jumpingAll intrinsics procedures are available regardless of selected standard-fallow-parameterless-variadic-functionsAllow variadic functions without named parameterPermit the use of the assert keywordAllow optimization for floating-point arithmetic which may change the result of the operation due to rounding.Generate unwind tables that are exact at each instruction boundaryGenerate auto-inc/dec instructionsUse sample profile information for call graph node weights. The default profile file is fbdata.afdo in 'pwd'.Whether to assume the sample profile is accurate.-fauto-profile-record-coverage-in-elfWhether to record annotation coverage info in elf.Use sample profile information for call graph node weights. The profile file is specified in the argument.Do not treat local variables and COMMON blocks as if they were named in SAVE statementsSpecify that backslash in string introduces an escape characterProduce a backtrace when a runtime error is encountered-fblas-matmul-limit=<n> Size of the smallest matrix for which matmul will use BLAS--bootclasspath=<path> Replace system pathGenerated should be loaded by bootstrap loaderGenerate code to check bounds before indexing arraysReplace add, compare, branch with branch on count registerUse profiling information for branch probabilitiesPerform branch target load optimization before prologue / epilogue threading-fbranch-target-load-optimize2Perform branch target load optimization after prologue / epilogue threadingRestrict target load migration not to re-use registers in any basic block-fcall-saved-<register> Mark <register> as being preserved across functions-fcall-used-<register> Mark <register> as being corrupted by function callsSave registers around function callsWhere shorter, use canonicalized paths to systems headers.Produce a warning at runtime if a array temporary has been created for a procedure argumentCompare branch prediction result and autofdo profile information, store the result in a section in the generated elf file.-fcheck-branch-annotation-threshold=The number of executions a basic block needs to reach before GCC dumps its branch prediction information with -fcheck-branch-annotation.Compare the results of several data dependence analyzers.Check the return value of new in C++Generate checks for references to NULL-fcheck=[...] Specify which runtime checks are to be performed--classpath=<path> Set class path-fcoarray=[...] Specify which coarray parallelization should be usedLooks for opportunities to reduce stack adjustments and stack references.Do not put uninitialized globals in the common sectionRun only the second compilation of -fcompare-debug-fcompare-debug[=<opts>] Compile with and without e.g. -gtoggle, and compare the final-insns dumpPerform comparison elimination after register allocation has finishedAllow the arguments of the '?' operator to have different typesDoes nothing. Preserved for backward compatibility.Do not perform optimizations increasing noticeably stack usage-fconst-string-class=<name> Use class <name> for constant stringsno class name specified with %qs-fconstexpr-depth=<number> Specify maximum constexpr recursion depthUse big-endian format for unformatted filesUse little-endian format for unformatted filesUse native format for unformatted filesSwap endianness for unformatted filesPerform a register copy-propagation optimization passUse the Cray Pointer extensionPerform cross-jumping optimizationWhen running CSE, follow jumps to their targetsComplex multiplication and division follow Fortran rulesOmit range reduction step when performing complex divisionIgnore 'D' in column one in fixed formTreat lines with 'D' in column one as commentsPlace data items into their own sectionList all available debugging counters with their limits and counts.-fdbg-cnt=<counter>:<limit>[,<counter>:<limit>,...] Set the debug counter limit. Use the RTL dead code elimination passEmit debug annotations during preprocessingMap one directory name to another in debug informationOutput .debug_types section when using DWARF v4 debuginfo.Factor complex constructors and destructors to favor space over speed-fdeduce-init-list enable deduction of std::initializer_list for a template type parameter from a brace-enclosed initializer-listSet the default double precision kind to an 8 byte wide typeSet the default integer kind to an 8 byte wide typeMake functions no-throw/noexcept by defaultSet the default real kind to an 8 byte wide typeDefer popping functions args from stack until laterAttempt to fill delay slots of branch instructionsDelete dead instructions that may throw exceptionsDelete useless null pointer checksTry to convert virtual calls to direct ones.Perform speculative devirtualization-fdiagnostics-color=[never|always|auto] Colorize diagnosticsShow the source line with a caret indicating the column-fdiagnostics-show-location=[once|every-line] How often to emit source location at the beginning of line-wrapped diagnosticsAmend appropriate diagnostic messages with the command line option that controls them-fdisable-[tree|rtl|ipa]-<pass>=range1+range2 disables an optimization passAllow dollar signs in entity namesPermit '$' as an identifier characterUse the RTL dead store elimination pass-fdump-<type> Dump various compiler internals to a fileWrite all declarations as Ada code transitivelyWrite all declarations as Ada code for the given file only-fdump-final-insns=filename Dump to filename the insns at the end of translationDisplay the code tree after front end optimizationDisplay the code tree after parsing-fdump-go-spec=filename Write all declarations to file as Go codeSuppress output of addresses in debugging dumpsDisplay the code tree after parsing; deprecated optionSuppress output of instruction numbers, line number notes and addresses in debugging dumpsSuppress output of previous and next insn numbers in debugging dumpsEnable CFI tables via GAS assembler directives.Perform DWARF2 duplicate elimination-feliminate-unused-debug-symbolsPerform unused type elimination in debug info-feliminate-unused-debug-typesDo not suppress C++ class debug information.Print to stderr the mapping from module name and function id to assembler function name when -ftest-coverage, -fprofile-generate or -fprofile-use are active, for use in correlating function ids in gcda files with the function name.-femit-struct-debug-baseonly Aggressive reduced debug info for structs-femit-struct-debug-detailed=<spec-list> Detailed reduced debug info for structs-femit-struct-debug-reduced Conservative reduced debug info for structs-fenable-[tree|rtl|ipa]-<pass>=range1+range2 enables an optimization pass--encoding=<encoding> Choose input encoding (defaults from your locale)Generate code to check exception specifications-fexcess-precision=[fast|standard] Specify handling of excess floating-point precision-fexec-charset=<cset> Convert all strings and character constants to character set <cset>Perform a number of minor, expensive optimizationsInterpret imaginary, fixed-point, or other gnu number suffix as the corresponding number literal rather than a user-defined number literal.--extdirs=<path> Set the extension directory pathPermit universal character names (\u and \U) in identifiersSupport dynamic initialization of thread-local variables in a different translation unitSpecify that an external BLAS library should be used for matmul calls on large-size arraysOutput lto objects containing both the intermediate language and binary output.Input file is a file with a list of filenames to compileAssume no NaNs or infinities are generated-ffixed-<register> Mark <register> as being unavailable to the compilerAssume that the source file is fixed form-ffixed-line-length-<n> Use n as character line width in fixed modeAllow arbitrary character line width in fixed modeDon't allocate floats and doubles in extended-precision registersScope of for-init-statement variables is local to the loopAlways check for non gcj generated classes archivesPerform a forward propagation pass on RTL-ffp-contract=[off|on|fast] Perform floating-point expression contraction.-ffpe-summary=[...] Print summary of floating point exceptions-ffpe-trap=[...] Stop on following floating point exceptionsAssume that the source file is free form-ffree-line-length-<n> Use n as character line width in free modeAllow arbitrary character line width in free modeDo not assume that standard C libraries and "main" existInject friend functions into enclosing namespace-ffunction-attribute-list=attribute:name,... Add attribute to named functionsAllow function addresses to be held in registersPlace each function into its own sectionPerform global common subexpression eliminationPerform global common subexpression elimination after register allocation has finishedPerform redundant load after store elimination in global common subexpression eliminationPerform enhanced load motion during global common subexpression eliminationPerform store motion after global common subexpression eliminationRecognize GNU-defined keywordsGenerate code for GNU runtime environmentEnable support for GNU transactional memoryUse STB_GNU_UNIQUE if supported by the assemblerUse traditional GNU semantics for inline functionsAdd explicit checks for division overflow in INT_MIN / -1Add explicit checks for division by zero-fgo-dump-<type> Dump Go frontend internal information-fgo-optimize-<type> Turn on optimization passes in the frontend-fgo-pkgpath=<string> Set Go package path-fgo-prefix=<string> Set package-specific prefix for exported Go names-fgo-relative-import-path=<path> Treat a relative import as relative to pathEnable in and out of Graphite representationEnable Graphite Identity transformationEnable guessing of branch probabilities-fhandle-exceptions has been renamed -fexceptions (and is now on by default)Assume the runtime uses a hash table to map an object to its synchronization structureEnable hoisting adjacent loads to encourage generating conditional move instructionsAssume normal C execution environmentPerform conversion of conditional jumps to branchless equivalentsPerform conversion of conditional jumps to conditional executionExport functions even if they can be inlinedEmit implicit instantiations of inline templatesSpecify that no implicit typing is allowed, unless overridden by explicit IMPLICIT statementsEmit implicit instantiations of templatesGenerate instances of Class at runtimeUse offset tables for virtual method callsDo not generate .size directives-finit-character=<n> Initialize local character variables to ASCII value n-finit-integer=<n> Initialize local integer variables to nInitialize local variables to zero (from g77)-finit-logical=<true|false> Initialize local logical variables-finit-real=<zero|nan|inf|-inf> Initialize local real variablesEnable inlining of function declared "inline", disabling disables all inliningInline __atomic operations when a lock free instruction sequence is available.Integrate functions not declared "inline" into their callers when profitable-finline-functions-called-onceIntegrate functions only required by their single caller-finline-limit=<number> Limit the size of inlined functions to <number>Integrate functions into their callers when code size is known not to grow-finput-charset=<cset> Specify the default character set for source filesInstrument function entry and exit with profiling calls-finstrument-functions-exclude-file-list=-finstrument-functions-exclude-file-list=filename,... Do not instrument functions listed in files-finstrument-functions-exclude-function-list=-finstrument-functions-exclude-function-list=name,... Do not instrument listed functionsInterpret any INTEGER(4) as an INTEGER(8)Specify where to find the compiled intrinsic modulesPerform interprocedural constant propagationPerform cloning to make Interprocedural constant propagation strongerPerform interprocedural profile propagationPerform interprocedural points-to analysisDiscover pure and const functionsDiscover readonly and non addressable static variablesPerform interprocedural reduction of aggregates-fira-algorithm=[CB|priority] Set the used IRA algorithmUse IRA based register pressure calculation in RTL hoist optimizations.Use IRA based register pressure calculation in RTL loop optimizations.-fira-region=[one|all|mixed] Set regions for IRAShare slots for saving different hard registers.Share stack slots for spilled pseudo-registers.-fira-verbose=<number> Control IRA's level of diagnostic messages.-fisolate-erroneous-paths-attributeDetect paths which trigger erroneous or undefined behaviour due a NULL value being used in a way which is forbidden by a returns_nonnull or nonnull attribute. Isolate those paths from the main control flow and turn the statement with erroneous or undefined behaviour into a trap. -fisolate-erroneous-paths-dereferenceDetect paths which trigger erroneous or undefined behaviour due to dereferencing a NULL pointer. Isolate those paths from the main control flow and turn the statement with erroneous or undefined behaviour into a trap.Optimize induction variables on treesAssume native functions are implemented using JNIUse jump tables for sufficiently large switch statementsDon't emit dllexported inline functions unless neededGenerate code for functions even if they are fully inlinedEmit static const variables even if they are not usedAllow implicit conversions between vectors with differing numbers of subparts and/or differing element types.Give external symbols a leading underscoreTell DSE that the storage for a C++ object is dead when the constructor starts and when the destructor finishes.Relief of register pressure through live range shrinkageEnable Loop Blocking transformationEnable Loop Interchange transformationEnable the ISL based loop nest optimizerEnable Loop Strip Mining transformationEnable link-time optimization.-flto-compression-level=<number> Use zlib compression level <number> for ILPartition symbols and vars at linktime based on object files they originate fromPartition functions and vars at linktime into approximately same sized bucketsPut every symbol into separate partitionDisable partioning and streamingReport various link-time optimization statisticsReport various link-time optimization statistics for WPA onlyLink-time optimization with number of parallel jobs or jobserver.Run the link-time optimizer in local transformation (LTRANS) mode.Specify a file to which a list of files output by LTRANS is written.Set errno after built-in math functions-fmax-array-constructor=<n> Maximum number of objects in an array constructor-fmax-errors=<number> Maximum number of errors to report-fmax-identifier-length=<n> Maximum identifier length-fmax-stack-var-size=<n> Size in bytes of the largest array that will be put on the stack-fmax-subrecord-length=<n> Maximum length for subrecordsReport on permanent memory allocationReport on permanent memory allocation in WPA onlyAttempt to merge identical constants and constant variablesAttempt to merge identical constants across compilation unitsAttempt to merge identical debug strings across compilation units-fmessage-length=<number> Limit diagnostics to <number> characters per line. 0 suppresses line-wrappingSet default accessibility of module entities to PRIVATE.Perform SMS based modulo scheduling before the first scheduling passPerform SMS based modulo scheduling with register moves allowedMove loop invariant computations out of loopsDon't warn about uses of Microsoft extensionsGenerate code for NeXT (Apple Mac OS X) runtime environmentAssume that receivers of Objective-C messages may be nilEnables the unlimited vectorizer cost model. Preserved for backward compatibility.Support synchronous non-call exceptionsTreat a throw() exception specification as noexcept to improve code sizeSpecify which ABI to use for Objective-C family code and meta-data generation.Generate special Objective-C methods to initialize/destroy non-POD C++ ivars, if neededAllow fast jumps to the message dispatcherEnable Objective-C exception and synchronization syntaxEnable garbage collection (GC) in Objective-C/Objective-C++ programsEnable inline checks for nil receivers with the NeXT runtime and ABI version 2.Enable Objective-C setjmp exception handling runtimeConform to the Objective-C 1.0 language as implemented in GCC 4.0When possible do not generate stack framesEnable OpenMP (implies -frecursive in Fortran)Enable OpenMP's SIMD directivesRecognize C++ keywords like "compl" and "xor"Enable all optimization info dumps on stderr-fopt-info[-<type>=filename] Dump compiler optimization detailsOptimize sibling and tail recursive calls-foptimize-static-class-initializationEnable optimization of static class initialization codeEnable string length optimizations on treesTry to lay out derived types as compactly as possiblePack structure members together without holes-fpack-struct=<number> Set initial maximum structure member alignmentReturn small aggregates in memory, not registersLook for and use PCH files even when preprocessingLimit non-const non-FP loop peeling under profile estimates of large code footprintEnable machine specific peephole optimizationsEnable an RTL peephole pass before sched2Downgrade conformance errors to warningsGenerate position-independent code if possible (small mode)Generate position-independent code for executables if possible (small mode)Enable Plan 9 language extensionsUse PLT for PIC calls (-fno-plt: load the address from GOT at call site)-fplugin-arg-<name>-<key>[=<value>] Specify argument <key>=<value> for plugin <name>Report on memory allocation before interprocedural optimizationRun predictive commoning optimization.Generate prefetch instructions, if available, for arrays in loopsTreat the input file as already preprocessed-fno-pretty-templates Do not pretty-print template specializations as the template signature followed by the argumentsEnable basic program profiling codeInsert arc-based program profiling codeEnable correction of flow inconsistent profile data inputSet the top-level directory for storing the profile data. The default is 'pwd'.Dump CFG profile for comparison.Enable common options for generating profile info for profile feedback directed optimizationsfprofile-generate-atomic=[0..3] Atomically increments for profile counters.-fprofile-generate-buildinfo=filename Read build info to include in gcda file from filenameTurn on instrumentation sampling with -fprofile-generate with rate set by --param profile-generate-sampling-rate or environment variable GCOV_SAMPLING_RATEEnable common options for generating profile info for profile feedback directed optimizations, and set -fprofile-dir=Enable function reordering that improves code placementReport on consistency of profileSpecify a substring to be stripped from the profile base file nameEnable common options for performing profile feedback directed optimizationsEnable common options for performing profile feedback directed optimizations, and set -fprofile-dir=Insert code to profile values of expressionsProtect parentheses in expressions-frandom-seed=<string> Make compile reproducible using <string>Enable range checking during compilationInterpret any REAL(4) as a REAL(10)Interpret any REAL(4) as a REAL(16)Interpret any REAL(4) as a REAL(8)Interpret any REAL(8) as a REAL(10)Interpret any REAL(8) as a REAL(16)Interpret any REAL(8) as a REAL(4)Reallocate the LHS in assignmentsSame as -fassociative-math for expressions which include division.-frecord-compilation-info-in-elfRecord the compiler optimizations in a .gnu.switches.text section.Record gcc command line switches in the object file.Use a 4-byte record marker for unformatted filesUse an 8-byte record marker for unformatted filesAllocate local variables on the stack to allow indirect recursionReduce the amount of reflection meta-data generatedTurn on Redundant Extensions Elimination pass.Return small aggregates in registersPerform a register renaming optimization passReorder basic blocks to improve code placement-freorder-blocks-and-partitionReorder basic blocks and partition into hot and cold sectionsReorder functions to improve code placement-freorder-functions=[callgraph] Select the scheme for function reordering. This invokes a linker plugin. Generate .gnu.callgraph.text sections listing callees and edge counts.Copy array sections into a contiguous block on procedure entryUsed in Fix-and-Continue mode to indicate that object files may be swapped in at runtimeEnable automatic template instantiationFunctions which return values must end with return statementsAdd a common subexpression elimination pass after loop optimizations-freschedule-modulo-scheduled-loopsEnable/Disable the traditional scheduling in loops that already passed modulo schedulingPerform Dynamic Inter-Procedural Analysis.Allow -g enablement for -fripa -fprofile-generate compiles.Don't import an auxiliary module if it contains asm statementsDon't import an auxiliary module if the command line options mismatch with the primary moduleSubstitute substring in include paths with a new string to allow reuse profile data-fripa-no-promote-always-inline-funcDon't promote always inline static functions assuming they will be inlined and no copy is needed.Disable optimizations that assume default FP rounding behaviorGenerate run time type descriptor informationEnable coverage-guided fuzzing code instrumentation. Inserts call to __sanitizer_cov_trace_pc into every basic block.-fsched-critical-path-heuristicEnable the critical path heuristic in the schedulerEnable the dependent count heuristic in the schedulerEnable the group heuristic in the schedulerEnable scheduling across basic blocksEnable the last instruction heuristic in the schedulerEnable register pressure sensitive insn schedulingEnable the rank heuristic in the schedulerAllow speculative motion of non-loadsEnable the speculative instruction heuristic in the schedulerAllow speculative motion of some loadsAllow speculative motion of more loadsAllow premature scheduling of queued insnsSet dependence distance checking in premature scheduling of queued insns-fsched-stalled-insns-dep=<number> Set dependence distance checking in premature scheduling of queued insns-fsched-stalled-insns=<number> Set number of queued insns that can be prematurely scheduled-fsched-verbose=<number> Set the verbosity level of the schedulerIf scheduling post reload, do superblock schedulingReschedule instructions before register allocationReschedule instructions after register allocationAppend a second underscore if the name already contains an underscoreAccess data in the same section from shared anchor pointsPerform software pipelining of inner loops during selective scheduling-fsel-sched-pipelining-outer-loopsPerform software pipelining of outer loops during selective scheduling-fsel-sched-reschedule-pipelinedReschedule pipelined regions without pipeliningSchedule instructions using selective scheduling algorithmRun selective scheduling after reloadUse the same size for double as for floatUse the narrowest integer type possible for enumeration typesForce the underlying type for "wchar_t" to be "unsigned short"Show column numbers in diagnostics, when available. Default onEmit function prologues only before parts of the function that need it, rather than at the top of the function.Framepointer shrinkwrapping optimization.Apply negative sign to zero valuesDisable optimizations observable by IEEE signaling NaNsWhen "signed" or "unsigned" is not given make the bitfield signedDisable floating point optimizations that ignore the IEEE signedness of zeroSpecifies the vectorization cost model for code marked with a simd directiveConvert floating point constants to single precision constantsSupport delete operator with objetc's size as the second parameter.Set the source language versionSplit lifetimes of induction variables when loops are unrolledGenerate discontiguous stack framesSplit wide types into independent registersPut all local arrays on stack.Insert stack checking code into the program. Same as -fstack-check=specific-fstack-check=[no|generic|specific] Insert stack checking code into the program-fstack-limit-register=<register> Trap if the stack goes past <register>-fstack-limit-symbol=<name> Trap if the stack goes past symbol <name>Use propolice as a stack protection methodUse a stack protection method for every functionUse a smart stack protection method for certain functions-fstack-reuse=[all|named_vars|none] Set stack reuse level for local variables.Output stack usage information on a per-function basisDisplay statistics accumulated during compilationEnable assignability checks for stores into object arraysAssume strict aliasing rules applyPerform transformations based on enum precisionAssume that values of enumeration type are always within the minimum range of that typeTreat signed overflow as undefinedForce bitfield accesses to match their type widthImplement __atomic operations via libcalls to legacy __sync functionsCheck for syntax errors, then stop-ftabstop=<number> Distance between tab stops for column reportingSet the maximum number of template instantiation notes for a single warning or error-ftemplate-depth=<number> Specify maximum template instantiation depthCreate data files needed by "gcov"Perform jump threading optimizations-fno-threadsafe-statics Do not generate thread-safe code for initializing local staticsReport the time taken by each compiler pass-ftls-model=[global-dynamic|local-dynamic|initial-exec|local-exec] Set the default thread-local storage code generation modelReorder top level functions, variables, and asmsPerform superblock formation via tail duplication-ftrack-macro-expansion=<0|1|2> Track locations of tokens coming from macro expansion and display them in error messagesAssume floating-point operations can trapTrap for signed overflow in addition, subtraction and multiplicationEnable SSA-BIT-CCP optimization on treesEnable conditional dead code elimination for builtin callsEnable SSA-CCP optimization on treesEnable loop header copying on treesEnable coalescing of copy-related user variables that are inlinedEnable coalescing of all copy-related user variablesEnable copy propagation on treesReplace SSA temporaries with better names in copiesTransform condition stores into unconditional onesEnable SSA dead code elimination optimization on treesEnable dominator optimizationsEnable forward propagation on treesEnable Full Redundancy Elimination (FRE) on trees-ftree-loop-distribute-patternsEnable loop distribution for patterns transformed into a library callEnable loop distribution on treesConvert conditional jumps in innermost loops to branchless equivalentsAlso if-convert conditional jumps containing memory writesEnable loop invariant motion on treesCreate canonical induction variables in loopsEnable loop interchange transforms. Same as -floop-interchangeEnable loop optimizations on tree levelEnable loop vectorization on treesPerform live range splitting during the SSA->normal passEnable automatic parallelization of loopsIn SSA-PRE optimization on trees, enable partial-partial redundancy eliminationEnable hoisting loads from conditional pointers.Enable SSA-PRE optimization on treesPerform function-local points-to analysis on trees.Enable reassociation on tree levelEnable copy propagation of scalar-evolution information.Enable SSA code sinking on treesEnable basic block vectorization (SLP) on treesPerform straight-line strength reductionPerform scalar replacement of aggregatesPerform conversions of switch initializations.Replace temporary expressions in the SSA->normal passPerform Value Range Propagation on treesWhen generating two-level line tables in DWARF (experimental), add linkage names for all functions (not just inlined functions).Use two-level line tables in DWARF (experimental).Append underscores to externally visible namesCompile whole compilation unit at a timePerform loop unrolling for all loopsLimit non-const non-FP loop unrolling under profile estimates of large code footprintPerform loop unrolling when iteration count is knownAllow loop optimizations to assume that the loops behave in normal wayAllow math optimizations that may violate IEEE or ISO standardsWhen "signed" or "unsigned" is not given make the bitfield unsignedMake "char" unsigned by defaultJust generate unwind tables for exception handlingGenerate code for built-in atomic operationsGenerate code for the Boehm GCUse __cxa_atexit to register destructorsUse __cxa_get_exception_ptr in exception handlingCall a library routine to do integer divisionsUse the bfd linker instead of the default linkerUse the gold linker instead of the default linkerUse the mcld linker instead of the default linkerPerform variable tracking by annotating assignments-fvar-tracking-assignments-toggleToggle -fvar-tracking-assignmentsPerform variable tracking and also tag variables that are uninitialized-fvariable-expansion-in-unrollerApply variable expansion when loops are unrolledEnables the dynamic vectorizer cost model. Preserved for backward compatibility.Specifies the cost model for vectorizationAdd extra commentary to assembler outputMarks all inlined functions and methods as having hidden visibilityChanges visibility to match Microsoft Visual Studio by default-fvisibility=[default|internal|hidden|protected] Set the default symbol visibilityUse expression value profiles in optimizationsValidate vtable pointers before using them.Output vtable verification counters.Output vtable verification pointer sets information.Emit common-like symbols as weak symbolsConstruct webs and split unrelated uses of single variablePerform whole program optimizations-fwide-exec-charset=<cset> Convert all wide strings and character constants to character set <cset>Generate a #line directive pointing at the current working directoryRun the link-time optimizer in whole program analysis (WPA) mode.Whole program analysis (WPA) mode with number of parallel jobs specified.Assume signed arithmetic overflow wraps aroundPut zero initialized data in the bss sectionGenerate lazy class lookup (via objc_getClass()) for use in Zero-Link modeGenerate debug information in default formatGenerate debug information in COFF formatGenerate debug information in default version of DWARF formatGenerate debug information in DWARF v2 (or later) formatDump declarations to a .decl fileGenerate debug information in default extended formatGenerate DWARF pubnames and pubtypes sections with GNU extensions.Generate DWARF line number tables and no other debug sectionsGenerate debug information at level 1 with minimal line table-gnat<options> Specify options to GNATSet name of output ALI file (internal switch)Don't generate DWARF pubnames and pubtypes sections.Don't record gcc command line switches in DWARF DW_AT_producer.Don't generate debug information in separate .dwo filesEmit DWARF additions beyond selected versionGenerate DWARF pubnames and pubtypes sections.Record gcc command line switches in DWARF DW_AT_producer.Generate debug information in separate .dwo filesGenerate debug information in STABS formatGenerate debug information in extended STABS formatDon't emit DWARF additions beyond selected versionToggle debug information generationGenerate debug information in VMS formatGenerate debug information in XCOFF formatGenerate debug information in extended XCOFF format-idirafter <dir> Add <dir> to the end of the system include path-imacros <file> Accept definition of macros in <file>-imultiarch <dir> Set <dir> to be the multiarch include subdirectory-imultilib <dir> Set <dir> to be the multilib include subdirectory-include <file> Include the contents of <file> before other files-iplugindir=<dir> Set <dir> to be the default plugin directory-iprefix <path> Specify <path> as a prefix for next two options-iquote <dir> Add <dir> to the end of the quote include path-isysroot <dir> Set <dir> to be the system root directory-isystem <dir> Add <dir> to the start of the system include path-iwithprefix <dir> Add <dir> to the end of the system include path-iwithprefixbefore <dir> Add <dir> to the end of the main include pathGenerate a call to abort if a noreturn function returnsGenerate code for the Android platform.Pass FP arguments in FP registersGenerate APCS conformant stack framesSpecify the name of the target architectureGenerate code in 32 bit ARM state.Assume target CPU is configured as big endianThumb: Assume non-static functions may be called from ARM codeThumb: Assume function pointers may go to non-Thumb aware codeSpecify the name of the target CPUAvoid overlapping destination and address registers on LDRD instructions that may trigger Cortex-M3 errata.Specify if floating point hardware should be usedSpecify the __fp16 floating-point formatSpecify the name of the target floating point hardware/formatAssume target CPU is configured as little endianGenerate call insns as indirect calls, if necessaryUse LRA instead of reload (transitional)Use Neon to perform 64-bits operations rather than core registers.Use the new generic RTX cost tables if new core-specific cost table not available (transitional).Use the old RTX costing tables (transitional).Assume data segments are relative to text segment.Specify the register to be used for PIC addressingStore function names in object codeGenerate IT blocks appropriate for ARMv8.Permit scheduling of a function's prologue sequenceDo not load the PIC register in function prologuesAssume loading data from flash is slower than fetching instructions.Specify the minimum bit alignment of structuresSupport calls between Thumb and ARM instruction setsSpecify thread local storage schemeSpecify how to access the thread pointerThumb: Generate (non-leaf) stack frames even if not neededThumb: Generate (leaf) stack frames even if not neededTune code for the given processorEnable unaligned word and halfword accesses to packed data.Use Neon double-word (rather than quad-word) registers for vectorizationUse Neon quad-word (rather than double-word) registers for vectorizationOnly generate absolute relocations on word sized values.Assume big endian bytes, little endian words. This option is deprecated.Create a position dependent executableDo not search standard system include directories (those specified with -isystem will still be used)Do not search standard system include directories for C++Do not look for object files in standard path-o <file> Place output into <file>Like -pedantic but issue them as errorsCreate a position independent executableGenerate C header of platform-specific featuresDo not display functions compiled or elapsed timeRemap file names when including filesStatically link the GNU Fortran helper library (libgfortran)Conform to the ISO 1998 C++ standard revised by the 2003 technical corrigendumDeprecated in favor of -std=c++11Conform to the ISO 2011 C++ standardConform to the ISO 2014(?) C++ draft standard (experimental and incomplete support)Conform to the ISO 2011 C standard (experimental and incomplete support)Deprecated in favor of -std=c11Conform to the ISO 1990 C standardConform to the ISO 1999 C standardDeprecated in favor of -std=c99Conform to the ISO Fortran 2003 standardConform to the ISO Fortran 2008 standardConform to the ISO Fortran 2008 standard including TS 29113Conform to the ISO Fortran 95 standardConform to nothing in particularConform to the ISO 1998 C++ standard revised by the 2003 technical corrigendum with GNU extensionsDeprecated in favor of -std=gnu++11Conform to the ISO 2011 C++ standard with GNU extensions (experimental and incomplete support)Conform to the ISO 201y(7?) C++ draft standard with GNU extensions (experimental and incomplete support)Conform to the ISO 2011 C standard with GNU extensions (experimental and incomplete support)Deprecated in favor of -std=gnu11Conform to the ISO 1990 C standard with GNU extensionsConform to the ISO 1999 C standard with GNU extensionsDeprecated in favor of -std=gnu99Conform to the ISO 1990 C standard as amended in 1994Deprecated in favor of -std=iso9899:1999Accept extensions to support legacy codeEnable traditional preprocessing-trigraphs Support ISO C trigraphsDo not predefine system-specific and GCC-specific macrosDisplay the compiler's versionKnown ARM ABIs (for use with the -mabi= option):Known ARM architectures (for use with the -march= option):Known __fp16 formats (for use with the -mfp16-format= option):Known ARM FPUs (for use with the -mfpu= option):unknown excess precision style %qsKnown floating-point ABIs (for use with the -mfloat-abi= option):unknown floating point contraction style %qsunrecognized function reorder value %qsKnown ARM CPUs (for use with the -mcpu= and -mtune= options):unrecognized visibility value %qsunknown vectorizer cost model %qsunknown vtable verify initialization priority %qs/buildbot/src/android/gcc/toolchain/build/../gcc/gcc-4.9/gcc/vec.c/buildbot/src/android/gcc/toolchain/build/../gcc/gcc-4.9/gcc/hooks.c%s: all warnings being treated as errors%s: some warnings being treated as errorsIn file included from %r%s:%d:%d%RIn file included from %r%s:%d%R,
/prebuilts/tools/common/m2/repository/org/mozilla/rhino/1.7R3/
H A D	rhino-1.7R3.jar	META-INF/ META-INF/MANIFEST.MF org/ org/mozilla/ org/mozilla/classfile/ org/mozilla/javascript/ org/ ...
/prebuilts/gdb/darwin-x86/lib/python2.7/pydoc_data/
H A D	topics.py	6 'attribute-access': '\nCustomizing attribute access\n****************************\n\nThe following methods can be defined to customize the meaning of\nattribute access (use of, assignment to, or deletion of ``x.name``)\nfor class instances.\n\nobject.__getattr__(self, name)\n\n Called when an attribute lookup has not found the attribute in the\n usual places (i.e. it is not an instance attribute nor is it found\n in the class tree for ``self``). ``name`` is the attribute name.\n This method should return the (computed) attribute value or raise\n an ``AttributeError`` exception.\n\n Note that if the attribute is found through the normal mechanism,\n ``__getattr__()`` is not called. (This is an intentional asymmetry\n between ``__getattr__()`` and ``__setattr__()``.) This is done both\n for efficiency reasons and because otherwise ``__getattr__()``\n would have no way to access other attributes of the instance. Note\n that at least for instance variables, you can fake total control by\n not inserting any values in the instance attribute dictionary (but\n instead inserting them in another object). See the\n ``__getattribute__()`` method below for a way to actually get total\n control in new-style classes.\n\nobject.__setattr__(self, name, value)\n\n Called when an attribute assignment is attempted. This is called\n instead of the normal mechanism (i.e. store the value in the\n instance dictionary). *name* is the attribute name, *value* is the\n value to be assigned to it.\n\n If ``__setattr__()`` wants to assign to an instance attribute, it\n should not simply execute ``self.name = value`` --- this would\n cause a recursive call to itself. Instead, it should insert the\n value in the dictionary of instance attributes, e.g.,\n ``self.__dict__[name] = value``. For new-style classes, rather\n than accessing the instance dictionary, it should call the base\n class method with the same name, for example,\n ``object.__setattr__(self, name, value)``.\n\nobject.__delattr__(self, name)\n\n Like ``__setattr__()`` but for attribute deletion instead of\n assignment. This should only be implemented if ``del obj.name`` is\n meaningful for the object.\n\n\nMore attribute access for new-style classes\n===========================================\n\nThe following methods only apply to new-style classes.\n\nobject.__getattribute__(self, name)\n\n Called unconditionally to implement attribute accesses for\n instances of the class. If the class also defines\n ``__getattr__()``, the latter will not be called unless\n ``__getattribute__()`` either calls it explicitly or raises an\n ``AttributeError``. This method should return the (computed)\n attribute value or raise an ``AttributeError`` exception. In order\n to avoid infinite recursion in this method, its implementation\n should always call the base class method with the same name to\n access any attributes it needs, for example,\n ``object.__getattribute__(self, name)``.\n\n Note: This method may still be bypassed when looking up special methods\n as the result of implicit invocation via language syntax or\n built-in functions. See *Special method lookup for new-style\n classes*.\n\n\nImplementing Descriptors\n========================\n\nThe following methods only apply when an instance of the class\ncontaining the method (a so-called *descriptor* class) appears in an\n*owner* class (the descriptor must be in either the owner\'s class\ndictionary or in the class dictionary for one of its parents). In the\nexamples below, "the attribute" refers to the attribute whose name is\nthe key of the property in the owner class\' ``__dict__``.\n\nobject.__get__(self, instance, owner)\n\n Called to get the attribute of the owner class (class attribute\n access) or of an instance of that class (instance attribute\n access). *owner* is always the owner class, while *instance* is the\n instance that the attribute was accessed through, or ``None`` when\n the attribute is accessed through the *owner*. This method should\n return the (computed) attribute value or raise an\n ``AttributeError`` exception.\n\nobject.__set__(self, instance, value)\n\n Called to set the attribute on an instance *instance* of the owner\n class to a new value, *value*.\n\nobject.__delete__(self, instance)\n\n Called to delete the attribute on an instance *instance* of the\n owner class.\n\n\nInvoking Descriptors\n====================\n\nIn general, a descriptor is an object attribute with "binding\nbehavior", one whose attribute access has been overridden by methods\nin the descriptor protocol: ``__get__()``, ``__set__()``, and\n``__delete__()``. If any of those methods are defined for an object,\nit is said to be a descriptor.\n\nThe default behavior for attribute access is to get, set, or delete\nthe attribute from an object\'s dictionary. For instance, ``a.x`` has a\nlookup chain starting with ``a.__dict__[\'x\']``, then\n``type(a).__dict__[\'x\']``, and continuing through the base classes of\n``type(a)`` excluding metaclasses.\n\nHowever, if the looked-up value is an object defining one of the\ndescriptor methods, then Python may override the default behavior and\ninvoke the descriptor method instead. Where this occurs in the\nprecedence chain depends on which descriptor methods were defined and\nhow they were called. Note that descriptors are only invoked for new\nstyle objects or classes (ones that subclass ``object()`` or\n``type()``).\n\nThe starting point for descriptor invocation is a binding, ``a.x``.\nHow the arguments are assembled depends on ``a``:\n\nDirect Call\n The simplest and least common call is when user code directly\n invokes a descriptor method: ``x.__get__(a)``.\n\nInstance Binding\n If binding to a new-style object instance, ``a.x`` is transformed\n into the call: ``type(a).__dict__[\'x\'].__get__(a, type(a))``.\n\nClass Binding\n If binding to a new-style class, ``A.x`` is transformed into the\n call: ``A.__dict__[\'x\'].__get__(None, A)``.\n\nSuper Binding\n If ``a`` is an instance of ``super``, then the binding ``super(B,\n obj).m()`` searches ``obj.__class__.__mro__`` for the base class\n ``A`` immediately preceding ``B`` and then invokes the descriptor\n with the call: ``A.__dict__[\'m\'].__get__(obj, obj.__class__)``.\n\nFor instance bindings, the precedence of descriptor invocation depends\non the which descriptor methods are defined. A descriptor can define\nany combination of ``__get__()``, ``__set__()`` and ``__delete__()``.\nIf it does not define ``__get__()``, then accessing the attribute will\nreturn the descriptor object itself unless there is a value in the\nobject\'s instance dictionary. If the descriptor defines ``__set__()``\nand/or ``__delete__()``, it is a data descriptor; if it defines\nneither, it is a non-data descriptor. Normally, data descriptors\ndefine both ``__get__()`` and ``__set__()``, while non-data\ndescriptors have just the ``__get__()`` method. Data descriptors with\n``__set__()`` and ``__get__()`` defined always override a redefinition\nin an instance dictionary. In contrast, non-data descriptors can be\noverridden by instances.\n\nPython methods (including ``staticmethod()`` and ``classmethod()``)\nare implemented as non-data descriptors. Accordingly, instances can\nredefine and override methods. This allows individual instances to\nacquire behaviors that differ from other instances of the same class.\n\nThe ``property()`` function is implemented as a data descriptor.\nAccordingly, instances cannot override the behavior of a property.\n\n\n__slots__\n=========\n\nBy default, instances of both old and new-style classes have a\ndictionary for attribute storage. This wastes space for objects\nhaving very few instance variables. The space consumption can become\nacute when creating large numbers of instances.\n\nThe default can be overridden by defining *__slots__* in a new-style\nclass definition. The *__slots__* declaration takes a sequence of\ninstance variables and reserves just enough space in each instance to\nhold a value for each variable. Space is saved because *__dict__* is\nnot created for each instance.\n\n__slots__\n\n This class variable can be assigned a string, iterable, or sequence\n of strings with variable names used by instances. If defined in a\n new-style class, *__slots__* reserves space for the declared\n variables and prevents the automatic creation of *__dict__* and\n *__weakref__* for each instance.\n\n New in version 2.2.\n\nNotes on using *__slots__*\n\n* When inheriting from a class without *__slots__*, the *__dict__*\n attribute of that class will always be accessible, so a *__slots__*\n definition in the subclass is meaningless.\n\n* Without a *__dict__* variable, instances cannot be assigned new\n variables not listed in the *__slots__* definition. Attempts to\n assign to an unlisted variable name raises ``AttributeError``. If\n dynamic assignment of new variables is desired, then add\n ``\'__dict__\'`` to the sequence of strings in the *__slots__*\n declaration.\n\n Changed in version 2.3: Previously, adding ``\'__dict__\'`` to the\n *__slots__* declaration would not enable the assignment of new\n attributes not specifically listed in the sequence of instance\n variable names.\n\n* Without a *__weakref__* variable for each instance, classes defining\n *__slots__* do not support weak references to its instances. If weak\n reference support is needed, then add ``\'__weakref__\'`` to the\n sequence of strings in the *__slots__* declaration.\n\n Changed in version 2.3: Previously, adding ``\'__weakref__\'`` to the\n *__slots__* declaration would not enable support for weak\n references.\n\n* *__slots__* are implemented at the class level by creating\n descriptors (*Implementing Descriptors*) for each variable name. As\n a result, class attributes cannot be used to set default values for\n instance variables defined by *__slots__*; otherwise, the class\n attribute would overwrite the descriptor assignment.\n\n* The action of a *__slots__* declaration is limited to the class\n where it is defined. As a result, subclasses will have a *__dict__*\n unless they also define *__slots__* (which must only contain names\n of any *additional* slots).\n\n* If a class defines a slot also defined in a base class, the instance\n variable defined by the base class slot is inaccessible (except by\n retrieving its descriptor directly from the base class). This\n renders the meaning of the program undefined. In the future, a\n check may be added to prevent this.\n\n* Nonempty *__slots__* does not work for classes derived from\n "variable-length" built-in types such as ``long``, ``str`` and\n ``tuple``.\n\n* Any non-string iterable may be assigned to *__slots__*. Mappings may\n also be used; however, in the future, special meaning may be\n assigned to the values corresponding to each key.\n\n* *__class__* assignment works only if both classes have the same\n *__slots__*.\n\n Changed in version 2.6: Previously, *__class__* assignment raised an\n error if either new or old class had *__slots__*.\n', 18 'calls': '\nCalls\n*****\n\nA call calls a callable object (e.g., a *function*) with a possibly\nempty series of *arguments*:\n\n call ::= primary "(" [argument_list [","]\n | expression genexpr_for] ")"\n argument_list ::= positional_arguments ["," keyword_arguments]\n ["," "*" expression] ["," keyword_arguments]\n ["," "**" expression]\n | keyword_arguments ["," "*" expression]\n ["," "**" expression]\n | "*" expression ["," "*" expression] ["," "**" expression]\n | "**" expression\n positional_arguments ::= expression ("," expression)*\n keyword_arguments ::= keyword_item ("," keyword_item)*\n keyword_item ::= identifier "=" expression\n\nA trailing comma may be present after the positional and keyword\narguments but does not affect the semantics.\n\nThe primary must evaluate to a callable object (user-defined\nfunctions, built-in functions, methods of built-in objects, class\nobjects, methods of class instances, and certain class instances\nthemselves are callable; extensions may define additional callable\nobject types). All argument expressions are evaluated before the call\nis attempted. Please refer to section *Function definitions* for the\nsyntax of formal *parameter* lists.\n\nIf keyword arguments are present, they are first converted to\npositional arguments, as follows. First, a list of unfilled slots is\ncreated for the formal parameters. If there are N positional\narguments, they are placed in the first N slots. Next, for each\nkeyword argument, the identifier is used to determine the\ncorresponding slot (if the identifier is the same as the first formal\nparameter name, the first slot is used, and so on). If the slot is\nalready filled, a ``TypeError`` exception is raised. Otherwise, the\nvalue of the argument is placed in the slot, filling it (even if the\nexpression is ``None``, it fills the slot). When all arguments have\nbeen processed, the slots that are still unfilled are filled with the\ncorresponding default value from the function definition. (Default\nvalues are calculated, once, when the function is defined; thus, a\nmutable object such as a list or dictionary used as default value will\nbe shared by all calls that don\'t specify an argument value for the\ncorresponding slot; this should usually be avoided.) If there are any\nunfilled slots for which no default value is specified, a\n``TypeError`` exception is raised. Otherwise, the list of filled\nslots is used as the argument list for the call.\n\n**CPython implementation detail:** An implementation may provide\nbuilt-in functions whose positional parameters do not have names, even\nif they are \'named\' for the purpose of documentation, and which\ntherefore cannot be supplied by keyword. In CPython, this is the case\nfor functions implemented in C that use ``PyArg_ParseTuple()`` to\nparse their arguments.\n\nIf there are more positional arguments than there are formal parameter\nslots, a ``TypeError`` exception is raised, unless a formal parameter\nusing the syntax ``*identifier`` is present; in this case, that formal\nparameter receives a tuple containing the excess positional arguments\n(or an empty tuple if there were no excess positional arguments).\n\nIf any keyword argument does not correspond to a formal parameter\nname, a ``TypeError`` exception is raised, unless a formal parameter\nusing the syntax ``**identifier`` is present; in this case, that\nformal parameter receives a dictionary containing the excess keyword\narguments (using the keywords as keys and the argument values as\ncorresponding values), or a (new) empty dictionary if there were no\nexcess keyword arguments.\n\nIf the syntax ``*expression`` appears in the function call,\n``expression`` must evaluate to an iterable. Elements from this\niterable are treated as if they were additional positional arguments;\nif there are positional arguments *x1*, ..., *xN*, and ``expression``\nevaluates to a sequence *y1*, ..., *yM*, this is equivalent to a call\nwith M+N positional arguments *x1*, ..., *xN*, *y1*, ..., *yM*.\n\nA consequence of this is that although the ``*expression`` syntax may\nappear *after* some keyword arguments, it is processed *before* the\nkeyword arguments (and the ``**expression`` argument, if any -- see\nbelow). So:\n\n >>> def f(a, b):\n ... print a, b\n ...\n >>> f(b=1, *(2,))\n 2 1\n >>> f(a=1, *(2,))\n Traceback (most recent call last):\n File "<stdin>", line 1, in ?\n TypeError: f() got multiple values for keyword argument \'a\'\n >>> f(1, *(2,))\n 1 2\n\nIt is unusual for both keyword arguments and the ``*expression``\nsyntax to be used in the same call, so in practice this confusion does\nnot arise.\n\nIf the syntax ``**expression`` appears in the function call,\n``expression`` must evaluate to a mapping, the contents of which are\ntreated as additional keyword arguments. In the case of a keyword\nappearing in both ``expression`` and as an explicit keyword argument,\na ``TypeError`` exception is raised.\n\nFormal parameters using the syntax ``*identifier`` or ``**identifier``\ncannot be used as positional argument slots or as keyword argument\nnames. Formal parameters using the syntax ``(sublist)`` cannot be\nused as keyword argument names; the outermost sublist corresponds to a\nsingle unnamed argument slot, and the argument value is assigned to\nthe sublist using the usual tuple assignment rules after all other\nparameter processing is done.\n\nA call always returns some value, possibly ``None``, unless it raises\nan exception. How this value is computed depends on the type of the\ncallable object.\n\nIf it is---\n\na user-defined function:\n The code block for the function is executed, passing it the\n argument list. The first thing the code block will do is bind the\n formal parameters to the arguments; this is described in section\n *Function definitions*. When the code block executes a ``return``\n statement, this specifies the return value of the function call.\n\na built-in function or method:\n The result is up to the interpreter; see *Built-in Functions* for\n the descriptions of built-in functions and methods.\n\na class object:\n A new instance of that class is returned.\n\na class instance method:\n The corresponding user-defined function is called, with an argument\n list that is one longer than the argument list of the call: the\n instance becomes the first argument.\n\na class instance:\n The class must define a ``__call__()`` method; the effect is then\n the same as if that method was called.\n', 61 'specialnames': '\nSpecial method names\n********************\n\nA class can implement certain operations that are invoked by special\nsyntax (such as arithmetic operations or subscripting and slicing) by\ndefining methods with special names. This is Python\'s approach to\n*operator overloading*, allowing classes to define their own behavior\nwith respect to language operators. For instance, if a class defines\na method named ``__getitem__()``, and ``x`` is an instance of this\nclass, then ``x[i]`` is roughly equivalent to ``x.__getitem__(i)`` for\nold-style classes and ``type(x).__getitem__(x, i)`` for new-style\nclasses. Except where mentioned, attempts to execute an operation\nraise an exception when no appropriate method is defined (typically\n``AttributeError`` or ``TypeError``).\n\nWhen implementing a class that emulates any built-in type, it is\nimportant that the emulation only be implemented to the degree that it\nmakes sense for the object being modelled. For example, some\nsequences may work well with retrieval of individual elements, but\nextracting a slice may not make sense. (One example of this is the\n``NodeList`` interface in the W3C\'s Document Object Model.)\n\n\nBasic customization\n===================\n\nobject.__new__(cls[, ...])\n\n Called to create a new instance of class *cls*. ``__new__()`` is a\n static method (special-cased so you need not declare it as such)\n that takes the class of which an instance was requested as its\n first argument. The remaining arguments are those passed to the\n object constructor expression (the call to the class). The return\n value of ``__new__()`` should be the new object instance (usually\n an instance of *cls*).\n\n Typical implementations create a new instance of the class by\n invoking the superclass\'s ``__new__()`` method using\n ``super(currentclass, cls).__new__(cls[, ...])`` with appropriate\n arguments and then modifying the newly-created instance as\n necessary before returning it.\n\n If ``__new__()`` returns an instance of *cls*, then the new\n instance\'s ``__init__()`` method will be invoked like\n ``__init__(self[, ...])``, where *self* is the new instance and the\n remaining arguments are the same as were passed to ``__new__()``.\n\n If ``__new__()`` does not return an instance of *cls*, then the new\n instance\'s ``__init__()`` method will not be invoked.\n\n ``__new__()`` is intended mainly to allow subclasses of immutable\n types (like int, str, or tuple) to customize instance creation. It\n is also commonly overridden in custom metaclasses in order to\n customize class creation.\n\nobject.__init__(self[, ...])\n\n Called when the instance is created. The arguments are those\n passed to the class constructor expression. If a base class has an\n ``__init__()`` method, the derived class\'s ``__init__()`` method,\n if any, must explicitly call it to ensure proper initialization of\n the base class part of the instance; for example:\n ``BaseClass.__init__(self, [args...])``. As a special constraint\n on constructors, no value may be returned; doing so will cause a\n ``TypeError`` to be raised at runtime.\n\nobject.__del__(self)\n\n Called when the instance is about to be destroyed. This is also\n called a destructor. If a base class has a ``__del__()`` method,\n the derived class\'s ``__del__()`` method, if any, must explicitly\n call it to ensure proper deletion of the base class part of the\n instance. Note that it is possible (though not recommended!) for\n the ``__del__()`` method to postpone destruction of the instance by\n creating a new reference to it. It may then be called at a later\n time when this new reference is deleted. It is not guaranteed that\n ``__del__()`` methods are called for objects that still exist when\n the interpreter exits.\n\n Note: ``del x`` doesn\'t directly call ``x.__del__()`` --- the former\n decrements the reference count for ``x`` by one, and the latter\n is only called when ``x``\'s reference count reaches zero. Some\n common situations that may prevent the reference count of an\n object from going to zero include: circular references between\n objects (e.g., a doubly-linked list or a tree data structure with\n parent and child pointers); a reference to the object on the\n stack frame of a function that caught an exception (the traceback\n stored in ``sys.exc_traceback`` keeps the stack frame alive); or\n a reference to the object on the stack frame that raised an\n unhandled exception in interactive mode (the traceback stored in\n ``sys.last_traceback`` keeps the stack frame alive). The first\n situation can only be remedied by explicitly breaking the cycles;\n the latter two situations can be resolved by storing ``None`` in\n ``sys.exc_traceback`` or ``sys.last_traceback``. Circular\n references which are garbage are detected when the option cycle\n detector is enabled (it\'s on by default), but can only be cleaned\n up if there are no Python-level ``__del__()`` methods involved.\n Refer to the documentation for the ``gc`` module for more\n information about how ``__del__()`` methods are handled by the\n cycle detector, particularly the description of the ``garbage``\n value.\n\n Warning: Due to the precarious circumstances under which ``__del__()``\n methods are invoked, exceptions that occur during their execution\n are ignored, and a warning is printed to ``sys.stderr`` instead.\n Also, when ``__del__()`` is invoked in response to a module being\n deleted (e.g., when execution of the program is done), other\n globals referenced by the ``__del__()`` method may already have\n been deleted or in the process of being torn down (e.g. the\n import machinery shutting down). For this reason, ``__del__()``\n methods should do the absolute minimum needed to maintain\n external invariants. Starting with version 1.5, Python\n guarantees that globals whose name begins with a single\n underscore are deleted from their module before other globals are\n deleted; if no other references to such globals exist, this may\n help in assuring that imported modules are still available at the\n time when the ``__del__()`` method is called.\n\n See also the *-R* command-line option.\n\nobject.__repr__(self)\n\n Called by the ``repr()`` built-in function and by string\n conversions (reverse quotes) to compute the "official" string\n representation of an object. If at all possible, this should look\n like a valid Python expression that could be used to recreate an\n object with the same value (given an appropriate environment). If\n this is not possible, a string of the form ``<...some useful\n description...>`` should be returned. The return value must be a\n string object. If a class defines ``__repr__()`` but not\n ``__str__()``, then ``__repr__()`` is also used when an "informal"\n string representation of instances of that class is required.\n\n This is typically used for debugging, so it is important that the\n representation is information-rich and unambiguous.\n\nobject.__str__(self)\n\n Called by the ``str()`` built-in function and by the ``print``\n statement to compute the "informal" string representation of an\n object. This differs from ``__repr__()`` in that it does not have\n to be a valid Python expression: a more convenient or concise\n representation may be used instead. The return value must be a\n string object.\n\nobject.__lt__(self, other)\nobject.__le__(self, other)\nobject.__eq__(self, other)\nobject.__ne__(self, other)\nobject.__gt__(self, other)\nobject.__ge__(self, other)\n\n New in version 2.1.\n\n These are the so-called "rich comparison" methods, and are called\n for comparison operators in preference to ``__cmp__()`` below. The\n correspondence between operator symbols and method names is as\n follows: ``x<y`` calls ``x.__lt__(y)``, ``x<=y`` calls\n ``x.__le__(y)``, ``x==y`` calls ``x.__eq__(y)``, ``x!=y`` and\n ``x<>y`` call ``x.__ne__(y)``, ``x>y`` calls ``x.__gt__(y)``, and\n ``x>=y`` calls ``x.__ge__(y)``.\n\n A rich comparison method may return the singleton\n ``NotImplemented`` if it does not implement the operation for a\n given pair of arguments. By convention, ``False`` and ``True`` are\n returned for a successful comparison. However, these methods can\n return any value, so if the comparison operator is used in a\n Boolean context (e.g., in the condition of an ``if`` statement),\n Python will call ``bool()`` on the value to determine if the result\n is true or false.\n\n There are no implied relationships among the comparison operators.\n The truth of ``x==y`` does not imply that ``x!=y`` is false.\n Accordingly, when defining ``__eq__()``, one should also define\n ``__ne__()`` so that the operators will behave as expected. See\n the paragraph on ``__hash__()`` for some important notes on\n creating *hashable* objects which support custom comparison\n operations and are usable as dictionary keys.\n\n There are no swapped-argument versions of these methods (to be used\n when the left argument does not support the operation but the right\n argument does); rather, ``__lt__()`` and ``__gt__()`` are each\n other\'s reflection, ``__le__()`` and ``__ge__()`` are each other\'s\n reflection, and ``__eq__()`` and ``__ne__()`` are their own\n reflection.\n\n Arguments to rich comparison methods are never coerced.\n\n To automatically generate ordering operations from a single root\n operation, see ``functools.total_ordering()``.\n\nobject.__cmp__(self, other)\n\n Called by comparison operations if rich comparison (see above) is\n not defined. Should return a negative integer if ``self < other``,\n zero if ``self == other``, a positive integer if ``self > other``.\n If no ``__cmp__()``, ``__eq__()`` or ``__ne__()`` operation is\n defined, class instances are compared by object identity\n ("address"). See also the description of ``__hash__()`` for some\n important notes on creating *hashable* objects which support custom\n comparison operations and are usable as dictionary keys. (Note: the\n restriction that exceptions are not propagated by ``__cmp__()`` has\n been removed since Python 1.5.)\n\nobject.__rcmp__(self, other)\n\n Changed in version 2.1: No longer supported.\n\nobject.__hash__(self)\n\n Called by built-in function ``hash()`` and for operations on\n members of hashed collections including ``set``, ``frozenset``, and\n ``dict``. ``__hash__()`` should return an integer. The only\n required property is that objects which compare equal have the same\n hash value; it is advised to somehow mix together (e.g. using\n exclusive or) the hash values for the components of the object that\n also play a part in comparison of objects.\n\n If a class does not define a ``__cmp__()`` or ``__eq__()`` method\n it should not define a ``__hash__()`` operation either; if it\n defines ``__cmp__()`` or ``__eq__()`` but not ``__hash__()``, its\n instances will not be usable in hashed collections. If a class\n defines mutable objects and implements a ``__cmp__()`` or\n ``__eq__()`` method, it should not implement ``__hash__()``, since\n hashable collection implementations require that a object\'s hash\n value is immutable (if the object\'s hash value changes, it will be\n in the wrong hash bucket).\n\n User-defined classes have ``__cmp__()`` and ``__hash__()`` methods\n by default; with them, all objects compare unequal (except with\n themselves) and ``x.__hash__()`` returns ``id(x)``.\n\n Classes which inherit a ``__hash__()`` method from a parent class\n but change the meaning of ``__cmp__()`` or ``__eq__()`` such that\n the hash value returned is no longer appropriate (e.g. by switching\n to a value-based concept of equality instead of the default\n identity based equality) can explicitly flag themselves as being\n unhashable by setting ``__hash__ = None`` in the class definition.\n Doing so means that not only will instances of the class raise an\n appropriate ``TypeError`` when a program attempts to retrieve their\n hash value, but they will also be correctly identified as\n unhashable when checking ``isinstance(obj, collections.Hashable)``\n (unlike classes which define their own ``__hash__()`` to explicitly\n raise ``TypeError``).\n\n Changed in version 2.5: ``__hash__()`` may now also return a long\n integer object; the 32-bit integer is then derived from the hash of\n that object.\n\n Changed in version 2.6: ``__hash__`` may now be set to ``None`` to\n explicitly flag instances of a class as unhashable.\n\nobject.__nonzero__(self)\n\n Called to implement truth value testing and the built-in operation\n ``bool()``; should return ``False`` or ``True``, or their integer\n equivalents ``0`` or ``1``. When this method is not defined,\n ``__len__()`` is called, if it is defined, and the object is\n considered true if its result is nonzero. If a class defines\n neither ``__len__()`` nor ``__nonzero__()``, all its instances are\n considered true.\n\nobject.__unicode__(self)\n\n Called to implement ``unicode()`` built-in; should return a Unicode\n object. When this method is not defined, string conversion is\n attempted, and the result of string conversion is converted to\n Unicode using the system default encoding.\n\n\nCustomizing attribute access\n============================\n\nThe following methods can be defined to customize the meaning of\nattribute access (use of, assignment to, or deletion of ``x.name``)\nfor class instances.\n\nobject.__getattr__(self, name)\n\n Called when an attribute lookup has not found the attribute in the\n usual places (i.e. it is not an instance attribute nor is it found\n in the class tree for ``self``). ``name`` is the attribute name.\n This method should return the (computed) attribute value or raise\n an ``AttributeError`` exception.\n\n Note that if the attribute is found through the normal mechanism,\n ``__getattr__()`` is not called. (This is an intentional asymmetry\n between ``__getattr__()`` and ``__setattr__()``.) This is done both\n for efficiency reasons and because otherwise ``__getattr__()``\n would have no way to access other attributes of the instance. Note\n that at least for instance variables, you can fake total control by\n not inserting any values in the instance attribute dictionary (but\n instead inserting them in another object). See the\n ``__getattribute__()`` method below for a way to actually get total\n control in new-style classes.\n\nobject.__setattr__(self, name, value)\n\n Called when an attribute assignment is attempted. This is called\n instead of the normal mechanism (i.e. store the value in the\n instance dictionary). *name* is the attribute name, *value* is the\n value to be assigned to it.\n\n If ``__setattr__()`` wants to assign to an instance attribute, it\n should not simply execute ``self.name = value`` --- this would\n cause a recursive call to itself. Instead, it should insert the\n value in the dictionary of instance attributes, e.g.,\n ``self.__dict__[name] = value``. For new-style classes, rather\n than accessing the instance dictionary, it should call the base\n class method with the same name, for example,\n ``object.__setattr__(self, name, value)``.\n\nobject.__delattr__(self, name)\n\n Like ``__setattr__()`` but for attribute deletion instead of\n assignment. This should only be implemented if ``del obj.name`` is\n meaningful for the object.\n\n\nMore attribute access for new-style classes\n-------------------------------------------\n\nThe following methods only apply to new-style classes.\n\nobject.__getattribute__(self, name)\n\n Called unconditionally to implement attribute accesses for\n instances of the class. If the class also defines\n ``__getattr__()``, the latter will not be called unless\n ``__getattribute__()`` either calls it explicitly or raises an\n ``AttributeError``. This method should return the (computed)\n attribute value or raise an ``AttributeError`` exception. In order\n to avoid infinite recursion in this method, its implementation\n should always call the base class method with the same name to\n access any attributes it needs, for example,\n ``object.__getattribute__(self, name)``.\n\n Note: This method may still be bypassed when looking up special methods\n as the result of implicit invocation via language syntax or\n built-in functions. See *Special method lookup for new-style\n classes*.\n\n\nImplementing Descriptors\n------------------------\n\nThe following methods only apply when an instance of the class\ncontaining the method (a so-called *descriptor* class) appears in an\n*owner* class (the descriptor must be in either the owner\'s class\ndictionary or in the class dictionary for one of its parents). In the\nexamples below, "the attribute" refers to the attribute whose name is\nthe key of the property in the owner class\' ``__dict__``.\n\nobject.__get__(self, instance, owner)\n\n Called to get the attribute of the owner class (class attribute\n access) or of an instance of that class (instance attribute\n access). *owner* is always the owner class, while *instance* is the\n instance that the attribute was accessed through, or ``None`` when\n the attribute is accessed through the *owner*. This method should\n return the (computed) attribute value or raise an\n ``AttributeError`` exception.\n\nobject.__set__(self, instance, value)\n\n Called to set the attribute on an instance *instance* of the owner\n class to a new value, *value*.\n\nobject.__delete__(self, instance)\n\n Called to delete the attribute on an instance *instance* of the\n owner class.\n\n\nInvoking Descriptors\n--------------------\n\nIn general, a descriptor is an object attribute with "binding\nbehavior", one whose attribute access has been overridden by methods\nin the descriptor protocol: ``__get__()``, ``__set__()``, and\n``__delete__()``. If any of those methods are defined for an object,\nit is said to be a descriptor.\n\nThe default behavior for attribute access is to get, set, or delete\nthe attribute from an object\'s dictionary. For instance, ``a.x`` has a\nlookup chain starting with ``a.__dict__[\'x\']``, then\n``type(a).__dict__[\'x\']``, and continuing through the base classes of\n``type(a)`` excluding metaclasses.\n\nHowever, if the looked-up value is an object defining one of the\ndescriptor methods, then Python may override the default behavior and\ninvoke the descriptor method instead. Where this occurs in the\nprecedence chain depends on which descriptor methods were defined and\nhow they were called. Note that descriptors are only invoked for new\nstyle objects or classes (ones that subclass ``object()`` or\n``type()``).\n\nThe starting point for descriptor invocation is a binding, ``a.x``.\nHow the arguments are assembled depends on ``a``:\n\nDirect Call\n The simplest and least common call is when user code directly\n invokes a descriptor method: ``x.__get__(a)``.\n\nInstance Binding\n If binding to a new-style object instance, ``a.x`` is transformed\n into the call: ``type(a).__dict__[\'x\'].__get__(a, type(a))``.\n\nClass Binding\n If binding to a new-style class, ``A.x`` is transformed into the\n call: ``A.__dict__[\'x\'].__get__(None, A)``.\n\nSuper Binding\n If ``a`` is an instance of ``super``, then the binding ``super(B,\n obj).m()`` searches ``obj.__class__.__mro__`` for the base class\n ``A`` immediately preceding ``B`` and then invokes the descriptor\n with the call: ``A.__dict__[\'m\'].__get__(obj, obj.__class__)``.\n\nFor instance bindings, the precedence of descriptor invocation depends\non the which descriptor methods are defined. A descriptor can define\nany combination of ``__get__()``, ``__set__()`` and ``__delete__()``.\nIf it does not define ``__get__()``, then accessing the attribute will\nreturn the descriptor object itself unless there is a value in the\nobject\'s instance dictionary. If the descriptor defines ``__set__()``\nand/or ``__delete__()``, it is a data descriptor; if it defines\nneither, it is a non-data descriptor. Normally, data descriptors\ndefine both ``__get__()`` and ``__set__()``, while non-data\ndescriptors have just the ``__get__()`` method. Data descriptors with\n``__set__()`` and ``__get__()`` defined always override a redefinition\nin an instance dictionary. In contrast, non-data descriptors can be\noverridden by instances.\n\nPython methods (including ``staticmethod()`` and ``classmethod()``)\nare implemented as non-data descriptors. Accordingly, instances can\nredefine and override methods. This allows individual instances to\nacquire behaviors that differ from other instances of the same class.\n\nThe ``property()`` function is implemented as a data descriptor.\nAccordingly, instances cannot override the behavior of a property.\n\n\n__slots__\n---------\n\nBy default, instances of both old and new-style classes have a\ndictionary for attribute storage. This wastes space for objects\nhaving very few instance variables. The space consumption can become\nacute when creating large numbers of instances.\n\nThe default can be overridden by defining *__slots__* in a new-style\nclass definition. The *__slots__* declaration takes a sequence of\ninstance variables and reserves just enough space in each instance to\nhold a value for each variable. Space is saved because *__dict__* is\nnot created for each instance.\n\n__slots__\n\n This class variable can be assigned a string, iterable, or sequence\n of strings with variable names used by instances. If defined in a\n new-style class, *__slots__* reserves space for the declared\n variables and prevents the automatic creation of *__dict__* and\n *__weakref__* for each instance.\n\n New in version 2.2.\n\nNotes on using *__slots__*\n\n* When inheriting from a class without *__slots__*, the *__dict__*\n attribute of that class will always be accessible, so a *__slots__*\n definition in the subclass is meaningless.\n\n* Without a *__dict__* variable, instances cannot be assigned new\n variables not listed in the *__slots__* definition. Attempts to\n assign to an unlisted variable name raises ``AttributeError``. If\n dynamic assignment of new variables is desired, then add\n ``\'__dict__\'`` to the sequence of strings in the *__slots__*\n declaration.\n\n Changed in version 2.3: Previously, adding ``\'__dict__\'`` to the\n *__slots__* declaration would not enable the assignment of new\n attributes not specifically listed in the sequence of instance\n variable names.\n\n* Without a *__weakref__* variable for each instance, classes defining\n *__slots__* do not support weak references to its instances. If weak\n reference support is needed, then add ``\'__weakref__\'`` to the\n sequence of strings in the *__slots__* declaration.\n\n Changed in version 2.3: Previously, adding ``\'__weakref__\'`` to the\n *__slots__* declaration would not enable support for weak\n references.\n\n* *__slots__* are implemented at the class level by creating\n descriptors (*Implementing Descriptors*) for each variable name. As\n a result, class attributes cannot be used to set default values for\n instance variables defined by *__slots__*; otherwise, the class\n attribute would overwrite the descriptor assignment.\n\n* The action of a *__slots__* declaration is limited to the class\n where it is defined. As a result, subclasses will have a *__dict__*\n unless they also define *__slots__* (which must only contain names\n of any *additional* slots).\n\n* If a class defines a slot also defined in a base class, the instance\n variable defined by the base class slot is inaccessible (except by\n retrieving its descriptor directly from the base class). This\n renders the meaning of the program undefined. In the future, a\n check may be added to prevent this.\n\n* Nonempty *__slots__* does not work for classes derived from\n "variable-length" built-in types such as ``long``, ``str`` and\n ``tuple``.\n\n* Any non-string iterable may be assigned to *__slots__*. Mappings may\n also be used; however, in the future, special meaning may be\n assigned to the values corresponding to each key.\n\n* *__class__* assignment works only if both classes have the same\n *__slots__*.\n\n Changed in version 2.6: Previously, *__class__* assignment raised an\n error if either new or old class had *__slots__*.\n\n\nCustomizing class creation\n==========================\n\nBy default, new-style classes are constructed using ``type()``. A\nclass definition is read into a separate namespace and the value of\nclass name is bound to the result of ``type(name, bases, dict)``.\n\nWhen the class definition is read, if *__metaclass__* is defined then\nthe callable assigned to it will be called instead of ``type()``. This\nallows classes or functions to be written which monitor or alter the\nclass creation process:\n\n* Modifying the class dictionary prior to the class being created.\n\n* Returning an instance of another class -- essentially performing the\n role of a factory function.\n\nThese steps will have to be performed in the metaclass\'s ``__new__()``\nmethod -- ``type.__new__()`` can then be called from this method to\ncreate a class with different properties. This example adds a new\nelement to the class dictionary before creating the class:\n\n class metacls(type):\n def __new__(mcs, name, bases, dict):\n dict[\'foo\'] = \'metacls was here\'\n return type.__new__(mcs, name, bases, dict)\n\nYou can of course also override other class methods (or add new\nmethods); for example defining a custom ``__call__()`` method in the\nmetaclass allows custom behavior when the class is called, e.g. not\nalways creating a new instance.\n\n__metaclass__\n\n This variable can be any callable accepting arguments for ``name``,\n ``bases``, and ``dict``. Upon class creation, the callable is used\n instead of the built-in ``type()``.\n\n New in version 2.2.\n\nThe appropriate metaclass is determined by the following precedence\nrules:\n\n* If ``dict[\'__metaclass__\']`` exists, it is used.\n\n* Otherwise, if there is at least one base class, its metaclass is\n used (this looks for a *__class__* attribute first and if not found,\n uses its type).\n\n* Otherwise, if a global variable named __metaclass__ exists, it is\n used.\n\n* Otherwise, the old-style, classic metaclass (types.ClassType) is\n used.\n\nThe potential uses for metaclasses are boundless. Some ideas that have\nbeen explored including logging, interface checking, automatic\ndelegation, automatic property creation, proxies, frameworks, and\nautomatic resource locking/synchronization.\n\n\nCustomizing instance and subclass checks\n========================================\n\nNew in version 2.6.\n\nThe following methods are used to override the default behavior of the\n``isinstance()`` and ``issubclass()`` built-in functions.\n\nIn particular, the metaclass ``abc.ABCMeta`` implements these methods\nin order to allow the addition of Abstract Base Classes (ABCs) as\n"virtual base classes" to any class or type (including built-in\ntypes), including other ABCs.\n\nclass.__instancecheck__(self, instance)\n\n Return true if *instance* should be considered a (direct or\n indirect) instance of *class*. If defined, called to implement\n ``isinstance(instance, class)``.\n\nclass.__subclasscheck__(self, subclass)\n\n Return true if *subclass* should be considered a (direct or\n indirect) subclass of *class*. If defined, called to implement\n ``issubclass(subclass, class)``.\n\nNote that these methods are looked up on the type (metaclass) of a\nclass. They cannot be defined as class methods in the actual class.\nThis is consistent with the lookup of special methods that are called\non instances, only in this case the instance is itself a class.\n\nSee also:\n\n **PEP 3119** - Introducing Abstract Base Classes\n Includes the specification for customizing ``isinstance()`` and\n ``issubclass()`` behavior through ``__instancecheck__()`` and\n ``__subclasscheck__()``, with motivation for this functionality\n in the context of adding Abstract Base Classes (see the ``abc``\n module) to the language.\n\n\nEmulating callable objects\n==========================\n\nobject.__call__(self[, args...])\n\n Called when the instance is "called" as a function; if this method\n is defined, ``x(arg1, arg2, ...)`` is a shorthand for\n ``x.__call__(arg1, arg2, ...)``.\n\n\nEmulating container types\n=========================\n\nThe following methods can be defined to implement container objects.\nContainers usually are sequences (such as lists or tuples) or mappings\n(like dictionaries), but can represent other containers as well. The\nfirst set of methods is used either to emulate a sequence or to\nemulate a mapping; the difference is that for a sequence, the\nallowable keys should be the integers *k* for which ``0 <= k < N``\nwhere *N* is the length of the sequence, or slice objects, which\ndefine a range of items. (For backwards compatibility, the method\n``__getslice__()`` (see below) can also be defined to handle simple,\nbut not extended slices.) It is also recommended that mappings provide\nthe methods ``keys()``, ``values()``, ``items()``, ``has_key()``,\n``get()``, ``clear()``, ``setdefault()``, ``iterkeys()``,\n``itervalues()``, ``iteritems()``, ``pop()``, ``popitem()``,\n``copy()``, and ``update()`` behaving similar to those for Python\'s\nstandard dictionary objects. The ``UserDict`` module provides a\n``DictMixin`` class to help create those methods from a base set of\n``__getitem__()``, ``__setitem__()``, ``__delitem__()``, and\n``keys()``. Mutable sequences should provide methods ``append()``,\n``count()``, ``index()``, ``extend()``, ``insert()``, ``pop()``,\n``remove()``, ``reverse()`` and ``sort()``, like Python standard list\nobjects. Finally, sequence types should implement addition (meaning\nconcatenation) and multiplication (meaning repetition) by defining the\nmethods ``__add__()``, ``__radd__()``, ``__iadd__()``, ``__mul__()``,\n``__rmul__()`` and ``__imul__()`` described below; they should not\ndefine ``__coerce__()`` or other numerical operators. It is\nrecommended that both mappings and sequences implement the\n``__contains__()`` method to allow efficient use of the ``in``\noperator; for mappings, ``in`` should be equivalent of ``has_key()``;\nfor sequences, it should search through the values. It is further\nrecommended that both mappings and sequences implement the\n``__iter__()`` method to allow efficient iteration through the\ncontainer; for mappings, ``__iter__()`` should be the same as\n``iterkeys()``; for sequences, it should iterate through the values.\n\nobject.__len__(self)\n\n Called to implement the built-in function ``len()``. Should return\n the length of the object, an integer ``>=`` 0. Also, an object\n that doesn\'t define a ``__nonzero__()`` method and whose\n ``__len__()`` method returns zero is considered to be false in a\n Boolean context.\n\nobject.__getitem__(self, key)\n\n Called to implement evaluation of ``self[key]``. For sequence\n types, the accepted keys should be integers and slice objects.\n Note that the special interpretation of negative indexes (if the\n class wishes to emulate a sequence type) is up to the\n ``__getitem__()`` method. If *key* is of an inappropriate type,\n ``TypeError`` may be raised; if of a value outside the set of\n indexes for the sequence (after any special interpretation of\n negative values), ``IndexError`` should be raised. For mapping\n types, if *key* is missing (not in the container), ``KeyError``\n should be raised.\n\n Note: ``for`` loops expect that an ``IndexError`` will be raised for\n illegal indexes to allow proper detection of the end of the\n sequence.\n\nobject.__setitem__(self, key, value)\n\n Called to implement assignment to ``self[key]``. Same note as for\n ``__getitem__()``. This should only be implemented for mappings if\n the objects support changes to the values for keys, or if new keys\n can be added, or for sequences if elements can be replaced. The\n same exceptions should be raised for improper *key* values as for\n the ``__getitem__()`` method.\n\nobject.__delitem__(self, key)\n\n Called to implement deletion of ``self[key]``. Same note as for\n ``__getitem__()``. This should only be implemented for mappings if\n the objects support removal of keys, or for sequences if elements\n can be removed from the sequence. The same exceptions should be\n raised for improper *key* values as for the ``__getitem__()``\n method.\n\nobject.__iter__(self)\n\n This method is called when an iterator is required for a container.\n This method should return a new iterator object that can iterate\n over all the objects in the container. For mappings, it should\n iterate over the keys of the container, and should also be made\n available as the method ``iterkeys()``.\n\n Iterator objects also need to implement this method; they are\n required to return themselves. For more information on iterator\n objects, see *Iterator Types*.\n\nobject.__reversed__(self)\n\n Called (if present) by the ``reversed()`` built-in to implement\n reverse iteration. It should return a new iterator object that\n iterates over all the objects in the container in reverse order.\n\n If the ``__reversed__()`` method is not provided, the\n ``reversed()`` built-in will fall back to using the sequence\n protocol (``__len__()`` and ``__getitem__()``). Objects that\n support the sequence protocol should only provide\n ``__reversed__()`` if they can provide an implementation that is\n more efficient than the one provided by ``reversed()``.\n\n New in version 2.6.\n\nThe membership test operators (``in`` and ``not in``) are normally\nimplemented as an iteration through a sequence. However, container\nobjects can supply the following special method with a more efficient\nimplementation, which also does not require the object be a sequence.\n\nobject.__contains__(self, item)\n\n Called to implement membership test operators. Should return true\n if *item* is in *self*, false otherwise. For mapping objects, this\n should consider the keys of the mapping rather than the values or\n the key-item pairs.\n\n For objects that don\'t define ``__contains__()``, the membership\n test first tries iteration via ``__iter__()``, then the old\n sequence iteration protocol via ``__getitem__()``, see *this\n section in the language reference*.\n\n\nAdditional methods for emulation of sequence types\n==================================================\n\nThe following optional methods can be defined to further emulate\nsequence objects. Immutable sequences methods should at most only\ndefine ``__getslice__()``; mutable sequences might define all three\nmethods.\n\nobject.__getslice__(self, i, j)\n\n Deprecated since version 2.0: Support slice objects as parameters\n to the ``__getitem__()`` method. (However, built-in types in\n CPython currently still implement ``__getslice__()``. Therefore,\n you have to override it in derived classes when implementing\n slicing.)\n\n Called to implement evaluation of ``self[i:j]``. The returned\n object should be of the same type as *self*. Note that missing *i*\n or *j* in the slice expression are replaced by zero or\n ``sys.maxint``, respectively. If negative indexes are used in the\n slice, the length of the sequence is added to that index. If the\n instance does not implement the ``__len__()`` method, an\n ``AttributeError`` is raised. No guarantee is made that indexes\n adjusted this way are not still negative. Indexes which are\n greater than the length of the sequence are not modified. If no\n ``__getslice__()`` is found, a slice object is created instead, and\n passed to ``__getitem__()`` instead.\n\nobject.__setslice__(self, i, j, sequence)\n\n Called to implement assignment to ``self[i:j]``. Same notes for *i*\n and *j* as for ``__getslice__()``.\n\n This method is deprecated. If no ``__setslice__()`` is found, or\n for extended slicing of the form ``self[i:j:k]``, a slice object is\n created, and passed to ``__setitem__()``, instead of\n ``__setslice__()`` being called.\n\nobject.__delslice__(self, i, j)\n\n Called to implement deletion of ``self[i:j]``. Same notes for *i*\n and *j* as for ``__getslice__()``. This method is deprecated. If no\n ``__delslice__()`` is found, or for extended slicing of the form\n ``self[i:j:k]``, a slice object is created, and passed to\n ``__delitem__()``, instead of ``__delslice__()`` being called.\n\nNotice that these methods are only invoked when a single slice with a\nsingle colon is used, and the slice method is available. For slice\noperations involving extended slice notation, or in absence of the\nslice methods, ``__getitem__()``, ``__setitem__()`` or\n``__delitem__()`` is called with a slice object as argument.\n\nThe following example demonstrate how to make your program or module\ncompatible with earlier versions of Python (assuming that methods\n``__getitem__()``, ``__setitem__()`` and ``__delitem__()`` support\nslice objects as arguments):\n\n class MyClass:\n ...\n def __getitem__(self, index):\n ...\n def __setitem__(self, index, value):\n ...\n def __delitem__(self, index):\n ...\n\n if sys.version_info < (2, 0):\n # They won\'t be defined if version is at least 2.0 final\n\n def __getslice__(self, i, j):\n return self[max(0, i):max(0, j):]\n def __setslice__(self, i, j, seq):\n self[max(0, i):max(0, j):] = seq\n def __delslice__(self, i, j):\n del self[max(0, i):max(0, j):]\n ...\n\nNote the calls to ``max()``; these are necessary because of the\nhandling of negative indices before the ``__*slice__()`` methods are\ncalled. When negative indexes are used, the ``__*item__()`` methods\nreceive them as provided, but the ``__*slice__()`` methods get a\n"cooked" form of the index values. For each negative index value, the\nlength of the sequence is added to the index before calling the method\n(which may still result in a negative index); this is the customary\nhandling of negative indexes by the built-in sequence types, and the\n``__*item__()`` methods are expected to do this as well. However,\nsince they should already be doing that, negative indexes cannot be\npassed in; they must be constrained to the bounds of the sequence\nbefore being passed to the ``__*item__()`` methods. Calling ``max(0,\ni)`` conveniently returns the proper value.\n\n\nEmulating numeric types\n=======================\n\nThe following methods can be defined to emulate numeric objects.\nMethods corresponding to operations that are not supported by the\nparticular kind of number implemented (e.g., bitwise operations for\nnon-integral numbers) should be left undefined.\n\nobject.__add__(self, other)\nobject.__sub__(self, other)\nobject.__mul__(self, other)\nobject.__floordiv__(self, other)\nobject.__mod__(self, other)\nobject.__divmod__(self, other)\nobject.__pow__(self, other[, modulo])\nobject.__lshift__(self, other)\nobject.__rshift__(self, other)\nobject.__and__(self, other)\nobject.__xor__(self, other)\nobject.__or__(self, other)\n\n These methods are called to implement the binary arithmetic\n operations (``+``, ``-``, ``*``, ``//``, ``%``, ``divmod()``,\n ``pow()``, ``**``, ``<<``, ``>>``, ``&``, ``^``, ``|``). For\n instance, to evaluate the expression ``x + y``, where *x* is an\n instance of a class that has an ``__add__()`` method,\n ``x.__add__(y)`` is called. The ``__divmod__()`` method should be\n the equivalent to using ``__floordiv__()`` and ``__mod__()``; it\n should not be related to ``__truediv__()`` (described below). Note\n that ``__pow__()`` should be defined to accept an optional third\n argument if the ternary version of the built-in ``pow()`` function\n is to be supported.\n\n If one of those methods does not support the operation with the\n supplied arguments, it should return ``NotImplemented``.\n\nobject.__div__(self, other)\nobject.__truediv__(self, other)\n\n The division operator (``/``) is implemented by these methods. The\n ``__truediv__()`` method is used when ``__future__.division`` is in\n effect, otherwise ``__div__()`` is used. If only one of these two\n methods is defined, the object will not support division in the\n alternate context; ``TypeError`` will be raised instead.\n\nobject.__radd__(self, other)\nobject.__rsub__(self, other)\nobject.__rmul__(self, other)\nobject.__rdiv__(self, other)\nobject.__rtruediv__(self, other)\nobject.__rfloordiv__(self, other)\nobject.__rmod__(self, other)\nobject.__rdivmod__(self, other)\nobject.__rpow__(self, other)\nobject.__rlshift__(self, other)\nobject.__rrshift__(self, other)\nobject.__rand__(self, other)\nobject.__rxor__(self, other)\nobject.__ror__(self, other)\n\n These methods are called to implement the binary arithmetic\n operations (``+``, ``-``, ``*``, ``/``, ``%``, ``divmod()``,\n ``pow()``, ``**``, ``<<``, ``>>``, ``&``, ``^``, ``|``) with\n reflected (swapped) operands. These functions are only called if\n the left operand does not support the corresponding operation and\n the operands are of different types. [2] For instance, to evaluate\n the expression ``x - y``, where *y* is an instance of a class that\n has an ``__rsub__()`` method, ``y.__rsub__(x)`` is called if\n ``x.__sub__(y)`` returns *NotImplemented*.\n\n Note that ternary ``pow()`` will not try calling ``__rpow__()``\n (the coercion rules would become too complicated).\n\n Note: If the right operand\'s type is a subclass of the left operand\'s\n type and that subclass provides the reflected method for the\n operation, this method will be called before the left operand\'s\n non-reflected method. This behavior allows subclasses to\n override their ancestors\' operations.\n\nobject.__iadd__(self, other)\nobject.__isub__(self, other)\nobject.__imul__(self, other)\nobject.__idiv__(self, other)\nobject.__itruediv__(self, other)\nobject.__ifloordiv__(self, other)\nobject.__imod__(self, other)\nobject.__ipow__(self, other[, modulo])\nobject.__ilshift__(self, other)\nobject.__irshift__(self, other)\nobject.__iand__(self, other)\nobject.__ixor__(self, other)\nobject.__ior__(self, other)\n\n These methods are called to implement the augmented arithmetic\n assignments (``+=``, ``-=``, ``*=``, ``/=``, ``//=``, ``%=``,\n ``**=``, ``<<=``, ``>>=``, ``&=``, ``^=``, ``|=``). These methods\n should attempt to do the operation in-place (modifying *self*) and\n return the result (which could be, but does not have to be,\n *self*). If a specific method is not defined, the augmented\n assignment falls back to the normal methods. For instance, to\n execute the statement ``x += y``, where *x* is an instance of a\n class that has an ``__iadd__()`` method, ``x.__iadd__(y)`` is\n called. If *x* is an instance of a class that does not define a\n ``__iadd__()`` method, ``x.__add__(y)`` and ``y.__radd__(x)`` are\n considered, as with the evaluation of ``x + y``.\n\nobject.__neg__(self)\nobject.__pos__(self)\nobject.__abs__(self)\nobject.__invert__(self)\n\n Called to implement the unary arithmetic operations (``-``, ``+``,\n ``abs()`` and ``~``).\n\nobject.__complex__(self)\nobject.__int__(self)\nobject.__long__(self)\nobject.__float__(self)\n\n Called to implement the built-in functions ``complex()``,\n ``int()``, ``long()``, and ``float()``. Should return a value of\n the appropriate type.\n\nobject.__oct__(self)\nobject.__hex__(self)\n\n Called to implement the built-in functions ``oct()`` and ``hex()``.\n Should return a string value.\n\nobject.__index__(self)\n\n Called to implement ``operator.index()``. Also called whenever\n Python needs an integer object (such as in slicing). Must return\n an integer (int or long).\n\n New in version 2.5.\n\nobject.__coerce__(self, other)\n\n Called to implement "mixed-mode" numeric arithmetic. Should either\n return a 2-tuple containing *self* and *other* converted to a\n common numeric type, or ``None`` if conversion is impossible. When\n the common type would be the type of ``other``, it is sufficient to\n return ``None``, since the interpreter will also ask the other\n object to attempt a coercion (but sometimes, if the implementation\n of the other type cannot be changed, it is useful to do the\n conversion to the other type here). A return value of\n ``NotImplemented`` is equivalent to returning ``None``.\n\n\nCoercion rules\n==============\n\nThis section used to document the rules for coercion. As the language\nhas evolved, the coercion rules have become hard to document\nprecisely; documenting what one version of one particular\nimplementation does is undesirable. Instead, here are some informal\nguidelines regarding coercion. In Python 3, coercion will not be\nsupported.\n\n* If the left operand of a % operator is a string or Unicode object,\n no coercion takes place and the string formatting operation is\n invoked instead.\n\n* It is no longer recommended to define a coercion operation. Mixed-\n mode operations on types that don\'t define coercion pass the\n original arguments to the operation.\n\n* New-style classes (those derived from ``object``) never invoke the\n ``__coerce__()`` method in response to a binary operator; the only\n time ``__coerce__()`` is invoked is when the built-in function\n ``coerce()`` is called.\n\n* For most intents and purposes, an operator that returns\n ``NotImplemented`` is treated the same as one that is not\n implemented at all.\n\n* Below, ``__op__()`` and ``__rop__()`` are used to signify the\n generic method names corresponding to an operator; ``__iop__()`` is\n used for the corresponding in-place operator. For example, for the\n operator \'``+``\', ``__add__()`` and ``__radd__()`` are used for the\n left and right variant of the binary operator, and ``__iadd__()``\n for the in-place variant.\n\n* For objects *x* and *y*, first ``x.__op__(y)`` is tried. If this is\n not implemented or returns ``NotImplemented``, ``y.__rop__(x)`` is\n tried. If this is also not implemented or returns\n ``NotImplemented``, a ``TypeError`` exception is raised. But see\n the following exception:\n\n* Exception to the previous item: if the left operand is an instance\n of a built-in type or a new-style class, and the right operand is an\n instance of a proper subclass of that type or class and overrides\n the base\'s ``__rop__()`` method, the right operand\'s ``__rop__()``\n method is tried *before* the left operand\'s ``__op__()`` method.\n\n This is done so that a subclass can completely override binary\n operators. Otherwise, the left operand\'s ``__op__()`` method would\n always accept the right operand: when an instance of a given class\n is expected, an instance of a subclass of that class is always\n acceptable.\n\n* When either operand type defines a coercion, this coercion is called\n before that type\'s ``__op__()`` or ``__rop__()`` method is called,\n but no sooner. If the coercion returns an object of a different\n type for the operand whose coercion is invoked, part of the process\n is redone using the new object.\n\n* When an in-place operator (like \'``+=``\') is used, if the left\n operand implements ``__iop__()``, it is invoked without any\n coercion. When the operation falls back to ``__op__()`` and/or\n ``__rop__()``, the normal coercion rules apply.\n\n* In ``x + y``, if *x* is a sequence that implements sequence\n concatenation, sequence concatenation is invoked.\n\n* In ``x * y``, if one operand is a sequence that implements sequence\n repetition, and the other is an integer (``int`` or ``long``),\n sequence repetition is invoked.\n\n* Rich comparisons (implemented by methods ``__eq__()`` and so on)\n never use coercion. Three-way comparison (implemented by\n ``__cmp__()``) does use coercion under the same conditions as other\n binary operations use it.\n\n* In the current implementation, the built-in numeric types ``int``,\n ``long``, ``float``, and ``complex`` do not use coercion. All these\n types implement a ``__coerce__()`` method, for use by the built-in\n ``coerce()`` function.\n\n Changed in version 2.7.\n\n\nWith Statement Context Managers\n===============================\n\nNew in version 2.5.\n\nA *context manager* is an object that defines the runtime context to\nbe established when executing a ``with`` statement. The context\nmanager handles the entry into, and the exit from, the desired runtime\ncontext for the execution of the block of code. Context managers are\nnormally invoked using the ``with`` statement (described in section\n*The with statement*), but can also be used by directly invoking their\nmethods.\n\nTypical uses of context managers include saving and restoring various\nkinds of global state, locking and unlocking resources, closing opened\nfiles, etc.\n\nFor more information on context managers, see *Context Manager Types*.\n\nobject.__enter__(self)\n\n Enter the runtime context related to this object. The ``with``\n statement will bind this method\'s return value to the target(s)\n specified in the ``as`` clause of the statement, if any.\n\nobject.__exit__(self, exc_type, exc_value, traceback)\n\n Exit the runtime context related to this object. The parameters\n describe the exception that caused the context to be exited. If the\n context was exited without an exception, all three arguments will\n be ``None``.\n\n If an exception is supplied, and the method wishes to suppress the\n exception (i.e., prevent it from being propagated), it should\n return a true value. Otherwise, the exception will be processed\n normally upon exit from this method.\n\n Note that ``__exit__()`` methods should not reraise the passed-in\n exception; this is the caller\'s responsibility.\n\nSee also:\n\n **PEP 0343** - The "with" statement\n The specification, background, and examples for the Python\n ``with`` statement.\n\n\nSpecial method lookup for old-style classes\n===========================================\n\nFor old-style classes, special methods are always looked up in exactly\nthe same way as any other method or attribute. This is the case\nregardless of whether the method is being looked up explicitly as in\n``x.__getitem__(i)`` or implicitly as in ``x[i]``.\n\nThis behaviour means that special methods may exhibit different\nbehaviour for different instances of a single old-style class if the\nappropriate special attributes are set differently:\n\n >>> class C:\n ... pass\n ...\n >>> c1 = C()\n >>> c2 = C()\n >>> c1.__len__ = lambda: 5\n >>> c2.__len__ = lambda: 9\n >>> len(c1)\n 5\n >>> len(c2)\n 9\n\n\nSpecial method lookup for new-style classes\n===========================================\n\nFor new-style classes, implicit invocations of special methods are\nonly guaranteed to work correctly if defined on an object\'s type, not\nin the object\'s instance dictionary. That behaviour is the reason why\nthe following code raises an exception (unlike the equivalent example\nwith old-style classes):\n\n >>> class C(object):\n ... pass\n ...\n >>> c = C()\n >>> c.__len__ = lambda: 5\n >>> len(c)\n Traceback (most recent call last):\n File "<stdin>", line 1, in <module>\n TypeError: object of type \'C\' has no len()\n\nThe rationale behind this behaviour lies with a number of special\nmethods such as ``__hash__()`` and ``__repr__()`` that are implemented\nby all objects, including type objects. If the implicit lookup of\nthese methods used the conventional lookup process, they would fail\nwhen invoked on the type object itself:\n\n >>> 1 .__hash__() == hash(1)\n True\n >>> int.__hash__() == hash(int)\n Traceback (most recent call last):\n File "<stdin>", line 1, in <module>\n TypeError: descriptor \'__hash__\' of \'int\' object needs an argument\n\nIncorrectly attempting to invoke an unbound method of a class in this\nway is sometimes referred to as \'metaclass confusion\', and is avoided\nby bypassing the instance when looking up special methods:\n\n >>> type(1).__hash__(1) == hash(1)\n True\n >>> type(int).__hash__(int) == hash(int)\n True\n\nIn addition to bypassing any instance attributes in the interest of\ncorrectness, implicit special method lookup generally also bypasses\nthe ``__getattribute__()`` method even of the object\'s metaclass:\n\n >>> class Meta(type):\n ... def __getattribute__(*args):\n ... print "Metaclass getattribute invoked"\n ... return type.__getattribute__(*args)\n ...\n >>> class C(object):\n ... __metaclass__ = Meta\n ... def __len__(self):\n ... return 10\n ... def __getattribute__(*args):\n ... print "Class getattribute invoked"\n ... return object.__getattribute__(*args)\n ...\n >>> c = C()\n >>> c.__len__() # Explicit lookup via instance\n Class getattribute invoked\n 10\n >>> type(c).__len__(c) # Explicit lookup via type\n Metaclass getattribute invoked\n 10\n >>> len(c) # Implicit lookup\n 10\n\nBypassing the ``__getattribute__()`` machinery in this fashion\nprovides significant scope for speed optimisations within the\ninterpreter, at the cost of some flexibility in the handling of\nspecial methods (the special method *must* be set on the class object\nitself in order to be consistently invoked by the interpreter).\n\n-[ Footnotes ]-\n\n[1] It *is* possible in some cases to change an object\'s type, under\n certain controlled conditions. It generally isn\'t a good idea\n though, since it can lead to some very strange behaviour if it is\n handled incorrectly.\n\n[2] For operands of the same type, it is assumed that if the non-\n reflected method (such as ``__add__()``) fails the operation is\n not supported, which is why the reflected method is not called.\n',
/prebuilts/gdb/linux-x86/lib/python2.7/pydoc_data/
H A D	topics.py	6 'attribute-access': '\nCustomizing attribute access\n****************************\n\nThe following methods can be defined to customize the meaning of\nattribute access (use of, assignment to, or deletion of ``x.name``)\nfor class instances.\n\nobject.__getattr__(self, name)\n\n Called when an attribute lookup has not found the attribute in the\n usual places (i.e. it is not an instance attribute nor is it found\n in the class tree for ``self``). ``name`` is the attribute name.\n This method should return the (computed) attribute value or raise\n an ``AttributeError`` exception.\n\n Note that if the attribute is found through the normal mechanism,\n ``__getattr__()`` is not called. (This is an intentional asymmetry\n between ``__getattr__()`` and ``__setattr__()``.) This is done both\n for efficiency reasons and because otherwise ``__getattr__()``\n would have no way to access other attributes of the instance. Note\n that at least for instance variables, you can fake total control by\n not inserting any values in the instance attribute dictionary (but\n instead inserting them in another object). See the\n ``__getattribute__()`` method below for a way to actually get total\n control in new-style classes.\n\nobject.__setattr__(self, name, value)\n\n Called when an attribute assignment is attempted. This is called\n instead of the normal mechanism (i.e. store the value in the\n instance dictionary). *name* is the attribute name, *value* is the\n value to be assigned to it.\n\n If ``__setattr__()`` wants to assign to an instance attribute, it\n should not simply execute ``self.name = value`` --- this would\n cause a recursive call to itself. Instead, it should insert the\n value in the dictionary of instance attributes, e.g.,\n ``self.__dict__[name] = value``. For new-style classes, rather\n than accessing the instance dictionary, it should call the base\n class method with the same name, for example,\n ``object.__setattr__(self, name, value)``.\n\nobject.__delattr__(self, name)\n\n Like ``__setattr__()`` but for attribute deletion instead of\n assignment. This should only be implemented if ``del obj.name`` is\n meaningful for the object.\n\n\nMore attribute access for new-style classes\n===========================================\n\nThe following methods only apply to new-style classes.\n\nobject.__getattribute__(self, name)\n\n Called unconditionally to implement attribute accesses for\n instances of the class. If the class also defines\n ``__getattr__()``, the latter will not be called unless\n ``__getattribute__()`` either calls it explicitly or raises an\n ``AttributeError``. This method should return the (computed)\n attribute value or raise an ``AttributeError`` exception. In order\n to avoid infinite recursion in this method, its implementation\n should always call the base class method with the same name to\n access any attributes it needs, for example,\n ``object.__getattribute__(self, name)``.\n\n Note: This method may still be bypassed when looking up special methods\n as the result of implicit invocation via language syntax or\n built-in functions. See *Special method lookup for new-style\n classes*.\n\n\nImplementing Descriptors\n========================\n\nThe following methods only apply when an instance of the class\ncontaining the method (a so-called *descriptor* class) appears in an\n*owner* class (the descriptor must be in either the owner\'s class\ndictionary or in the class dictionary for one of its parents). In the\nexamples below, "the attribute" refers to the attribute whose name is\nthe key of the property in the owner class\' ``__dict__``.\n\nobject.__get__(self, instance, owner)\n\n Called to get the attribute of the owner class (class attribute\n access) or of an instance of that class (instance attribute\n access). *owner* is always the owner class, while *instance* is the\n instance that the attribute was accessed through, or ``None`` when\n the attribute is accessed through the *owner*. This method should\n return the (computed) attribute value or raise an\n ``AttributeError`` exception.\n\nobject.__set__(self, instance, value)\n\n Called to set the attribute on an instance *instance* of the owner\n class to a new value, *value*.\n\nobject.__delete__(self, instance)\n\n Called to delete the attribute on an instance *instance* of the\n owner class.\n\n\nInvoking Descriptors\n====================\n\nIn general, a descriptor is an object attribute with "binding\nbehavior", one whose attribute access has been overridden by methods\nin the descriptor protocol: ``__get__()``, ``__set__()``, and\n``__delete__()``. If any of those methods are defined for an object,\nit is said to be a descriptor.\n\nThe default behavior for attribute access is to get, set, or delete\nthe attribute from an object\'s dictionary. For instance, ``a.x`` has a\nlookup chain starting with ``a.__dict__[\'x\']``, then\n``type(a).__dict__[\'x\']``, and continuing through the base classes of\n``type(a)`` excluding metaclasses.\n\nHowever, if the looked-up value is an object defining one of the\ndescriptor methods, then Python may override the default behavior and\ninvoke the descriptor method instead. Where this occurs in the\nprecedence chain depends on which descriptor methods were defined and\nhow they were called. Note that descriptors are only invoked for new\nstyle objects or classes (ones that subclass ``object()`` or\n``type()``).\n\nThe starting point for descriptor invocation is a binding, ``a.x``.\nHow the arguments are assembled depends on ``a``:\n\nDirect Call\n The simplest and least common call is when user code directly\n invokes a descriptor method: ``x.__get__(a)``.\n\nInstance Binding\n If binding to a new-style object instance, ``a.x`` is transformed\n into the call: ``type(a).__dict__[\'x\'].__get__(a, type(a))``.\n\nClass Binding\n If binding to a new-style class, ``A.x`` is transformed into the\n call: ``A.__dict__[\'x\'].__get__(None, A)``.\n\nSuper Binding\n If ``a`` is an instance of ``super``, then the binding ``super(B,\n obj).m()`` searches ``obj.__class__.__mro__`` for the base class\n ``A`` immediately preceding ``B`` and then invokes the descriptor\n with the call: ``A.__dict__[\'m\'].__get__(obj, obj.__class__)``.\n\nFor instance bindings, the precedence of descriptor invocation depends\non the which descriptor methods are defined. A descriptor can define\nany combination of ``__get__()``, ``__set__()`` and ``__delete__()``.\nIf it does not define ``__get__()``, then accessing the attribute will\nreturn the descriptor object itself unless there is a value in the\nobject\'s instance dictionary. If the descriptor defines ``__set__()``\nand/or ``__delete__()``, it is a data descriptor; if it defines\nneither, it is a non-data descriptor. Normally, data descriptors\ndefine both ``__get__()`` and ``__set__()``, while non-data\ndescriptors have just the ``__get__()`` method. Data descriptors with\n``__set__()`` and ``__get__()`` defined always override a redefinition\nin an instance dictionary. In contrast, non-data descriptors can be\noverridden by instances.\n\nPython methods (including ``staticmethod()`` and ``classmethod()``)\nare implemented as non-data descriptors. Accordingly, instances can\nredefine and override methods. This allows individual instances to\nacquire behaviors that differ from other instances of the same class.\n\nThe ``property()`` function is implemented as a data descriptor.\nAccordingly, instances cannot override the behavior of a property.\n\n\n__slots__\n=========\n\nBy default, instances of both old and new-style classes have a\ndictionary for attribute storage. This wastes space for objects\nhaving very few instance variables. The space consumption can become\nacute when creating large numbers of instances.\n\nThe default can be overridden by defining *__slots__* in a new-style\nclass definition. The *__slots__* declaration takes a sequence of\ninstance variables and reserves just enough space in each instance to\nhold a value for each variable. Space is saved because *__dict__* is\nnot created for each instance.\n\n__slots__\n\n This class variable can be assigned a string, iterable, or sequence\n of strings with variable names used by instances. If defined in a\n new-style class, *__slots__* reserves space for the declared\n variables and prevents the automatic creation of *__dict__* and\n *__weakref__* for each instance.\n\n New in version 2.2.\n\nNotes on using *__slots__*\n\n* When inheriting from a class without *__slots__*, the *__dict__*\n attribute of that class will always be accessible, so a *__slots__*\n definition in the subclass is meaningless.\n\n* Without a *__dict__* variable, instances cannot be assigned new\n variables not listed in the *__slots__* definition. Attempts to\n assign to an unlisted variable name raises ``AttributeError``. If\n dynamic assignment of new variables is desired, then add\n ``\'__dict__\'`` to the sequence of strings in the *__slots__*\n declaration.\n\n Changed in version 2.3: Previously, adding ``\'__dict__\'`` to the\n *__slots__* declaration would not enable the assignment of new\n attributes not specifically listed in the sequence of instance\n variable names.\n\n* Without a *__weakref__* variable for each instance, classes defining\n *__slots__* do not support weak references to its instances. If weak\n reference support is needed, then add ``\'__weakref__\'`` to the\n sequence of strings in the *__slots__* declaration.\n\n Changed in version 2.3: Previously, adding ``\'__weakref__\'`` to the\n *__slots__* declaration would not enable support for weak\n references.\n\n* *__slots__* are implemented at the class level by creating\n descriptors (*Implementing Descriptors*) for each variable name. As\n a result, class attributes cannot be used to set default values for\n instance variables defined by *__slots__*; otherwise, the class\n attribute would overwrite the descriptor assignment.\n\n* The action of a *__slots__* declaration is limited to the class\n where it is defined. As a result, subclasses will have a *__dict__*\n unless they also define *__slots__* (which must only contain names\n of any *additional* slots).\n\n* If a class defines a slot also defined in a base class, the instance\n variable defined by the base class slot is inaccessible (except by\n retrieving its descriptor directly from the base class). This\n renders the meaning of the program undefined. In the future, a\n check may be added to prevent this.\n\n* Nonempty *__slots__* does not work for classes derived from\n "variable-length" built-in types such as ``long``, ``str`` and\n ``tuple``.\n\n* Any non-string iterable may be assigned to *__slots__*. Mappings may\n also be used; however, in the future, special meaning may be\n assigned to the values corresponding to each key.\n\n* *__class__* assignment works only if both classes have the same\n *__slots__*.\n\n Changed in version 2.6: Previously, *__class__* assignment raised an\n error if either new or old class had *__slots__*.\n', 18 'calls': '\nCalls\n*****\n\nA call calls a callable object (e.g., a *function*) with a possibly\nempty series of *arguments*:\n\n call ::= primary "(" [argument_list [","]\n | expression genexpr_for] ")"\n argument_list ::= positional_arguments ["," keyword_arguments]\n ["," "*" expression] ["," keyword_arguments]\n ["," "**" expression]\n | keyword_arguments ["," "*" expression]\n ["," "**" expression]\n | "*" expression ["," "*" expression] ["," "**" expression]\n | "**" expression\n positional_arguments ::= expression ("," expression)*\n keyword_arguments ::= keyword_item ("," keyword_item)*\n keyword_item ::= identifier "=" expression\n\nA trailing comma may be present after the positional and keyword\narguments but does not affect the semantics.\n\nThe primary must evaluate to a callable object (user-defined\nfunctions, built-in functions, methods of built-in objects, class\nobjects, methods of class instances, and certain class instances\nthemselves are callable; extensions may define additional callable\nobject types). All argument expressions are evaluated before the call\nis attempted. Please refer to section *Function definitions* for the\nsyntax of formal *parameter* lists.\n\nIf keyword arguments are present, they are first converted to\npositional arguments, as follows. First, a list of unfilled slots is\ncreated for the formal parameters. If there are N positional\narguments, they are placed in the first N slots. Next, for each\nkeyword argument, the identifier is used to determine the\ncorresponding slot (if the identifier is the same as the first formal\nparameter name, the first slot is used, and so on). If the slot is\nalready filled, a ``TypeError`` exception is raised. Otherwise, the\nvalue of the argument is placed in the slot, filling it (even if the\nexpression is ``None``, it fills the slot). When all arguments have\nbeen processed, the slots that are still unfilled are filled with the\ncorresponding default value from the function definition. (Default\nvalues are calculated, once, when the function is defined; thus, a\nmutable object such as a list or dictionary used as default value will\nbe shared by all calls that don\'t specify an argument value for the\ncorresponding slot; this should usually be avoided.) If there are any\nunfilled slots for which no default value is specified, a\n``TypeError`` exception is raised. Otherwise, the list of filled\nslots is used as the argument list for the call.\n\n**CPython implementation detail:** An implementation may provide\nbuilt-in functions whose positional parameters do not have names, even\nif they are \'named\' for the purpose of documentation, and which\ntherefore cannot be supplied by keyword. In CPython, this is the case\nfor functions implemented in C that use ``PyArg_ParseTuple()`` to\nparse their arguments.\n\nIf there are more positional arguments than there are formal parameter\nslots, a ``TypeError`` exception is raised, unless a formal parameter\nusing the syntax ``*identifier`` is present; in this case, that formal\nparameter receives a tuple containing the excess positional arguments\n(or an empty tuple if there were no excess positional arguments).\n\nIf any keyword argument does not correspond to a formal parameter\nname, a ``TypeError`` exception is raised, unless a formal parameter\nusing the syntax ``**identifier`` is present; in this case, that\nformal parameter receives a dictionary containing the excess keyword\narguments (using the keywords as keys and the argument values as\ncorresponding values), or a (new) empty dictionary if there were no\nexcess keyword arguments.\n\nIf the syntax ``*expression`` appears in the function call,\n``expression`` must evaluate to an iterable. Elements from this\niterable are treated as if they were additional positional arguments;\nif there are positional arguments *x1*, ..., *xN*, and ``expression``\nevaluates to a sequence *y1*, ..., *yM*, this is equivalent to a call\nwith M+N positional arguments *x1*, ..., *xN*, *y1*, ..., *yM*.\n\nA consequence of this is that although the ``*expression`` syntax may\nappear *after* some keyword arguments, it is processed *before* the\nkeyword arguments (and the ``**expression`` argument, if any -- see\nbelow). So:\n\n >>> def f(a, b):\n ... print a, b\n ...\n >>> f(b=1, *(2,))\n 2 1\n >>> f(a=1, *(2,))\n Traceback (most recent call last):\n File "<stdin>", line 1, in ?\n TypeError: f() got multiple values for keyword argument \'a\'\n >>> f(1, *(2,))\n 1 2\n\nIt is unusual for both keyword arguments and the ``*expression``\nsyntax to be used in the same call, so in practice this confusion does\nnot arise.\n\nIf the syntax ``**expression`` appears in the function call,\n``expression`` must evaluate to a mapping, the contents of which are\ntreated as additional keyword arguments. In the case of a keyword\nappearing in both ``expression`` and as an explicit keyword argument,\na ``TypeError`` exception is raised.\n\nFormal parameters using the syntax ``*identifier`` or ``**identifier``\ncannot be used as positional argument slots or as keyword argument\nnames. Formal parameters using the syntax ``(sublist)`` cannot be\nused as keyword argument names; the outermost sublist corresponds to a\nsingle unnamed argument slot, and the argument value is assigned to\nthe sublist using the usual tuple assignment rules after all other\nparameter processing is done.\n\nA call always returns some value, possibly ``None``, unless it raises\nan exception. How this value is computed depends on the type of the\ncallable object.\n\nIf it is---\n\na user-defined function:\n The code block for the function is executed, passing it the\n argument list. The first thing the code block will do is bind the\n formal parameters to the arguments; this is described in section\n *Function definitions*. When the code block executes a ``return``\n statement, this specifies the return value of the function call.\n\na built-in function or method:\n The result is up to the interpreter; see *Built-in Functions* for\n the descriptions of built-in functions and methods.\n\na class object:\n A new instance of that class is returned.\n\na class instance method:\n The corresponding user-defined function is called, with an argument\n list that is one longer than the argument list of the call: the\n instance becomes the first argument.\n\na class instance:\n The class must define a ``__call__()`` method; the effect is then\n the same as if that method was called.\n', 61 'specialnames': '\nSpecial method names\n********************\n\nA class can implement certain operations that are invoked by special\nsyntax (such as arithmetic operations or subscripting and slicing) by\ndefining methods with special names. This is Python\'s approach to\n*operator overloading*, allowing classes to define their own behavior\nwith respect to language operators. For instance, if a class defines\na method named ``__getitem__()``, and ``x`` is an instance of this\nclass, then ``x[i]`` is roughly equivalent to ``x.__getitem__(i)`` for\nold-style classes and ``type(x).__getitem__(x, i)`` for new-style\nclasses. Except where mentioned, attempts to execute an operation\nraise an exception when no appropriate method is defined (typically\n``AttributeError`` or ``TypeError``).\n\nWhen implementing a class that emulates any built-in type, it is\nimportant that the emulation only be implemented to the degree that it\nmakes sense for the object being modelled. For example, some\nsequences may work well with retrieval of individual elements, but\nextracting a slice may not make sense. (One example of this is the\n``NodeList`` interface in the W3C\'s Document Object Model.)\n\n\nBasic customization\n===================\n\nobject.__new__(cls[, ...])\n\n Called to create a new instance of class *cls*. ``__new__()`` is a\n static method (special-cased so you need not declare it as such)\n that takes the class of which an instance was requested as its\n first argument. The remaining arguments are those passed to the\n object constructor expression (the call to the class). The return\n value of ``__new__()`` should be the new object instance (usually\n an instance of *cls*).\n\n Typical implementations create a new instance of the class by\n invoking the superclass\'s ``__new__()`` method using\n ``super(currentclass, cls).__new__(cls[, ...])`` with appropriate\n arguments and then modifying the newly-created instance as\n necessary before returning it.\n\n If ``__new__()`` returns an instance of *cls*, then the new\n instance\'s ``__init__()`` method will be invoked like\n ``__init__(self[, ...])``, where *self* is the new instance and the\n remaining arguments are the same as were passed to ``__new__()``.\n\n If ``__new__()`` does not return an instance of *cls*, then the new\n instance\'s ``__init__()`` method will not be invoked.\n\n ``__new__()`` is intended mainly to allow subclasses of immutable\n types (like int, str, or tuple) to customize instance creation. It\n is also commonly overridden in custom metaclasses in order to\n customize class creation.\n\nobject.__init__(self[, ...])\n\n Called when the instance is created. The arguments are those\n passed to the class constructor expression. If a base class has an\n ``__init__()`` method, the derived class\'s ``__init__()`` method,\n if any, must explicitly call it to ensure proper initialization of\n the base class part of the instance; for example:\n ``BaseClass.__init__(self, [args...])``. As a special constraint\n on constructors, no value may be returned; doing so will cause a\n ``TypeError`` to be raised at runtime.\n\nobject.__del__(self)\n\n Called when the instance is about to be destroyed. This is also\n called a destructor. If a base class has a ``__del__()`` method,\n the derived class\'s ``__del__()`` method, if any, must explicitly\n call it to ensure proper deletion of the base class part of the\n instance. Note that it is possible (though not recommended!) for\n the ``__del__()`` method to postpone destruction of the instance by\n creating a new reference to it. It may then be called at a later\n time when this new reference is deleted. It is not guaranteed that\n ``__del__()`` methods are called for objects that still exist when\n the interpreter exits.\n\n Note: ``del x`` doesn\'t directly call ``x.__del__()`` --- the former\n decrements the reference count for ``x`` by one, and the latter\n is only called when ``x``\'s reference count reaches zero. Some\n common situations that may prevent the reference count of an\n object from going to zero include: circular references between\n objects (e.g., a doubly-linked list or a tree data structure with\n parent and child pointers); a reference to the object on the\n stack frame of a function that caught an exception (the traceback\n stored in ``sys.exc_traceback`` keeps the stack frame alive); or\n a reference to the object on the stack frame that raised an\n unhandled exception in interactive mode (the traceback stored in\n ``sys.last_traceback`` keeps the stack frame alive). The first\n situation can only be remedied by explicitly breaking the cycles;\n the latter two situations can be resolved by storing ``None`` in\n ``sys.exc_traceback`` or ``sys.last_traceback``. Circular\n references which are garbage are detected when the option cycle\n detector is enabled (it\'s on by default), but can only be cleaned\n up if there are no Python-level ``__del__()`` methods involved.\n Refer to the documentation for the ``gc`` module for more\n information about how ``__del__()`` methods are handled by the\n cycle detector, particularly the description of the ``garbage``\n value.\n\n Warning: Due to the precarious circumstances under which ``__del__()``\n methods are invoked, exceptions that occur during their execution\n are ignored, and a warning is printed to ``sys.stderr`` instead.\n Also, when ``__del__()`` is invoked in response to a module being\n deleted (e.g., when execution of the program is done), other\n globals referenced by the ``__del__()`` method may already have\n been deleted or in the process of being torn down (e.g. the\n import machinery shutting down). For this reason, ``__del__()``\n methods should do the absolute minimum needed to maintain\n external invariants. Starting with version 1.5, Python\n guarantees that globals whose name begins with a single\n underscore are deleted from their module before other globals are\n deleted; if no other references to such globals exist, this may\n help in assuring that imported modules are still available at the\n time when the ``__del__()`` method is called.\n\n See also the *-R* command-line option.\n\nobject.__repr__(self)\n\n Called by the ``repr()`` built-in function and by string\n conversions (reverse quotes) to compute the "official" string\n representation of an object. If at all possible, this should look\n like a valid Python expression that could be used to recreate an\n object with the same value (given an appropriate environment). If\n this is not possible, a string of the form ``<...some useful\n description...>`` should be returned. The return value must be a\n string object. If a class defines ``__repr__()`` but not\n ``__str__()``, then ``__repr__()`` is also used when an "informal"\n string representation of instances of that class is required.\n\n This is typically used for debugging, so it is important that the\n representation is information-rich and unambiguous.\n\nobject.__str__(self)\n\n Called by the ``str()`` built-in function and by the ``print``\n statement to compute the "informal" string representation of an\n object. This differs from ``__repr__()`` in that it does not have\n to be a valid Python expression: a more convenient or concise\n representation may be used instead. The return value must be a\n string object.\n\nobject.__lt__(self, other)\nobject.__le__(self, other)\nobject.__eq__(self, other)\nobject.__ne__(self, other)\nobject.__gt__(self, other)\nobject.__ge__(self, other)\n\n New in version 2.1.\n\n These are the so-called "rich comparison" methods, and are called\n for comparison operators in preference to ``__cmp__()`` below. The\n correspondence between operator symbols and method names is as\n follows: ``x<y`` calls ``x.__lt__(y)``, ``x<=y`` calls\n ``x.__le__(y)``, ``x==y`` calls ``x.__eq__(y)``, ``x!=y`` and\n ``x<>y`` call ``x.__ne__(y)``, ``x>y`` calls ``x.__gt__(y)``, and\n ``x>=y`` calls ``x.__ge__(y)``.\n\n A rich comparison method may return the singleton\n ``NotImplemented`` if it does not implement the operation for a\n given pair of arguments. By convention, ``False`` and ``True`` are\n returned for a successful comparison. However, these methods can\n return any value, so if the comparison operator is used in a\n Boolean context (e.g., in the condition of an ``if`` statement),\n Python will call ``bool()`` on the value to determine if the result\n is true or false.\n\n There are no implied relationships among the comparison operators.\n The truth of ``x==y`` does not imply that ``x!=y`` is false.\n Accordingly, when defining ``__eq__()``, one should also define\n ``__ne__()`` so that the operators will behave as expected. See\n the paragraph on ``__hash__()`` for some important notes on\n creating *hashable* objects which support custom comparison\n operations and are usable as dictionary keys.\n\n There are no swapped-argument versions of these methods (to be used\n when the left argument does not support the operation but the right\n argument does); rather, ``__lt__()`` and ``__gt__()`` are each\n other\'s reflection, ``__le__()`` and ``__ge__()`` are each other\'s\n reflection, and ``__eq__()`` and ``__ne__()`` are their own\n reflection.\n\n Arguments to rich comparison methods are never coerced.\n\n To automatically generate ordering operations from a single root\n operation, see ``functools.total_ordering()``.\n\nobject.__cmp__(self, other)\n\n Called by comparison operations if rich comparison (see above) is\n not defined. Should return a negative integer if ``self < other``,\n zero if ``self == other``, a positive integer if ``self > other``.\n If no ``__cmp__()``, ``__eq__()`` or ``__ne__()`` operation is\n defined, class instances are compared by object identity\n ("address"). See also the description of ``__hash__()`` for some\n important notes on creating *hashable* objects which support custom\n comparison operations and are usable as dictionary keys. (Note: the\n restriction that exceptions are not propagated by ``__cmp__()`` has\n been removed since Python 1.5.)\n\nobject.__rcmp__(self, other)\n\n Changed in version 2.1: No longer supported.\n\nobject.__hash__(self)\n\n Called by built-in function ``hash()`` and for operations on\n members of hashed collections including ``set``, ``frozenset``, and\n ``dict``. ``__hash__()`` should return an integer. The only\n required property is that objects which compare equal have the same\n hash value; it is advised to somehow mix together (e.g. using\n exclusive or) the hash values for the components of the object that\n also play a part in comparison of objects.\n\n If a class does not define a ``__cmp__()`` or ``__eq__()`` method\n it should not define a ``__hash__()`` operation either; if it\n defines ``__cmp__()`` or ``__eq__()`` but not ``__hash__()``, its\n instances will not be usable in hashed collections. If a class\n defines mutable objects and implements a ``__cmp__()`` or\n ``__eq__()`` method, it should not implement ``__hash__()``, since\n hashable collection implementations require that a object\'s hash\n value is immutable (if the object\'s hash value changes, it will be\n in the wrong hash bucket).\n\n User-defined classes have ``__cmp__()`` and ``__hash__()`` methods\n by default; with them, all objects compare unequal (except with\n themselves) and ``x.__hash__()`` returns ``id(x)``.\n\n Classes which inherit a ``__hash__()`` method from a parent class\n but change the meaning of ``__cmp__()`` or ``__eq__()`` such that\n the hash value returned is no longer appropriate (e.g. by switching\n to a value-based concept of equality instead of the default\n identity based equality) can explicitly flag themselves as being\n unhashable by setting ``__hash__ = None`` in the class definition.\n Doing so means that not only will instances of the class raise an\n appropriate ``TypeError`` when a program attempts to retrieve their\n hash value, but they will also be correctly identified as\n unhashable when checking ``isinstance(obj, collections.Hashable)``\n (unlike classes which define their own ``__hash__()`` to explicitly\n raise ``TypeError``).\n\n Changed in version 2.5: ``__hash__()`` may now also return a long\n integer object; the 32-bit integer is then derived from the hash of\n that object.\n\n Changed in version 2.6: ``__hash__`` may now be set to ``None`` to\n explicitly flag instances of a class as unhashable.\n\nobject.__nonzero__(self)\n\n Called to implement truth value testing and the built-in operation\n ``bool()``; should return ``False`` or ``True``, or their integer\n equivalents ``0`` or ``1``. When this method is not defined,\n ``__len__()`` is called, if it is defined, and the object is\n considered true if its result is nonzero. If a class defines\n neither ``__len__()`` nor ``__nonzero__()``, all its instances are\n considered true.\n\nobject.__unicode__(self)\n\n Called to implement ``unicode()`` built-in; should return a Unicode\n object. When this method is not defined, string conversion is\n attempted, and the result of string conversion is converted to\n Unicode using the system default encoding.\n\n\nCustomizing attribute access\n============================\n\nThe following methods can be defined to customize the meaning of\nattribute access (use of, assignment to, or deletion of ``x.name``)\nfor class instances.\n\nobject.__getattr__(self, name)\n\n Called when an attribute lookup has not found the attribute in the\n usual places (i.e. it is not an instance attribute nor is it found\n in the class tree for ``self``). ``name`` is the attribute name.\n This method should return the (computed) attribute value or raise\n an ``AttributeError`` exception.\n\n Note that if the attribute is found through the normal mechanism,\n ``__getattr__()`` is not called. (This is an intentional asymmetry\n between ``__getattr__()`` and ``__setattr__()``.) This is done both\n for efficiency reasons and because otherwise ``__getattr__()``\n would have no way to access other attributes of the instance. Note\n that at least for instance variables, you can fake total control by\n not inserting any values in the instance attribute dictionary (but\n instead inserting them in another object). See the\n ``__getattribute__()`` method below for a way to actually get total\n control in new-style classes.\n\nobject.__setattr__(self, name, value)\n\n Called when an attribute assignment is attempted. This is called\n instead of the normal mechanism (i.e. store the value in the\n instance dictionary). *name* is the attribute name, *value* is the\n value to be assigned to it.\n\n If ``__setattr__()`` wants to assign to an instance attribute, it\n should not simply execute ``self.name = value`` --- this would\n cause a recursive call to itself. Instead, it should insert the\n value in the dictionary of instance attributes, e.g.,\n ``self.__dict__[name] = value``. For new-style classes, rather\n than accessing the instance dictionary, it should call the base\n class method with the same name, for example,\n ``object.__setattr__(self, name, value)``.\n\nobject.__delattr__(self, name)\n\n Like ``__setattr__()`` but for attribute deletion instead of\n assignment. This should only be implemented if ``del obj.name`` is\n meaningful for the object.\n\n\nMore attribute access for new-style classes\n-------------------------------------------\n\nThe following methods only apply to new-style classes.\n\nobject.__getattribute__(self, name)\n\n Called unconditionally to implement attribute accesses for\n instances of the class. If the class also defines\n ``__getattr__()``, the latter will not be called unless\n ``__getattribute__()`` either calls it explicitly or raises an\n ``AttributeError``. This method should return the (computed)\n attribute value or raise an ``AttributeError`` exception. In order\n to avoid infinite recursion in this method, its implementation\n should always call the base class method with the same name to\n access any attributes it needs, for example,\n ``object.__getattribute__(self, name)``.\n\n Note: This method may still be bypassed when looking up special methods\n as the result of implicit invocation via language syntax or\n built-in functions. See *Special method lookup for new-style\n classes*.\n\n\nImplementing Descriptors\n------------------------\n\nThe following methods only apply when an instance of the class\ncontaining the method (a so-called *descriptor* class) appears in an\n*owner* class (the descriptor must be in either the owner\'s class\ndictionary or in the class dictionary for one of its parents). In the\nexamples below, "the attribute" refers to the attribute whose name is\nthe key of the property in the owner class\' ``__dict__``.\n\nobject.__get__(self, instance, owner)\n\n Called to get the attribute of the owner class (class attribute\n access) or of an instance of that class (instance attribute\n access). *owner* is always the owner class, while *instance* is the\n instance that the attribute was accessed through, or ``None`` when\n the attribute is accessed through the *owner*. This method should\n return the (computed) attribute value or raise an\n ``AttributeError`` exception.\n\nobject.__set__(self, instance, value)\n\n Called to set the attribute on an instance *instance* of the owner\n class to a new value, *value*.\n\nobject.__delete__(self, instance)\n\n Called to delete the attribute on an instance *instance* of the\n owner class.\n\n\nInvoking Descriptors\n--------------------\n\nIn general, a descriptor is an object attribute with "binding\nbehavior", one whose attribute access has been overridden by methods\nin the descriptor protocol: ``__get__()``, ``__set__()``, and\n``__delete__()``. If any of those methods are defined for an object,\nit is said to be a descriptor.\n\nThe default behavior for attribute access is to get, set, or delete\nthe attribute from an object\'s dictionary. For instance, ``a.x`` has a\nlookup chain starting with ``a.__dict__[\'x\']``, then\n``type(a).__dict__[\'x\']``, and continuing through the base classes of\n``type(a)`` excluding metaclasses.\n\nHowever, if the looked-up value is an object defining one of the\ndescriptor methods, then Python may override the default behavior and\ninvoke the descriptor method instead. Where this occurs in the\nprecedence chain depends on which descriptor methods were defined and\nhow they were called. Note that descriptors are only invoked for new\nstyle objects or classes (ones that subclass ``object()`` or\n``type()``).\n\nThe starting point for descriptor invocation is a binding, ``a.x``.\nHow the arguments are assembled depends on ``a``:\n\nDirect Call\n The simplest and least common call is when user code directly\n invokes a descriptor method: ``x.__get__(a)``.\n\nInstance Binding\n If binding to a new-style object instance, ``a.x`` is transformed\n into the call: ``type(a).__dict__[\'x\'].__get__(a, type(a))``.\n\nClass Binding\n If binding to a new-style class, ``A.x`` is transformed into the\n call: ``A.__dict__[\'x\'].__get__(None, A)``.\n\nSuper Binding\n If ``a`` is an instance of ``super``, then the binding ``super(B,\n obj).m()`` searches ``obj.__class__.__mro__`` for the base class\n ``A`` immediately preceding ``B`` and then invokes the descriptor\n with the call: ``A.__dict__[\'m\'].__get__(obj, obj.__class__)``.\n\nFor instance bindings, the precedence of descriptor invocation depends\non the which descriptor methods are defined. A descriptor can define\nany combination of ``__get__()``, ``__set__()`` and ``__delete__()``.\nIf it does not define ``__get__()``, then accessing the attribute will\nreturn the descriptor object itself unless there is a value in the\nobject\'s instance dictionary. If the descriptor defines ``__set__()``\nand/or ``__delete__()``, it is a data descriptor; if it defines\nneither, it is a non-data descriptor. Normally, data descriptors\ndefine both ``__get__()`` and ``__set__()``, while non-data\ndescriptors have just the ``__get__()`` method. Data descriptors with\n``__set__()`` and ``__get__()`` defined always override a redefinition\nin an instance dictionary. In contrast, non-data descriptors can be\noverridden by instances.\n\nPython methods (including ``staticmethod()`` and ``classmethod()``)\nare implemented as non-data descriptors. Accordingly, instances can\nredefine and override methods. This allows individual instances to\nacquire behaviors that differ from other instances of the same class.\n\nThe ``property()`` function is implemented as a data descriptor.\nAccordingly, instances cannot override the behavior of a property.\n\n\n__slots__\n---------\n\nBy default, instances of both old and new-style classes have a\ndictionary for attribute storage. This wastes space for objects\nhaving very few instance variables. The space consumption can become\nacute when creating large numbers of instances.\n\nThe default can be overridden by defining *__slots__* in a new-style\nclass definition. The *__slots__* declaration takes a sequence of\ninstance variables and reserves just enough space in each instance to\nhold a value for each variable. Space is saved because *__dict__* is\nnot created for each instance.\n\n__slots__\n\n This class variable can be assigned a string, iterable, or sequence\n of strings with variable names used by instances. If defined in a\n new-style class, *__slots__* reserves space for the declared\n variables and prevents the automatic creation of *__dict__* and\n *__weakref__* for each instance.\n\n New in version 2.2.\n\nNotes on using *__slots__*\n\n* When inheriting from a class without *__slots__*, the *__dict__*\n attribute of that class will always be accessible, so a *__slots__*\n definition in the subclass is meaningless.\n\n* Without a *__dict__* variable, instances cannot be assigned new\n variables not listed in the *__slots__* definition. Attempts to\n assign to an unlisted variable name raises ``AttributeError``. If\n dynamic assignment of new variables is desired, then add\n ``\'__dict__\'`` to the sequence of strings in the *__slots__*\n declaration.\n\n Changed in version 2.3: Previously, adding ``\'__dict__\'`` to the\n *__slots__* declaration would not enable the assignment of new\n attributes not specifically listed in the sequence of instance\n variable names.\n\n* Without a *__weakref__* variable for each instance, classes defining\n *__slots__* do not support weak references to its instances. If weak\n reference support is needed, then add ``\'__weakref__\'`` to the\n sequence of strings in the *__slots__* declaration.\n\n Changed in version 2.3: Previously, adding ``\'__weakref__\'`` to the\n *__slots__* declaration would not enable support for weak\n references.\n\n* *__slots__* are implemented at the class level by creating\n descriptors (*Implementing Descriptors*) for each variable name. As\n a result, class attributes cannot be used to set default values for\n instance variables defined by *__slots__*; otherwise, the class\n attribute would overwrite the descriptor assignment.\n\n* The action of a *__slots__* declaration is limited to the class\n where it is defined. As a result, subclasses will have a *__dict__*\n unless they also define *__slots__* (which must only contain names\n of any *additional* slots).\n\n* If a class defines a slot also defined in a base class, the instance\n variable defined by the base class slot is inaccessible (except by\n retrieving its descriptor directly from the base class). This\n renders the meaning of the program undefined. In the future, a\n check may be added to prevent this.\n\n* Nonempty *__slots__* does not work for classes derived from\n "variable-length" built-in types such as ``long``, ``str`` and\n ``tuple``.\n\n* Any non-string iterable may be assigned to *__slots__*. Mappings may\n also be used; however, in the future, special meaning may be\n assigned to the values corresponding to each key.\n\n* *__class__* assignment works only if both classes have the same\n *__slots__*.\n\n Changed in version 2.6: Previously, *__class__* assignment raised an\n error if either new or old class had *__slots__*.\n\n\nCustomizing class creation\n==========================\n\nBy default, new-style classes are constructed using ``type()``. A\nclass definition is read into a separate namespace and the value of\nclass name is bound to the result of ``type(name, bases, dict)``.\n\nWhen the class definition is read, if *__metaclass__* is defined then\nthe callable assigned to it will be called instead of ``type()``. This\nallows classes or functions to be written which monitor or alter the\nclass creation process:\n\n* Modifying the class dictionary prior to the class being created.\n\n* Returning an instance of another class -- essentially performing the\n role of a factory function.\n\nThese steps will have to be performed in the metaclass\'s ``__new__()``\nmethod -- ``type.__new__()`` can then be called from this method to\ncreate a class with different properties. This example adds a new\nelement to the class dictionary before creating the class:\n\n class metacls(type):\n def __new__(mcs, name, bases, dict):\n dict[\'foo\'] = \'metacls was here\'\n return type.__new__(mcs, name, bases, dict)\n\nYou can of course also override other class methods (or add new\nmethods); for example defining a custom ``__call__()`` method in the\nmetaclass allows custom behavior when the class is called, e.g. not\nalways creating a new instance.\n\n__metaclass__\n\n This variable can be any callable accepting arguments for ``name``,\n ``bases``, and ``dict``. Upon class creation, the callable is used\n instead of the built-in ``type()``.\n\n New in version 2.2.\n\nThe appropriate metaclass is determined by the following precedence\nrules:\n\n* If ``dict[\'__metaclass__\']`` exists, it is used.\n\n* Otherwise, if there is at least one base class, its metaclass is\n used (this looks for a *__class__* attribute first and if not found,\n uses its type).\n\n* Otherwise, if a global variable named __metaclass__ exists, it is\n used.\n\n* Otherwise, the old-style, classic metaclass (types.ClassType) is\n used.\n\nThe potential uses for metaclasses are boundless. Some ideas that have\nbeen explored including logging, interface checking, automatic\ndelegation, automatic property creation, proxies, frameworks, and\nautomatic resource locking/synchronization.\n\n\nCustomizing instance and subclass checks\n========================================\n\nNew in version 2.6.\n\nThe following methods are used to override the default behavior of the\n``isinstance()`` and ``issubclass()`` built-in functions.\n\nIn particular, the metaclass ``abc.ABCMeta`` implements these methods\nin order to allow the addition of Abstract Base Classes (ABCs) as\n"virtual base classes" to any class or type (including built-in\ntypes), including other ABCs.\n\nclass.__instancecheck__(self, instance)\n\n Return true if *instance* should be considered a (direct or\n indirect) instance of *class*. If defined, called to implement\n ``isinstance(instance, class)``.\n\nclass.__subclasscheck__(self, subclass)\n\n Return true if *subclass* should be considered a (direct or\n indirect) subclass of *class*. If defined, called to implement\n ``issubclass(subclass, class)``.\n\nNote that these methods are looked up on the type (metaclass) of a\nclass. They cannot be defined as class methods in the actual class.\nThis is consistent with the lookup of special methods that are called\non instances, only in this case the instance is itself a class.\n\nSee also:\n\n **PEP 3119** - Introducing Abstract Base Classes\n Includes the specification for customizing ``isinstance()`` and\n ``issubclass()`` behavior through ``__instancecheck__()`` and\n ``__subclasscheck__()``, with motivation for this functionality\n in the context of adding Abstract Base Classes (see the ``abc``\n module) to the language.\n\n\nEmulating callable objects\n==========================\n\nobject.__call__(self[, args...])\n\n Called when the instance is "called" as a function; if this method\n is defined, ``x(arg1, arg2, ...)`` is a shorthand for\n ``x.__call__(arg1, arg2, ...)``.\n\n\nEmulating container types\n=========================\n\nThe following methods can be defined to implement container objects.\nContainers usually are sequences (such as lists or tuples) or mappings\n(like dictionaries), but can represent other containers as well. The\nfirst set of methods is used either to emulate a sequence or to\nemulate a mapping; the difference is that for a sequence, the\nallowable keys should be the integers *k* for which ``0 <= k < N``\nwhere *N* is the length of the sequence, or slice objects, which\ndefine a range of items. (For backwards compatibility, the method\n``__getslice__()`` (see below) can also be defined to handle simple,\nbut not extended slices.) It is also recommended that mappings provide\nthe methods ``keys()``, ``values()``, ``items()``, ``has_key()``,\n``get()``, ``clear()``, ``setdefault()``, ``iterkeys()``,\n``itervalues()``, ``iteritems()``, ``pop()``, ``popitem()``,\n``copy()``, and ``update()`` behaving similar to those for Python\'s\nstandard dictionary objects. The ``UserDict`` module provides a\n``DictMixin`` class to help create those methods from a base set of\n``__getitem__()``, ``__setitem__()``, ``__delitem__()``, and\n``keys()``. Mutable sequences should provide methods ``append()``,\n``count()``, ``index()``, ``extend()``, ``insert()``, ``pop()``,\n``remove()``, ``reverse()`` and ``sort()``, like Python standard list\nobjects. Finally, sequence types should implement addition (meaning\nconcatenation) and multiplication (meaning repetition) by defining the\nmethods ``__add__()``, ``__radd__()``, ``__iadd__()``, ``__mul__()``,\n``__rmul__()`` and ``__imul__()`` described below; they should not\ndefine ``__coerce__()`` or other numerical operators. It is\nrecommended that both mappings and sequences implement the\n``__contains__()`` method to allow efficient use of the ``in``\noperator; for mappings, ``in`` should be equivalent of ``has_key()``;\nfor sequences, it should search through the values. It is further\nrecommended that both mappings and sequences implement the\n``__iter__()`` method to allow efficient iteration through the\ncontainer; for mappings, ``__iter__()`` should be the same as\n``iterkeys()``; for sequences, it should iterate through the values.\n\nobject.__len__(self)\n\n Called to implement the built-in function ``len()``. Should return\n the length of the object, an integer ``>=`` 0. Also, an object\n that doesn\'t define a ``__nonzero__()`` method and whose\n ``__len__()`` method returns zero is considered to be false in a\n Boolean context.\n\nobject.__getitem__(self, key)\n\n Called to implement evaluation of ``self[key]``. For sequence\n types, the accepted keys should be integers and slice objects.\n Note that the special interpretation of negative indexes (if the\n class wishes to emulate a sequence type) is up to the\n ``__getitem__()`` method. If *key* is of an inappropriate type,\n ``TypeError`` may be raised; if of a value outside the set of\n indexes for the sequence (after any special interpretation of\n negative values), ``IndexError`` should be raised. For mapping\n types, if *key* is missing (not in the container), ``KeyError``\n should be raised.\n\n Note: ``for`` loops expect that an ``IndexError`` will be raised for\n illegal indexes to allow proper detection of the end of the\n sequence.\n\nobject.__setitem__(self, key, value)\n\n Called to implement assignment to ``self[key]``. Same note as for\n ``__getitem__()``. This should only be implemented for mappings if\n the objects support changes to the values for keys, or if new keys\n can be added, or for sequences if elements can be replaced. The\n same exceptions should be raised for improper *key* values as for\n the ``__getitem__()`` method.\n\nobject.__delitem__(self, key)\n\n Called to implement deletion of ``self[key]``. Same note as for\n ``__getitem__()``. This should only be implemented for mappings if\n the objects support removal of keys, or for sequences if elements\n can be removed from the sequence. The same exceptions should be\n raised for improper *key* values as for the ``__getitem__()``\n method.\n\nobject.__iter__(self)\n\n This method is called when an iterator is required for a container.\n This method should return a new iterator object that can iterate\n over all the objects in the container. For mappings, it should\n iterate over the keys of the container, and should also be made\n available as the method ``iterkeys()``.\n\n Iterator objects also need to implement this method; they are\n required to return themselves. For more information on iterator\n objects, see *Iterator Types*.\n\nobject.__reversed__(self)\n\n Called (if present) by the ``reversed()`` built-in to implement\n reverse iteration. It should return a new iterator object that\n iterates over all the objects in the container in reverse order.\n\n If the ``__reversed__()`` method is not provided, the\n ``reversed()`` built-in will fall back to using the sequence\n protocol (``__len__()`` and ``__getitem__()``). Objects that\n support the sequence protocol should only provide\n ``__reversed__()`` if they can provide an implementation that is\n more efficient than the one provided by ``reversed()``.\n\n New in version 2.6.\n\nThe membership test operators (``in`` and ``not in``) are normally\nimplemented as an iteration through a sequence. However, container\nobjects can supply the following special method with a more efficient\nimplementation, which also does not require the object be a sequence.\n\nobject.__contains__(self, item)\n\n Called to implement membership test operators. Should return true\n if *item* is in *self*, false otherwise. For mapping objects, this\n should consider the keys of the mapping rather than the values or\n the key-item pairs.\n\n For objects that don\'t define ``__contains__()``, the membership\n test first tries iteration via ``__iter__()``, then the old\n sequence iteration protocol via ``__getitem__()``, see *this\n section in the language reference*.\n\n\nAdditional methods for emulation of sequence types\n==================================================\n\nThe following optional methods can be defined to further emulate\nsequence objects. Immutable sequences methods should at most only\ndefine ``__getslice__()``; mutable sequences might define all three\nmethods.\n\nobject.__getslice__(self, i, j)\n\n Deprecated since version 2.0: Support slice objects as parameters\n to the ``__getitem__()`` method. (However, built-in types in\n CPython currently still implement ``__getslice__()``. Therefore,\n you have to override it in derived classes when implementing\n slicing.)\n\n Called to implement evaluation of ``self[i:j]``. The returned\n object should be of the same type as *self*. Note that missing *i*\n or *j* in the slice expression are replaced by zero or\n ``sys.maxint``, respectively. If negative indexes are used in the\n slice, the length of the sequence is added to that index. If the\n instance does not implement the ``__len__()`` method, an\n ``AttributeError`` is raised. No guarantee is made that indexes\n adjusted this way are not still negative. Indexes which are\n greater than the length of the sequence are not modified. If no\n ``__getslice__()`` is found, a slice object is created instead, and\n passed to ``__getitem__()`` instead.\n\nobject.__setslice__(self, i, j, sequence)\n\n Called to implement assignment to ``self[i:j]``. Same notes for *i*\n and *j* as for ``__getslice__()``.\n\n This method is deprecated. If no ``__setslice__()`` is found, or\n for extended slicing of the form ``self[i:j:k]``, a slice object is\n created, and passed to ``__setitem__()``, instead of\n ``__setslice__()`` being called.\n\nobject.__delslice__(self, i, j)\n\n Called to implement deletion of ``self[i:j]``. Same notes for *i*\n and *j* as for ``__getslice__()``. This method is deprecated. If no\n ``__delslice__()`` is found, or for extended slicing of the form\n ``self[i:j:k]``, a slice object is created, and passed to\n ``__delitem__()``, instead of ``__delslice__()`` being called.\n\nNotice that these methods are only invoked when a single slice with a\nsingle colon is used, and the slice method is available. For slice\noperations involving extended slice notation, or in absence of the\nslice methods, ``__getitem__()``, ``__setitem__()`` or\n``__delitem__()`` is called with a slice object as argument.\n\nThe following example demonstrate how to make your program or module\ncompatible with earlier versions of Python (assuming that methods\n``__getitem__()``, ``__setitem__()`` and ``__delitem__()`` support\nslice objects as arguments):\n\n class MyClass:\n ...\n def __getitem__(self, index):\n ...\n def __setitem__(self, index, value):\n ...\n def __delitem__(self, index):\n ...\n\n if sys.version_info < (2, 0):\n # They won\'t be defined if version is at least 2.0 final\n\n def __getslice__(self, i, j):\n return self[max(0, i):max(0, j):]\n def __setslice__(self, i, j, seq):\n self[max(0, i):max(0, j):] = seq\n def __delslice__(self, i, j):\n del self[max(0, i):max(0, j):]\n ...\n\nNote the calls to ``max()``; these are necessary because of the\nhandling of negative indices before the ``__*slice__()`` methods are\ncalled. When negative indexes are used, the ``__*item__()`` methods\nreceive them as provided, but the ``__*slice__()`` methods get a\n"cooked" form of the index values. For each negative index value, the\nlength of the sequence is added to the index before calling the method\n(which may still result in a negative index); this is the customary\nhandling of negative indexes by the built-in sequence types, and the\n``__*item__()`` methods are expected to do this as well. However,\nsince they should already be doing that, negative indexes cannot be\npassed in; they must be constrained to the bounds of the sequence\nbefore being passed to the ``__*item__()`` methods. Calling ``max(0,\ni)`` conveniently returns the proper value.\n\n\nEmulating numeric types\n=======================\n\nThe following methods can be defined to emulate numeric objects.\nMethods corresponding to operations that are not supported by the\nparticular kind of number implemented (e.g., bitwise operations for\nnon-integral numbers) should be left undefined.\n\nobject.__add__(self, other)\nobject.__sub__(self, other)\nobject.__mul__(self, other)\nobject.__floordiv__(self, other)\nobject.__mod__(self, other)\nobject.__divmod__(self, other)\nobject.__pow__(self, other[, modulo])\nobject.__lshift__(self, other)\nobject.__rshift__(self, other)\nobject.__and__(self, other)\nobject.__xor__(self, other)\nobject.__or__(self, other)\n\n These methods are called to implement the binary arithmetic\n operations (``+``, ``-``, ``*``, ``//``, ``%``, ``divmod()``,\n ``pow()``, ``**``, ``<<``, ``>>``, ``&``, ``^``, ``|``). For\n instance, to evaluate the expression ``x + y``, where *x* is an\n instance of a class that has an ``__add__()`` method,\n ``x.__add__(y)`` is called. The ``__divmod__()`` method should be\n the equivalent to using ``__floordiv__()`` and ``__mod__()``; it\n should not be related to ``__truediv__()`` (described below). Note\n that ``__pow__()`` should be defined to accept an optional third\n argument if the ternary version of the built-in ``pow()`` function\n is to be supported.\n\n If one of those methods does not support the operation with the\n supplied arguments, it should return ``NotImplemented``.\n\nobject.__div__(self, other)\nobject.__truediv__(self, other)\n\n The division operator (``/``) is implemented by these methods. The\n ``__truediv__()`` method is used when ``__future__.division`` is in\n effect, otherwise ``__div__()`` is used. If only one of these two\n methods is defined, the object will not support division in the\n alternate context; ``TypeError`` will be raised instead.\n\nobject.__radd__(self, other)\nobject.__rsub__(self, other)\nobject.__rmul__(self, other)\nobject.__rdiv__(self, other)\nobject.__rtruediv__(self, other)\nobject.__rfloordiv__(self, other)\nobject.__rmod__(self, other)\nobject.__rdivmod__(self, other)\nobject.__rpow__(self, other)\nobject.__rlshift__(self, other)\nobject.__rrshift__(self, other)\nobject.__rand__(self, other)\nobject.__rxor__(self, other)\nobject.__ror__(self, other)\n\n These methods are called to implement the binary arithmetic\n operations (``+``, ``-``, ``*``, ``/``, ``%``, ``divmod()``,\n ``pow()``, ``**``, ``<<``, ``>>``, ``&``, ``^``, ``|``) with\n reflected (swapped) operands. These functions are only called if\n the left operand does not support the corresponding operation and\n the operands are of different types. [2] For instance, to evaluate\n the expression ``x - y``, where *y* is an instance of a class that\n has an ``__rsub__()`` method, ``y.__rsub__(x)`` is called if\n ``x.__sub__(y)`` returns *NotImplemented*.\n\n Note that ternary ``pow()`` will not try calling ``__rpow__()``\n (the coercion rules would become too complicated).\n\n Note: If the right operand\'s type is a subclass of the left operand\'s\n type and that subclass provides the reflected method for the\n operation, this method will be called before the left operand\'s\n non-reflected method. This behavior allows subclasses to\n override their ancestors\' operations.\n\nobject.__iadd__(self, other)\nobject.__isub__(self, other)\nobject.__imul__(self, other)\nobject.__idiv__(self, other)\nobject.__itruediv__(self, other)\nobject.__ifloordiv__(self, other)\nobject.__imod__(self, other)\nobject.__ipow__(self, other[, modulo])\nobject.__ilshift__(self, other)\nobject.__irshift__(self, other)\nobject.__iand__(self, other)\nobject.__ixor__(self, other)\nobject.__ior__(self, other)\n\n These methods are called to implement the augmented arithmetic\n assignments (``+=``, ``-=``, ``*=``, ``/=``, ``//=``, ``%=``,\n ``**=``, ``<<=``, ``>>=``, ``&=``, ``^=``, ``|=``). These methods\n should attempt to do the operation in-place (modifying *self*) and\n return the result (which could be, but does not have to be,\n *self*). If a specific method is not defined, the augmented\n assignment falls back to the normal methods. For instance, to\n execute the statement ``x += y``, where *x* is an instance of a\n class that has an ``__iadd__()`` method, ``x.__iadd__(y)`` is\n called. If *x* is an instance of a class that does not define a\n ``__iadd__()`` method, ``x.__add__(y)`` and ``y.__radd__(x)`` are\n considered, as with the evaluation of ``x + y``.\n\nobject.__neg__(self)\nobject.__pos__(self)\nobject.__abs__(self)\nobject.__invert__(self)\n\n Called to implement the unary arithmetic operations (``-``, ``+``,\n ``abs()`` and ``~``).\n\nobject.__complex__(self)\nobject.__int__(self)\nobject.__long__(self)\nobject.__float__(self)\n\n Called to implement the built-in functions ``complex()``,\n ``int()``, ``long()``, and ``float()``. Should return a value of\n the appropriate type.\n\nobject.__oct__(self)\nobject.__hex__(self)\n\n Called to implement the built-in functions ``oct()`` and ``hex()``.\n Should return a string value.\n\nobject.__index__(self)\n\n Called to implement ``operator.index()``. Also called whenever\n Python needs an integer object (such as in slicing). Must return\n an integer (int or long).\n\n New in version 2.5.\n\nobject.__coerce__(self, other)\n\n Called to implement "mixed-mode" numeric arithmetic. Should either\n return a 2-tuple containing *self* and *other* converted to a\n common numeric type, or ``None`` if conversion is impossible. When\n the common type would be the type of ``other``, it is sufficient to\n return ``None``, since the interpreter will also ask the other\n object to attempt a coercion (but sometimes, if the implementation\n of the other type cannot be changed, it is useful to do the\n conversion to the other type here). A return value of\n ``NotImplemented`` is equivalent to returning ``None``.\n\n\nCoercion rules\n==============\n\nThis section used to document the rules for coercion. As the language\nhas evolved, the coercion rules have become hard to document\nprecisely; documenting what one version of one particular\nimplementation does is undesirable. Instead, here are some informal\nguidelines regarding coercion. In Python 3, coercion will not be\nsupported.\n\n* If the left operand of a % operator is a string or Unicode object,\n no coercion takes place and the string formatting operation is\n invoked instead.\n\n* It is no longer recommended to define a coercion operation. Mixed-\n mode operations on types that don\'t define coercion pass the\n original arguments to the operation.\n\n* New-style classes (those derived from ``object``) never invoke the\n ``__coerce__()`` method in response to a binary operator; the only\n time ``__coerce__()`` is invoked is when the built-in function\n ``coerce()`` is called.\n\n* For most intents and purposes, an operator that returns\n ``NotImplemented`` is treated the same as one that is not\n implemented at all.\n\n* Below, ``__op__()`` and ``__rop__()`` are used to signify the\n generic method names corresponding to an operator; ``__iop__()`` is\n used for the corresponding in-place operator. For example, for the\n operator \'``+``\', ``__add__()`` and ``__radd__()`` are used for the\n left and right variant of the binary operator, and ``__iadd__()``\n for the in-place variant.\n\n* For objects *x* and *y*, first ``x.__op__(y)`` is tried. If this is\n not implemented or returns ``NotImplemented``, ``y.__rop__(x)`` is\n tried. If this is also not implemented or returns\n ``NotImplemented``, a ``TypeError`` exception is raised. But see\n the following exception:\n\n* Exception to the previous item: if the left operand is an instance\n of a built-in type or a new-style class, and the right operand is an\n instance of a proper subclass of that type or class and overrides\n the base\'s ``__rop__()`` method, the right operand\'s ``__rop__()``\n method is tried *before* the left operand\'s ``__op__()`` method.\n\n This is done so that a subclass can completely override binary\n operators. Otherwise, the left operand\'s ``__op__()`` method would\n always accept the right operand: when an instance of a given class\n is expected, an instance of a subclass of that class is always\n acceptable.\n\n* When either operand type defines a coercion, this coercion is called\n before that type\'s ``__op__()`` or ``__rop__()`` method is called,\n but no sooner. If the coercion returns an object of a different\n type for the operand whose coercion is invoked, part of the process\n is redone using the new object.\n\n* When an in-place operator (like \'``+=``\') is used, if the left\n operand implements ``__iop__()``, it is invoked without any\n coercion. When the operation falls back to ``__op__()`` and/or\n ``__rop__()``, the normal coercion rules apply.\n\n* In ``x + y``, if *x* is a sequence that implements sequence\n concatenation, sequence concatenation is invoked.\n\n* In ``x * y``, if one operand is a sequence that implements sequence\n repetition, and the other is an integer (``int`` or ``long``),\n sequence repetition is invoked.\n\n* Rich comparisons (implemented by methods ``__eq__()`` and so on)\n never use coercion. Three-way comparison (implemented by\n ``__cmp__()``) does use coercion under the same conditions as other\n binary operations use it.\n\n* In the current implementation, the built-in numeric types ``int``,\n ``long``, ``float``, and ``complex`` do not use coercion. All these\n types implement a ``__coerce__()`` method, for use by the built-in\n ``coerce()`` function.\n\n Changed in version 2.7.\n\n\nWith Statement Context Managers\n===============================\n\nNew in version 2.5.\n\nA *context manager* is an object that defines the runtime context to\nbe established when executing a ``with`` statement. The context\nmanager handles the entry into, and the exit from, the desired runtime\ncontext for the execution of the block of code. Context managers are\nnormally invoked using the ``with`` statement (described in section\n*The with statement*), but can also be used by directly invoking their\nmethods.\n\nTypical uses of context managers include saving and restoring various\nkinds of global state, locking and unlocking resources, closing opened\nfiles, etc.\n\nFor more information on context managers, see *Context Manager Types*.\n\nobject.__enter__(self)\n\n Enter the runtime context related to this object. The ``with``\n statement will bind this method\'s return value to the target(s)\n specified in the ``as`` clause of the statement, if any.\n\nobject.__exit__(self, exc_type, exc_value, traceback)\n\n Exit the runtime context related to this object. The parameters\n describe the exception that caused the context to be exited. If the\n context was exited without an exception, all three arguments will\n be ``None``.\n\n If an exception is supplied, and the method wishes to suppress the\n exception (i.e., prevent it from being propagated), it should\n return a true value. Otherwise, the exception will be processed\n normally upon exit from this method.\n\n Note that ``__exit__()`` methods should not reraise the passed-in\n exception; this is the caller\'s responsibility.\n\nSee also:\n\n **PEP 0343** - The "with" statement\n The specification, background, and examples for the Python\n ``with`` statement.\n\n\nSpecial method lookup for old-style classes\n===========================================\n\nFor old-style classes, special methods are always looked up in exactly\nthe same way as any other method or attribute. This is the case\nregardless of whether the method is being looked up explicitly as in\n``x.__getitem__(i)`` or implicitly as in ``x[i]``.\n\nThis behaviour means that special methods may exhibit different\nbehaviour for different instances of a single old-style class if the\nappropriate special attributes are set differently:\n\n >>> class C:\n ... pass\n ...\n >>> c1 = C()\n >>> c2 = C()\n >>> c1.__len__ = lambda: 5\n >>> c2.__len__ = lambda: 9\n >>> len(c1)\n 5\n >>> len(c2)\n 9\n\n\nSpecial method lookup for new-style classes\n===========================================\n\nFor new-style classes, implicit invocations of special methods are\nonly guaranteed to work correctly if defined on an object\'s type, not\nin the object\'s instance dictionary. That behaviour is the reason why\nthe following code raises an exception (unlike the equivalent example\nwith old-style classes):\n\n >>> class C(object):\n ... pass\n ...\n >>> c = C()\n >>> c.__len__ = lambda: 5\n >>> len(c)\n Traceback (most recent call last):\n File "<stdin>", line 1, in <module>\n TypeError: object of type \'C\' has no len()\n\nThe rationale behind this behaviour lies with a number of special\nmethods such as ``__hash__()`` and ``__repr__()`` that are implemented\nby all objects, including type objects. If the implicit lookup of\nthese methods used the conventional lookup process, they would fail\nwhen invoked on the type object itself:\n\n >>> 1 .__hash__() == hash(1)\n True\n >>> int.__hash__() == hash(int)\n Traceback (most recent call last):\n File "<stdin>", line 1, in <module>\n TypeError: descriptor \'__hash__\' of \'int\' object needs an argument\n\nIncorrectly attempting to invoke an unbound method of a class in this\nway is sometimes referred to as \'metaclass confusion\', and is avoided\nby bypassing the instance when looking up special methods:\n\n >>> type(1).__hash__(1) == hash(1)\n True\n >>> type(int).__hash__(int) == hash(int)\n True\n\nIn addition to bypassing any instance attributes in the interest of\ncorrectness, implicit special method lookup generally also bypasses\nthe ``__getattribute__()`` method even of the object\'s metaclass:\n\n >>> class Meta(type):\n ... def __getattribute__(*args):\n ... print "Metaclass getattribute invoked"\n ... return type.__getattribute__(*args)\n ...\n >>> class C(object):\n ... __metaclass__ = Meta\n ... def __len__(self):\n ... return 10\n ... def __getattribute__(*args):\n ... print "Class getattribute invoked"\n ... return object.__getattribute__(*args)\n ...\n >>> c = C()\n >>> c.__len__() # Explicit lookup via instance\n Class getattribute invoked\n 10\n >>> type(c).__len__(c) # Explicit lookup via type\n Metaclass getattribute invoked\n 10\n >>> len(c) # Implicit lookup\n 10\n\nBypassing the ``__getattribute__()`` machinery in this fashion\nprovides significant scope for speed optimisations within the\ninterpreter, at the cost of some flexibility in the handling of\nspecial methods (the special method *must* be set on the class object\nitself in order to be consistently invoked by the interpreter).\n\n-[ Footnotes ]-\n\n[1] It *is* possible in some cases to change an object\'s type, under\n certain controlled conditions. It generally isn\'t a good idea\n though, since it can lead to some very strange behaviour if it is\n handled incorrectly.\n\n[2] For operands of the same type, it is assumed that if the non-\n reflected method (such as ``__add__()``) fails the operation is\n not supported, which is why the reflected method is not called.\n',
/prebuilts/python/darwin-x86/2.7.5/lib/python2.7/pydoc_data/
H A D	topics.py	6 'attribute-access': '\nCustomizing attribute access\n****************************\n\nThe following methods can be defined to customize the meaning of\nattribute access (use of, assignment to, or deletion of ``x.name``)\nfor class instances.\n\nobject.__getattr__(self, name)\n\n Called when an attribute lookup has not found the attribute in the\n usual places (i.e. it is not an instance attribute nor is it found\n in the class tree for ``self``). ``name`` is the attribute name.\n This method should return the (computed) attribute value or raise\n an ``AttributeError`` exception.\n\n Note that if the attribute is found through the normal mechanism,\n ``__getattr__()`` is not called. (This is an intentional asymmetry\n between ``__getattr__()`` and ``__setattr__()``.) This is done both\n for efficiency reasons and because otherwise ``__getattr__()``\n would have no way to access other attributes of the instance. Note\n that at least for instance variables, you can fake total control by\n not inserting any values in the instance attribute dictionary (but\n instead inserting them in another object). See the\n ``__getattribute__()`` method below for a way to actually get total\n control in new-style classes.\n\nobject.__setattr__(self, name, value)\n\n Called when an attribute assignment is attempted. This is called\n instead of the normal mechanism (i.e. store the value in the\n instance dictionary). *name* is the attribute name, *value* is the\n value to be assigned to it.\n\n If ``__setattr__()`` wants to assign to an instance attribute, it\n should not simply execute ``self.name = value`` --- this would\n cause a recursive call to itself. Instead, it should insert the\n value in the dictionary of instance attributes, e.g.,\n ``self.__dict__[name] = value``. For new-style classes, rather\n than accessing the instance dictionary, it should call the base\n class method with the same name, for example,\n ``object.__setattr__(self, name, value)``.\n\nobject.__delattr__(self, name)\n\n Like ``__setattr__()`` but for attribute deletion instead of\n assignment. This should only be implemented if ``del obj.name`` is\n meaningful for the object.\n\n\nMore attribute access for new-style classes\n===========================================\n\nThe following methods only apply to new-style classes.\n\nobject.__getattribute__(self, name)\n\n Called unconditionally to implement attribute accesses for\n instances of the class. If the class also defines\n ``__getattr__()``, the latter will not be called unless\n ``__getattribute__()`` either calls it explicitly or raises an\n ``AttributeError``. This method should return the (computed)\n attribute value or raise an ``AttributeError`` exception. In order\n to avoid infinite recursion in this method, its implementation\n should always call the base class method with the same name to\n access any attributes it needs, for example,\n ``object.__getattribute__(self, name)``.\n\n Note: This method may still be bypassed when looking up special methods\n as the result of implicit invocation via language syntax or\n built-in functions. See *Special method lookup for new-style\n classes*.\n\n\nImplementing Descriptors\n========================\n\nThe following methods only apply when an instance of the class\ncontaining the method (a so-called *descriptor* class) appears in an\n*owner* class (the descriptor must be in either the owner\'s class\ndictionary or in the class dictionary for one of its parents). In the\nexamples below, "the attribute" refers to the attribute whose name is\nthe key of the property in the owner class\' ``__dict__``.\n\nobject.__get__(self, instance, owner)\n\n Called to get the attribute of the owner class (class attribute\n access) or of an instance of that class (instance attribute\n access). *owner* is always the owner class, while *instance* is the\n instance that the attribute was accessed through, or ``None`` when\n the attribute is accessed through the *owner*. This method should\n return the (computed) attribute value or raise an\n ``AttributeError`` exception.\n\nobject.__set__(self, instance, value)\n\n Called to set the attribute on an instance *instance* of the owner\n class to a new value, *value*.\n\nobject.__delete__(self, instance)\n\n Called to delete the attribute on an instance *instance* of the\n owner class.\n\n\nInvoking Descriptors\n====================\n\nIn general, a descriptor is an object attribute with "binding\nbehavior", one whose attribute access has been overridden by methods\nin the descriptor protocol: ``__get__()``, ``__set__()``, and\n``__delete__()``. If any of those methods are defined for an object,\nit is said to be a descriptor.\n\nThe default behavior for attribute access is to get, set, or delete\nthe attribute from an object\'s dictionary. For instance, ``a.x`` has a\nlookup chain starting with ``a.__dict__[\'x\']``, then\n``type(a).__dict__[\'x\']``, and continuing through the base classes of\n``type(a)`` excluding metaclasses.\n\nHowever, if the looked-up value is an object defining one of the\ndescriptor methods, then Python may override the default behavior and\ninvoke the descriptor method instead. Where this occurs in the\nprecedence chain depends on which descriptor methods were defined and\nhow they were called. Note that descriptors are only invoked for new\nstyle objects or classes (ones that subclass ``object()`` or\n``type()``).\n\nThe starting point for descriptor invocation is a binding, ``a.x``.\nHow the arguments are assembled depends on ``a``:\n\nDirect Call\n The simplest and least common call is when user code directly\n invokes a descriptor method: ``x.__get__(a)``.\n\nInstance Binding\n If binding to a new-style object instance, ``a.x`` is transformed\n into the call: ``type(a).__dict__[\'x\'].__get__(a, type(a))``.\n\nClass Binding\n If binding to a new-style class, ``A.x`` is transformed into the\n call: ``A.__dict__[\'x\'].__get__(None, A)``.\n\nSuper Binding\n If ``a`` is an instance of ``super``, then the binding ``super(B,\n obj).m()`` searches ``obj.__class__.__mro__`` for the base class\n ``A`` immediately preceding ``B`` and then invokes the descriptor\n with the call: ``A.__dict__[\'m\'].__get__(obj, obj.__class__)``.\n\nFor instance bindings, the precedence of descriptor invocation depends\non the which descriptor methods are defined. A descriptor can define\nany combination of ``__get__()``, ``__set__()`` and ``__delete__()``.\nIf it does not define ``__get__()``, then accessing the attribute will\nreturn the descriptor object itself unless there is a value in the\nobject\'s instance dictionary. If the descriptor defines ``__set__()``\nand/or ``__delete__()``, it is a data descriptor; if it defines\nneither, it is a non-data descriptor. Normally, data descriptors\ndefine both ``__get__()`` and ``__set__()``, while non-data\ndescriptors have just the ``__get__()`` method. Data descriptors with\n``__set__()`` and ``__get__()`` defined always override a redefinition\nin an instance dictionary. In contrast, non-data descriptors can be\noverridden by instances.\n\nPython methods (including ``staticmethod()`` and ``classmethod()``)\nare implemented as non-data descriptors. Accordingly, instances can\nredefine and override methods. This allows individual instances to\nacquire behaviors that differ from other instances of the same class.\n\nThe ``property()`` function is implemented as a data descriptor.\nAccordingly, instances cannot override the behavior of a property.\n\n\n__slots__\n=========\n\nBy default, instances of both old and new-style classes have a\ndictionary for attribute storage. This wastes space for objects\nhaving very few instance variables. The space consumption can become\nacute when creating large numbers of instances.\n\nThe default can be overridden by defining *__slots__* in a new-style\nclass definition. The *__slots__* declaration takes a sequence of\ninstance variables and reserves just enough space in each instance to\nhold a value for each variable. Space is saved because *__dict__* is\nnot created for each instance.\n\n__slots__\n\n This class variable can be assigned a string, iterable, or sequence\n of strings with variable names used by instances. If defined in a\n new-style class, *__slots__* reserves space for the declared\n variables and prevents the automatic creation of *__dict__* and\n *__weakref__* for each instance.\n\n New in version 2.2.\n\nNotes on using *__slots__*\n\n* When inheriting from a class without *__slots__*, the *__dict__*\n attribute of that class will always be accessible, so a *__slots__*\n definition in the subclass is meaningless.\n\n* Without a *__dict__* variable, instances cannot be assigned new\n variables not listed in the *__slots__* definition. Attempts to\n assign to an unlisted variable name raises ``AttributeError``. If\n dynamic assignment of new variables is desired, then add\n ``\'__dict__\'`` to the sequence of strings in the *__slots__*\n declaration.\n\n Changed in version 2.3: Previously, adding ``\'__dict__\'`` to the\n *__slots__* declaration would not enable the assignment of new\n attributes not specifically listed in the sequence of instance\n variable names.\n\n* Without a *__weakref__* variable for each instance, classes defining\n *__slots__* do not support weak references to its instances. If weak\n reference support is needed, then add ``\'__weakref__\'`` to the\n sequence of strings in the *__slots__* declaration.\n\n Changed in version 2.3: Previously, adding ``\'__weakref__\'`` to the\n *__slots__* declaration would not enable support for weak\n references.\n\n* *__slots__* are implemented at the class level by creating\n descriptors (*Implementing Descriptors*) for each variable name. As\n a result, class attributes cannot be used to set default values for\n instance variables defined by *__slots__*; otherwise, the class\n attribute would overwrite the descriptor assignment.\n\n* The action of a *__slots__* declaration is limited to the class\n where it is defined. As a result, subclasses will have a *__dict__*\n unless they also define *__slots__* (which must only contain names\n of any *additional* slots).\n\n* If a class defines a slot also defined in a base class, the instance\n variable defined by the base class slot is inaccessible (except by\n retrieving its descriptor directly from the base class). This\n renders the meaning of the program undefined. In the future, a\n check may be added to prevent this.\n\n* Nonempty *__slots__* does not work for classes derived from\n "variable-length" built-in types such as ``long``, ``str`` and\n ``tuple``.\n\n* Any non-string iterable may be assigned to *__slots__*. Mappings may\n also be used; however, in the future, special meaning may be\n assigned to the values corresponding to each key.\n\n* *__class__* assignment works only if both classes have the same\n *__slots__*.\n\n Changed in version 2.6: Previously, *__class__* assignment raised an\n error if either new or old class had *__slots__*.\n', 18 'calls': '\nCalls\n*****\n\nA call calls a callable object (e.g., a *function*) with a possibly\nempty series of *arguments*:\n\n call ::= primary "(" [argument_list [","]\n | expression genexpr_for] ")"\n argument_list ::= positional_arguments ["," keyword_arguments]\n ["," "*" expression] ["," keyword_arguments]\n ["," "**" expression]\n | keyword_arguments ["," "*" expression]\n ["," "**" expression]\n | "*" expression ["," "*" expression] ["," "**" expression]\n | "**" expression\n positional_arguments ::= expression ("," expression)*\n keyword_arguments ::= keyword_item ("," keyword_item)*\n keyword_item ::= identifier "=" expression\n\nA trailing comma may be present after the positional and keyword\narguments but does not affect the semantics.\n\nThe primary must evaluate to a callable object (user-defined\nfunctions, built-in functions, methods of built-in objects, class\nobjects, methods of class instances, and certain class instances\nthemselves are callable; extensions may define additional callable\nobject types). All argument expressions are evaluated before the call\nis attempted. Please refer to section *Function definitions* for the\nsyntax of formal *parameter* lists.\n\nIf keyword arguments are present, they are first converted to\npositional arguments, as follows. First, a list of unfilled slots is\ncreated for the formal parameters. If there are N positional\narguments, they are placed in the first N slots. Next, for each\nkeyword argument, the identifier is used to determine the\ncorresponding slot (if the identifier is the same as the first formal\nparameter name, the first slot is used, and so on). If the slot is\nalready filled, a ``TypeError`` exception is raised. Otherwise, the\nvalue of the argument is placed in the slot, filling it (even if the\nexpression is ``None``, it fills the slot). When all arguments have\nbeen processed, the slots that are still unfilled are filled with the\ncorresponding default value from the function definition. (Default\nvalues are calculated, once, when the function is defined; thus, a\nmutable object such as a list or dictionary used as default value will\nbe shared by all calls that don\'t specify an argument value for the\ncorresponding slot; this should usually be avoided.) If there are any\nunfilled slots for which no default value is specified, a\n``TypeError`` exception is raised. Otherwise, the list of filled\nslots is used as the argument list for the call.\n\n**CPython implementation detail:** An implementation may provide\nbuilt-in functions whose positional parameters do not have names, even\nif they are \'named\' for the purpose of documentation, and which\ntherefore cannot be supplied by keyword. In CPython, this is the case\nfor functions implemented in C that use ``PyArg_ParseTuple()`` to\nparse their arguments.\n\nIf there are more positional arguments than there are formal parameter\nslots, a ``TypeError`` exception is raised, unless a formal parameter\nusing the syntax ``*identifier`` is present; in this case, that formal\nparameter receives a tuple containing the excess positional arguments\n(or an empty tuple if there were no excess positional arguments).\n\nIf any keyword argument does not correspond to a formal parameter\nname, a ``TypeError`` exception is raised, unless a formal parameter\nusing the syntax ``**identifier`` is present; in this case, that\nformal parameter receives a dictionary containing the excess keyword\narguments (using the keywords as keys and the argument values as\ncorresponding values), or a (new) empty dictionary if there were no\nexcess keyword arguments.\n\nIf the syntax ``*expression`` appears in the function call,\n``expression`` must evaluate to an iterable. Elements from this\niterable are treated as if they were additional positional arguments;\nif there are positional arguments *x1*, ..., *xN*, and ``expression``\nevaluates to a sequence *y1*, ..., *yM*, this is equivalent to a call\nwith M+N positional arguments *x1*, ..., *xN*, *y1*, ..., *yM*.\n\nA consequence of this is that although the ``*expression`` syntax may\nappear *after* some keyword arguments, it is processed *before* the\nkeyword arguments (and the ``**expression`` argument, if any -- see\nbelow). So:\n\n >>> def f(a, b):\n ... print a, b\n ...\n >>> f(b=1, *(2,))\n 2 1\n >>> f(a=1, *(2,))\n Traceback (most recent call last):\n File "<stdin>", line 1, in ?\n TypeError: f() got multiple values for keyword argument \'a\'\n >>> f(1, *(2,))\n 1 2\n\nIt is unusual for both keyword arguments and the ``*expression``\nsyntax to be used in the same call, so in practice this confusion does\nnot arise.\n\nIf the syntax ``**expression`` appears in the function call,\n``expression`` must evaluate to a mapping, the contents of which are\ntreated as additional keyword arguments. In the case of a keyword\nappearing in both ``expression`` and as an explicit keyword argument,\na ``TypeError`` exception is raised.\n\nFormal parameters using the syntax ``*identifier`` or ``**identifier``\ncannot be used as positional argument slots or as keyword argument\nnames. Formal parameters using the syntax ``(sublist)`` cannot be\nused as keyword argument names; the outermost sublist corresponds to a\nsingle unnamed argument slot, and the argument value is assigned to\nthe sublist using the usual tuple assignment rules after all other\nparameter processing is done.\n\nA call always returns some value, possibly ``None``, unless it raises\nan exception. How this value is computed depends on the type of the\ncallable object.\n\nIf it is---\n\na user-defined function:\n The code block for the function is executed, passing it the\n argument list. The first thing the code block will do is bind the\n formal parameters to the arguments; this is described in section\n *Function definitions*. When the code block executes a ``return``\n statement, this specifies the return value of the function call.\n\na built-in function or method:\n The result is up to the interpreter; see *Built-in Functions* for\n the descriptions of built-in functions and methods.\n\na class object:\n A new instance of that class is returned.\n\na class instance method:\n The corresponding user-defined function is called, with an argument\n list that is one longer than the argument list of the call: the\n instance becomes the first argument.\n\na class instance:\n The class must define a ``__call__()`` method; the effect is then\n the same as if that method was called.\n', 61 'specialnames': '\nSpecial method names\n********************\n\nA class can implement certain operations that are invoked by special\nsyntax (such as arithmetic operations or subscripting and slicing) by\ndefining methods with special names. This is Python\'s approach to\n*operator overloading*, allowing classes to define their own behavior\nwith respect to language operators. For instance, if a class defines\na method named ``__getitem__()``, and ``x`` is an instance of this\nclass, then ``x[i]`` is roughly equivalent to ``x.__getitem__(i)`` for\nold-style classes and ``type(x).__getitem__(x, i)`` for new-style\nclasses. Except where mentioned, attempts to execute an operation\nraise an exception when no appropriate method is defined (typically\n``AttributeError`` or ``TypeError``).\n\nWhen implementing a class that emulates any built-in type, it is\nimportant that the emulation only be implemented to the degree that it\nmakes sense for the object being modelled. For example, some\nsequences may work well with retrieval of individual elements, but\nextracting a slice may not make sense. (One example of this is the\n``NodeList`` interface in the W3C\'s Document Object Model.)\n\n\nBasic customization\n===================\n\nobject.__new__(cls[, ...])\n\n Called to create a new instance of class *cls*. ``__new__()`` is a\n static method (special-cased so you need not declare it as such)\n that takes the class of which an instance was requested as its\n first argument. The remaining arguments are those passed to the\n object constructor expression (the call to the class). The return\n value of ``__new__()`` should be the new object instance (usually\n an instance of *cls*).\n\n Typical implementations create a new instance of the class by\n invoking the superclass\'s ``__new__()`` method using\n ``super(currentclass, cls).__new__(cls[, ...])`` with appropriate\n arguments and then modifying the newly-created instance as\n necessary before returning it.\n\n If ``__new__()`` returns an instance of *cls*, then the new\n instance\'s ``__init__()`` method will be invoked like\n ``__init__(self[, ...])``, where *self* is the new instance and the\n remaining arguments are the same as were passed to ``__new__()``.\n\n If ``__new__()`` does not return an instance of *cls*, then the new\n instance\'s ``__init__()`` method will not be invoked.\n\n ``__new__()`` is intended mainly to allow subclasses of immutable\n types (like int, str, or tuple) to customize instance creation. It\n is also commonly overridden in custom metaclasses in order to\n customize class creation.\n\nobject.__init__(self[, ...])\n\n Called when the instance is created. The arguments are those\n passed to the class constructor expression. If a base class has an\n ``__init__()`` method, the derived class\'s ``__init__()`` method,\n if any, must explicitly call it to ensure proper initialization of\n the base class part of the instance; for example:\n ``BaseClass.__init__(self, [args...])``. As a special constraint\n on constructors, no value may be returned; doing so will cause a\n ``TypeError`` to be raised at runtime.\n\nobject.__del__(self)\n\n Called when the instance is about to be destroyed. This is also\n called a destructor. If a base class has a ``__del__()`` method,\n the derived class\'s ``__del__()`` method, if any, must explicitly\n call it to ensure proper deletion of the base class part of the\n instance. Note that it is possible (though not recommended!) for\n the ``__del__()`` method to postpone destruction of the instance by\n creating a new reference to it. It may then be called at a later\n time when this new reference is deleted. It is not guaranteed that\n ``__del__()`` methods are called for objects that still exist when\n the interpreter exits.\n\n Note: ``del x`` doesn\'t directly call ``x.__del__()`` --- the former\n decrements the reference count for ``x`` by one, and the latter\n is only called when ``x``\'s reference count reaches zero. Some\n common situations that may prevent the reference count of an\n object from going to zero include: circular references between\n objects (e.g., a doubly-linked list or a tree data structure with\n parent and child pointers); a reference to the object on the\n stack frame of a function that caught an exception (the traceback\n stored in ``sys.exc_traceback`` keeps the stack frame alive); or\n a reference to the object on the stack frame that raised an\n unhandled exception in interactive mode (the traceback stored in\n ``sys.last_traceback`` keeps the stack frame alive). The first\n situation can only be remedied by explicitly breaking the cycles;\n the latter two situations can be resolved by storing ``None`` in\n ``sys.exc_traceback`` or ``sys.last_traceback``. Circular\n references which are garbage are detected when the option cycle\n detector is enabled (it\'s on by default), but can only be cleaned\n up if there are no Python-level ``__del__()`` methods involved.\n Refer to the documentation for the ``gc`` module for more\n information about how ``__del__()`` methods are handled by the\n cycle detector, particularly the description of the ``garbage``\n value.\n\n Warning: Due to the precarious circumstances under which ``__del__()``\n methods are invoked, exceptions that occur during their execution\n are ignored, and a warning is printed to ``sys.stderr`` instead.\n Also, when ``__del__()`` is invoked in response to a module being\n deleted (e.g., when execution of the program is done), other\n globals referenced by the ``__del__()`` method may already have\n been deleted or in the process of being torn down (e.g. the\n import machinery shutting down). For this reason, ``__del__()``\n methods should do the absolute minimum needed to maintain\n external invariants. Starting with version 1.5, Python\n guarantees that globals whose name begins with a single\n underscore are deleted from their module before other globals are\n deleted; if no other references to such globals exist, this may\n help in assuring that imported modules are still available at the\n time when the ``__del__()`` method is called.\n\n See also the *-R* command-line option.\n\nobject.__repr__(self)\n\n Called by the ``repr()`` built-in function and by string\n conversions (reverse quotes) to compute the "official" string\n representation of an object. If at all possible, this should look\n like a valid Python expression that could be used to recreate an\n object with the same value (given an appropriate environment). If\n this is not possible, a string of the form ``<...some useful\n description...>`` should be returned. The return value must be a\n string object. If a class defines ``__repr__()`` but not\n ``__str__()``, then ``__repr__()`` is also used when an "informal"\n string representation of instances of that class is required.\n\n This is typically used for debugging, so it is important that the\n representation is information-rich and unambiguous.\n\nobject.__str__(self)\n\n Called by the ``str()`` built-in function and by the ``print``\n statement to compute the "informal" string representation of an\n object. This differs from ``__repr__()`` in that it does not have\n to be a valid Python expression: a more convenient or concise\n representation may be used instead. The return value must be a\n string object.\n\nobject.__lt__(self, other)\nobject.__le__(self, other)\nobject.__eq__(self, other)\nobject.__ne__(self, other)\nobject.__gt__(self, other)\nobject.__ge__(self, other)\n\n New in version 2.1.\n\n These are the so-called "rich comparison" methods, and are called\n for comparison operators in preference to ``__cmp__()`` below. The\n correspondence between operator symbols and method names is as\n follows: ``x<y`` calls ``x.__lt__(y)``, ``x<=y`` calls\n ``x.__le__(y)``, ``x==y`` calls ``x.__eq__(y)``, ``x!=y`` and\n ``x<>y`` call ``x.__ne__(y)``, ``x>y`` calls ``x.__gt__(y)``, and\n ``x>=y`` calls ``x.__ge__(y)``.\n\n A rich comparison method may return the singleton\n ``NotImplemented`` if it does not implement the operation for a\n given pair of arguments. By convention, ``False`` and ``True`` are\n returned for a successful comparison. However, these methods can\n return any value, so if the comparison operator is used in a\n Boolean context (e.g., in the condition of an ``if`` statement),\n Python will call ``bool()`` on the value to determine if the result\n is true or false.\n\n There are no implied relationships among the comparison operators.\n The truth of ``x==y`` does not imply that ``x!=y`` is false.\n Accordingly, when defining ``__eq__()``, one should also define\n ``__ne__()`` so that the operators will behave as expected. See\n the paragraph on ``__hash__()`` for some important notes on\n creating *hashable* objects which support custom comparison\n operations and are usable as dictionary keys.\n\n There are no swapped-argument versions of these methods (to be used\n when the left argument does not support the operation but the right\n argument does); rather, ``__lt__()`` and ``__gt__()`` are each\n other\'s reflection, ``__le__()`` and ``__ge__()`` are each other\'s\n reflection, and ``__eq__()`` and ``__ne__()`` are their own\n reflection.\n\n Arguments to rich comparison methods are never coerced.\n\n To automatically generate ordering operations from a single root\n operation, see ``functools.total_ordering()``.\n\nobject.__cmp__(self, other)\n\n Called by comparison operations if rich comparison (see above) is\n not defined. Should return a negative integer if ``self < other``,\n zero if ``self == other``, a positive integer if ``self > other``.\n If no ``__cmp__()``, ``__eq__()`` or ``__ne__()`` operation is\n defined, class instances are compared by object identity\n ("address"). See also the description of ``__hash__()`` for some\n important notes on creating *hashable* objects which support custom\n comparison operations and are usable as dictionary keys. (Note: the\n restriction that exceptions are not propagated by ``__cmp__()`` has\n been removed since Python 1.5.)\n\nobject.__rcmp__(self, other)\n\n Changed in version 2.1: No longer supported.\n\nobject.__hash__(self)\n\n Called by built-in function ``hash()`` and for operations on\n members of hashed collections including ``set``, ``frozenset``, and\n ``dict``. ``__hash__()`` should return an integer. The only\n required property is that objects which compare equal have the same\n hash value; it is advised to somehow mix together (e.g. using\n exclusive or) the hash values for the components of the object that\n also play a part in comparison of objects.\n\n If a class does not define a ``__cmp__()`` or ``__eq__()`` method\n it should not define a ``__hash__()`` operation either; if it\n defines ``__cmp__()`` or ``__eq__()`` but not ``__hash__()``, its\n instances will not be usable in hashed collections. If a class\n defines mutable objects and implements a ``__cmp__()`` or\n ``__eq__()`` method, it should not implement ``__hash__()``, since\n hashable collection implementations require that a object\'s hash\n value is immutable (if the object\'s hash value changes, it will be\n in the wrong hash bucket).\n\n User-defined classes have ``__cmp__()`` and ``__hash__()`` methods\n by default; with them, all objects compare unequal (except with\n themselves) and ``x.__hash__()`` returns ``id(x)``.\n\n Classes which inherit a ``__hash__()`` method from a parent class\n but change the meaning of ``__cmp__()`` or ``__eq__()`` such that\n the hash value returned is no longer appropriate (e.g. by switching\n to a value-based concept of equality instead of the default\n identity based equality) can explicitly flag themselves as being\n unhashable by setting ``__hash__ = None`` in the class definition.\n Doing so means that not only will instances of the class raise an\n appropriate ``TypeError`` when a program attempts to retrieve their\n hash value, but they will also be correctly identified as\n unhashable when checking ``isinstance(obj, collections.Hashable)``\n (unlike classes which define their own ``__hash__()`` to explicitly\n raise ``TypeError``).\n\n Changed in version 2.5: ``__hash__()`` may now also return a long\n integer object; the 32-bit integer is then derived from the hash of\n that object.\n\n Changed in version 2.6: ``__hash__`` may now be set to ``None`` to\n explicitly flag instances of a class as unhashable.\n\nobject.__nonzero__(self)\n\n Called to implement truth value testing and the built-in operation\n ``bool()``; should return ``False`` or ``True``, or their integer\n equivalents ``0`` or ``1``. When this method is not defined,\n ``__len__()`` is called, if it is defined, and the object is\n considered true if its result is nonzero. If a class defines\n neither ``__len__()`` nor ``__nonzero__()``, all its instances are\n considered true.\n\nobject.__unicode__(self)\n\n Called to implement ``unicode()`` built-in; should return a Unicode\n object. When this method is not defined, string conversion is\n attempted, and the result of string conversion is converted to\n Unicode using the system default encoding.\n\n\nCustomizing attribute access\n============================\n\nThe following methods can be defined to customize the meaning of\nattribute access (use of, assignment to, or deletion of ``x.name``)\nfor class instances.\n\nobject.__getattr__(self, name)\n\n Called when an attribute lookup has not found the attribute in the\n usual places (i.e. it is not an instance attribute nor is it found\n in the class tree for ``self``). ``name`` is the attribute name.\n This method should return the (computed) attribute value or raise\n an ``AttributeError`` exception.\n\n Note that if the attribute is found through the normal mechanism,\n ``__getattr__()`` is not called. (This is an intentional asymmetry\n between ``__getattr__()`` and ``__setattr__()``.) This is done both\n for efficiency reasons and because otherwise ``__getattr__()``\n would have no way to access other attributes of the instance. Note\n that at least for instance variables, you can fake total control by\n not inserting any values in the instance attribute dictionary (but\n instead inserting them in another object). See the\n ``__getattribute__()`` method below for a way to actually get total\n control in new-style classes.\n\nobject.__setattr__(self, name, value)\n\n Called when an attribute assignment is attempted. This is called\n instead of the normal mechanism (i.e. store the value in the\n instance dictionary). *name* is the attribute name, *value* is the\n value to be assigned to it.\n\n If ``__setattr__()`` wants to assign to an instance attribute, it\n should not simply execute ``self.name = value`` --- this would\n cause a recursive call to itself. Instead, it should insert the\n value in the dictionary of instance attributes, e.g.,\n ``self.__dict__[name] = value``. For new-style classes, rather\n than accessing the instance dictionary, it should call the base\n class method with the same name, for example,\n ``object.__setattr__(self, name, value)``.\n\nobject.__delattr__(self, name)\n\n Like ``__setattr__()`` but for attribute deletion instead of\n assignment. This should only be implemented if ``del obj.name`` is\n meaningful for the object.\n\n\nMore attribute access for new-style classes\n-------------------------------------------\n\nThe following methods only apply to new-style classes.\n\nobject.__getattribute__(self, name)\n\n Called unconditionally to implement attribute accesses for\n instances of the class. If the class also defines\n ``__getattr__()``, the latter will not be called unless\n ``__getattribute__()`` either calls it explicitly or raises an\n ``AttributeError``. This method should return the (computed)\n attribute value or raise an ``AttributeError`` exception. In order\n to avoid infinite recursion in this method, its implementation\n should always call the base class method with the same name to\n access any attributes it needs, for example,\n ``object.__getattribute__(self, name)``.\n\n Note: This method may still be bypassed when looking up special methods\n as the result of implicit invocation via language syntax or\n built-in functions. See *Special method lookup for new-style\n classes*.\n\n\nImplementing Descriptors\n------------------------\n\nThe following methods only apply when an instance of the class\ncontaining the method (a so-called *descriptor* class) appears in an\n*owner* class (the descriptor must be in either the owner\'s class\ndictionary or in the class dictionary for one of its parents). In the\nexamples below, "the attribute" refers to the attribute whose name is\nthe key of the property in the owner class\' ``__dict__``.\n\nobject.__get__(self, instance, owner)\n\n Called to get the attribute of the owner class (class attribute\n access) or of an instance of that class (instance attribute\n access). *owner* is always the owner class, while *instance* is the\n instance that the attribute was accessed through, or ``None`` when\n the attribute is accessed through the *owner*. This method should\n return the (computed) attribute value or raise an\n ``AttributeError`` exception.\n\nobject.__set__(self, instance, value)\n\n Called to set the attribute on an instance *instance* of the owner\n class to a new value, *value*.\n\nobject.__delete__(self, instance)\n\n Called to delete the attribute on an instance *instance* of the\n owner class.\n\n\nInvoking Descriptors\n--------------------\n\nIn general, a descriptor is an object attribute with "binding\nbehavior", one whose attribute access has been overridden by methods\nin the descriptor protocol: ``__get__()``, ``__set__()``, and\n``__delete__()``. If any of those methods are defined for an object,\nit is said to be a descriptor.\n\nThe default behavior for attribute access is to get, set, or delete\nthe attribute from an object\'s dictionary. For instance, ``a.x`` has a\nlookup chain starting with ``a.__dict__[\'x\']``, then\n``type(a).__dict__[\'x\']``, and continuing through the base classes of\n``type(a)`` excluding metaclasses.\n\nHowever, if the looked-up value is an object defining one of the\ndescriptor methods, then Python may override the default behavior and\ninvoke the descriptor method instead. Where this occurs in the\nprecedence chain depends on which descriptor methods were defined and\nhow they were called. Note that descriptors are only invoked for new\nstyle objects or classes (ones that subclass ``object()`` or\n``type()``).\n\nThe starting point for descriptor invocation is a binding, ``a.x``.\nHow the arguments are assembled depends on ``a``:\n\nDirect Call\n The simplest and least common call is when user code directly\n invokes a descriptor method: ``x.__get__(a)``.\n\nInstance Binding\n If binding to a new-style object instance, ``a.x`` is transformed\n into the call: ``type(a).__dict__[\'x\'].__get__(a, type(a))``.\n\nClass Binding\n If binding to a new-style class, ``A.x`` is transformed into the\n call: ``A.__dict__[\'x\'].__get__(None, A)``.\n\nSuper Binding\n If ``a`` is an instance of ``super``, then the binding ``super(B,\n obj).m()`` searches ``obj.__class__.__mro__`` for the base class\n ``A`` immediately preceding ``B`` and then invokes the descriptor\n with the call: ``A.__dict__[\'m\'].__get__(obj, obj.__class__)``.\n\nFor instance bindings, the precedence of descriptor invocation depends\non the which descriptor methods are defined. A descriptor can define\nany combination of ``__get__()``, ``__set__()`` and ``__delete__()``.\nIf it does not define ``__get__()``, then accessing the attribute will\nreturn the descriptor object itself unless there is a value in the\nobject\'s instance dictionary. If the descriptor defines ``__set__()``\nand/or ``__delete__()``, it is a data descriptor; if it defines\nneither, it is a non-data descriptor. Normally, data descriptors\ndefine both ``__get__()`` and ``__set__()``, while non-data\ndescriptors have just the ``__get__()`` method. Data descriptors with\n``__set__()`` and ``__get__()`` defined always override a redefinition\nin an instance dictionary. In contrast, non-data descriptors can be\noverridden by instances.\n\nPython methods (including ``staticmethod()`` and ``classmethod()``)\nare implemented as non-data descriptors. Accordingly, instances can\nredefine and override methods. This allows individual instances to\nacquire behaviors that differ from other instances of the same class.\n\nThe ``property()`` function is implemented as a data descriptor.\nAccordingly, instances cannot override the behavior of a property.\n\n\n__slots__\n---------\n\nBy default, instances of both old and new-style classes have a\ndictionary for attribute storage. This wastes space for objects\nhaving very few instance variables. The space consumption can become\nacute when creating large numbers of instances.\n\nThe default can be overridden by defining *__slots__* in a new-style\nclass definition. The *__slots__* declaration takes a sequence of\ninstance variables and reserves just enough space in each instance to\nhold a value for each variable. Space is saved because *__dict__* is\nnot created for each instance.\n\n__slots__\n\n This class variable can be assigned a string, iterable, or sequence\n of strings with variable names used by instances. If defined in a\n new-style class, *__slots__* reserves space for the declared\n variables and prevents the automatic creation of *__dict__* and\n *__weakref__* for each instance.\n\n New in version 2.2.\n\nNotes on using *__slots__*\n\n* When inheriting from a class without *__slots__*, the *__dict__*\n attribute of that class will always be accessible, so a *__slots__*\n definition in the subclass is meaningless.\n\n* Without a *__dict__* variable, instances cannot be assigned new\n variables not listed in the *__slots__* definition. Attempts to\n assign to an unlisted variable name raises ``AttributeError``. If\n dynamic assignment of new variables is desired, then add\n ``\'__dict__\'`` to the sequence of strings in the *__slots__*\n declaration.\n\n Changed in version 2.3: Previously, adding ``\'__dict__\'`` to the\n *__slots__* declaration would not enable the assignment of new\n attributes not specifically listed in the sequence of instance\n variable names.\n\n* Without a *__weakref__* variable for each instance, classes defining\n *__slots__* do not support weak references to its instances. If weak\n reference support is needed, then add ``\'__weakref__\'`` to the\n sequence of strings in the *__slots__* declaration.\n\n Changed in version 2.3: Previously, adding ``\'__weakref__\'`` to the\n *__slots__* declaration would not enable support for weak\n references.\n\n* *__slots__* are implemented at the class level by creating\n descriptors (*Implementing Descriptors*) for each variable name. As\n a result, class attributes cannot be used to set default values for\n instance variables defined by *__slots__*; otherwise, the class\n attribute would overwrite the descriptor assignment.\n\n* The action of a *__slots__* declaration is limited to the class\n where it is defined. As a result, subclasses will have a *__dict__*\n unless they also define *__slots__* (which must only contain names\n of any *additional* slots).\n\n* If a class defines a slot also defined in a base class, the instance\n variable defined by the base class slot is inaccessible (except by\n retrieving its descriptor directly from the base class). This\n renders the meaning of the program undefined. In the future, a\n check may be added to prevent this.\n\n* Nonempty *__slots__* does not work for classes derived from\n "variable-length" built-in types such as ``long``, ``str`` and\n ``tuple``.\n\n* Any non-string iterable may be assigned to *__slots__*. Mappings may\n also be used; however, in the future, special meaning may be\n assigned to the values corresponding to each key.\n\n* *__class__* assignment works only if both classes have the same\n *__slots__*.\n\n Changed in version 2.6: Previously, *__class__* assignment raised an\n error if either new or old class had *__slots__*.\n\n\nCustomizing class creation\n==========================\n\nBy default, new-style classes are constructed using ``type()``. A\nclass definition is read into a separate namespace and the value of\nclass name is bound to the result of ``type(name, bases, dict)``.\n\nWhen the class definition is read, if *__metaclass__* is defined then\nthe callable assigned to it will be called instead of ``type()``. This\nallows classes or functions to be written which monitor or alter the\nclass creation process:\n\n* Modifying the class dictionary prior to the class being created.\n\n* Returning an instance of another class -- essentially performing the\n role of a factory function.\n\nThese steps will have to be performed in the metaclass\'s ``__new__()``\nmethod -- ``type.__new__()`` can then be called from this method to\ncreate a class with different properties. This example adds a new\nelement to the class dictionary before creating the class:\n\n class metacls(type):\n def __new__(mcs, name, bases, dict):\n dict[\'foo\'] = \'metacls was here\'\n return type.__new__(mcs, name, bases, dict)\n\nYou can of course also override other class methods (or add new\nmethods); for example defining a custom ``__call__()`` method in the\nmetaclass allows custom behavior when the class is called, e.g. not\nalways creating a new instance.\n\n__metaclass__\n\n This variable can be any callable accepting arguments for ``name``,\n ``bases``, and ``dict``. Upon class creation, the callable is used\n instead of the built-in ``type()``.\n\n New in version 2.2.\n\nThe appropriate metaclass is determined by the following precedence\nrules:\n\n* If ``dict[\'__metaclass__\']`` exists, it is used.\n\n* Otherwise, if there is at least one base class, its metaclass is\n used (this looks for a *__class__* attribute first and if not found,\n uses its type).\n\n* Otherwise, if a global variable named __metaclass__ exists, it is\n used.\n\n* Otherwise, the old-style, classic metaclass (types.ClassType) is\n used.\n\nThe potential uses for metaclasses are boundless. Some ideas that have\nbeen explored including logging, interface checking, automatic\ndelegation, automatic property creation, proxies, frameworks, and\nautomatic resource locking/synchronization.\n\n\nCustomizing instance and subclass checks\n========================================\n\nNew in version 2.6.\n\nThe following methods are used to override the default behavior of the\n``isinstance()`` and ``issubclass()`` built-in functions.\n\nIn particular, the metaclass ``abc.ABCMeta`` implements these methods\nin order to allow the addition of Abstract Base Classes (ABCs) as\n"virtual base classes" to any class or type (including built-in\ntypes), including other ABCs.\n\nclass.__instancecheck__(self, instance)\n\n Return true if *instance* should be considered a (direct or\n indirect) instance of *class*. If defined, called to implement\n ``isinstance(instance, class)``.\n\nclass.__subclasscheck__(self, subclass)\n\n Return true if *subclass* should be considered a (direct or\n indirect) subclass of *class*. If defined, called to implement\n ``issubclass(subclass, class)``.\n\nNote that these methods are looked up on the type (metaclass) of a\nclass. They cannot be defined as class methods in the actual class.\nThis is consistent with the lookup of special methods that are called\non instances, only in this case the instance is itself a class.\n\nSee also:\n\n **PEP 3119** - Introducing Abstract Base Classes\n Includes the specification for customizing ``isinstance()`` and\n ``issubclass()`` behavior through ``__instancecheck__()`` and\n ``__subclasscheck__()``, with motivation for this functionality\n in the context of adding Abstract Base Classes (see the ``abc``\n module) to the language.\n\n\nEmulating callable objects\n==========================\n\nobject.__call__(self[, args...])\n\n Called when the instance is "called" as a function; if this method\n is defined, ``x(arg1, arg2, ...)`` is a shorthand for\n ``x.__call__(arg1, arg2, ...)``.\n\n\nEmulating container types\n=========================\n\nThe following methods can be defined to implement container objects.\nContainers usually are sequences (such as lists or tuples) or mappings\n(like dictionaries), but can represent other containers as well. The\nfirst set of methods is used either to emulate a sequence or to\nemulate a mapping; the difference is that for a sequence, the\nallowable keys should be the integers *k* for which ``0 <= k < N``\nwhere *N* is the length of the sequence, or slice objects, which\ndefine a range of items. (For backwards compatibility, the method\n``__getslice__()`` (see below) can also be defined to handle simple,\nbut not extended slices.) It is also recommended that mappings provide\nthe methods ``keys()``, ``values()``, ``items()``, ``has_key()``,\n``get()``, ``clear()``, ``setdefault()``, ``iterkeys()``,\n``itervalues()``, ``iteritems()``, ``pop()``, ``popitem()``,\n``copy()``, and ``update()`` behaving similar to those for Python\'s\nstandard dictionary objects. The ``UserDict`` module provides a\n``DictMixin`` class to help create those methods from a base set of\n``__getitem__()``, ``__setitem__()``, ``__delitem__()``, and\n``keys()``. Mutable sequences should provide methods ``append()``,\n``count()``, ``index()``, ``extend()``, ``insert()``, ``pop()``,\n``remove()``, ``reverse()`` and ``sort()``, like Python standard list\nobjects. Finally, sequence types should implement addition (meaning\nconcatenation) and multiplication (meaning repetition) by defining the\nmethods ``__add__()``, ``__radd__()``, ``__iadd__()``, ``__mul__()``,\n``__rmul__()`` and ``__imul__()`` described below; they should not\ndefine ``__coerce__()`` or other numerical operators. It is\nrecommended that both mappings and sequences implement the\n``__contains__()`` method to allow efficient use of the ``in``\noperator; for mappings, ``in`` should be equivalent of ``has_key()``;\nfor sequences, it should search through the values. It is further\nrecommended that both mappings and sequences implement the\n``__iter__()`` method to allow efficient iteration through the\ncontainer; for mappings, ``__iter__()`` should be the same as\n``iterkeys()``; for sequences, it should iterate through the values.\n\nobject.__len__(self)\n\n Called to implement the built-in function ``len()``. Should return\n the length of the object, an integer ``>=`` 0. Also, an object\n that doesn\'t define a ``__nonzero__()`` method and whose\n ``__len__()`` method returns zero is considered to be false in a\n Boolean context.\n\nobject.__getitem__(self, key)\n\n Called to implement evaluation of ``self[key]``. For sequence\n types, the accepted keys should be integers and slice objects.\n Note that the special interpretation of negative indexes (if the\n class wishes to emulate a sequence type) is up to the\n ``__getitem__()`` method. If *key* is of an inappropriate type,\n ``TypeError`` may be raised; if of a value outside the set of\n indexes for the sequence (after any special interpretation of\n negative values), ``IndexError`` should be raised. For mapping\n types, if *key* is missing (not in the container), ``KeyError``\n should be raised.\n\n Note: ``for`` loops expect that an ``IndexError`` will be raised for\n illegal indexes to allow proper detection of the end of the\n sequence.\n\nobject.__setitem__(self, key, value)\n\n Called to implement assignment to ``self[key]``. Same note as for\n ``__getitem__()``. This should only be implemented for mappings if\n the objects support changes to the values for keys, or if new keys\n can be added, or for sequences if elements can be replaced. The\n same exceptions should be raised for improper *key* values as for\n the ``__getitem__()`` method.\n\nobject.__delitem__(self, key)\n\n Called to implement deletion of ``self[key]``. Same note as for\n ``__getitem__()``. This should only be implemented for mappings if\n the objects support removal of keys, or for sequences if elements\n can be removed from the sequence. The same exceptions should be\n raised for improper *key* values as for the ``__getitem__()``\n method.\n\nobject.__iter__(self)\n\n This method is called when an iterator is required for a container.\n This method should return a new iterator object that can iterate\n over all the objects in the container. For mappings, it should\n iterate over the keys of the container, and should also be made\n available as the method ``iterkeys()``.\n\n Iterator objects also need to implement this method; they are\n required to return themselves. For more information on iterator\n objects, see *Iterator Types*.\n\nobject.__reversed__(self)\n\n Called (if present) by the ``reversed()`` built-in to implement\n reverse iteration. It should return a new iterator object that\n iterates over all the objects in the container in reverse order.\n\n If the ``__reversed__()`` method is not provided, the\n ``reversed()`` built-in will fall back to using the sequence\n protocol (``__len__()`` and ``__getitem__()``). Objects that\n support the sequence protocol should only provide\n ``__reversed__()`` if they can provide an implementation that is\n more efficient than the one provided by ``reversed()``.\n\n New in version 2.6.\n\nThe membership test operators (``in`` and ``not in``) are normally\nimplemented as an iteration through a sequence. However, container\nobjects can supply the following special method with a more efficient\nimplementation, which also does not require the object be a sequence.\n\nobject.__contains__(self, item)\n\n Called to implement membership test operators. Should return true\n if *item* is in *self*, false otherwise. For mapping objects, this\n should consider the keys of the mapping rather than the values or\n the key-item pairs.\n\n For objects that don\'t define ``__contains__()``, the membership\n test first tries iteration via ``__iter__()``, then the old\n sequence iteration protocol via ``__getitem__()``, see *this\n section in the language reference*.\n\n\nAdditional methods for emulation of sequence types\n==================================================\n\nThe following optional methods can be defined to further emulate\nsequence objects. Immutable sequences methods should at most only\ndefine ``__getslice__()``; mutable sequences might define all three\nmethods.\n\nobject.__getslice__(self, i, j)\n\n Deprecated since version 2.0: Support slice objects as parameters\n to the ``__getitem__()`` method. (However, built-in types in\n CPython currently still implement ``__getslice__()``. Therefore,\n you have to override it in derived classes when implementing\n slicing.)\n\n Called to implement evaluation of ``self[i:j]``. The returned\n object should be of the same type as *self*. Note that missing *i*\n or *j* in the slice expression are replaced by zero or\n ``sys.maxint``, respectively. If negative indexes are used in the\n slice, the length of the sequence is added to that index. If the\n instance does not implement the ``__len__()`` method, an\n ``AttributeError`` is raised. No guarantee is made that indexes\n adjusted this way are not still negative. Indexes which are\n greater than the length of the sequence are not modified. If no\n ``__getslice__()`` is found, a slice object is created instead, and\n passed to ``__getitem__()`` instead.\n\nobject.__setslice__(self, i, j, sequence)\n\n Called to implement assignment to ``self[i:j]``. Same notes for *i*\n and *j* as for ``__getslice__()``.\n\n This method is deprecated. If no ``__setslice__()`` is found, or\n for extended slicing of the form ``self[i:j:k]``, a slice object is\n created, and passed to ``__setitem__()``, instead of\n ``__setslice__()`` being called.\n\nobject.__delslice__(self, i, j)\n\n Called to implement deletion of ``self[i:j]``. Same notes for *i*\n and *j* as for ``__getslice__()``. This method is deprecated. If no\n ``__delslice__()`` is found, or for extended slicing of the form\n ``self[i:j:k]``, a slice object is created, and passed to\n ``__delitem__()``, instead of ``__delslice__()`` being called.\n\nNotice that these methods are only invoked when a single slice with a\nsingle colon is used, and the slice method is available. For slice\noperations involving extended slice notation, or in absence of the\nslice methods, ``__getitem__()``, ``__setitem__()`` or\n``__delitem__()`` is called with a slice object as argument.\n\nThe following example demonstrate how to make your program or module\ncompatible with earlier versions of Python (assuming that methods\n``__getitem__()``, ``__setitem__()`` and ``__delitem__()`` support\nslice objects as arguments):\n\n class MyClass:\n ...\n def __getitem__(self, index):\n ...\n def __setitem__(self, index, value):\n ...\n def __delitem__(self, index):\n ...\n\n if sys.version_info < (2, 0):\n # They won\'t be defined if version is at least 2.0 final\n\n def __getslice__(self, i, j):\n return self[max(0, i):max(0, j):]\n def __setslice__(self, i, j, seq):\n self[max(0, i):max(0, j):] = seq\n def __delslice__(self, i, j):\n del self[max(0, i):max(0, j):]\n ...\n\nNote the calls to ``max()``; these are necessary because of the\nhandling of negative indices before the ``__*slice__()`` methods are\ncalled. When negative indexes are used, the ``__*item__()`` methods\nreceive them as provided, but the ``__*slice__()`` methods get a\n"cooked" form of the index values. For each negative index value, the\nlength of the sequence is added to the index before calling the method\n(which may still result in a negative index); this is the customary\nhandling of negative indexes by the built-in sequence types, and the\n``__*item__()`` methods are expected to do this as well. However,\nsince they should already be doing that, negative indexes cannot be\npassed in; they must be constrained to the bounds of the sequence\nbefore being passed to the ``__*item__()`` methods. Calling ``max(0,\ni)`` conveniently returns the proper value.\n\n\nEmulating numeric types\n=======================\n\nThe following methods can be defined to emulate numeric objects.\nMethods corresponding to operations that are not supported by the\nparticular kind of number implemented (e.g., bitwise operations for\nnon-integral numbers) should be left undefined.\n\nobject.__add__(self, other)\nobject.__sub__(self, other)\nobject.__mul__(self, other)\nobject.__floordiv__(self, other)\nobject.__mod__(self, other)\nobject.__divmod__(self, other)\nobject.__pow__(self, other[, modulo])\nobject.__lshift__(self, other)\nobject.__rshift__(self, other)\nobject.__and__(self, other)\nobject.__xor__(self, other)\nobject.__or__(self, other)\n\n These methods are called to implement the binary arithmetic\n operations (``+``, ``-``, ``*``, ``//``, ``%``, ``divmod()``,\n ``pow()``, ``**``, ``<<``, ``>>``, ``&``, ``^``, ``|``). For\n instance, to evaluate the expression ``x + y``, where *x* is an\n instance of a class that has an ``__add__()`` method,\n ``x.__add__(y)`` is called. The ``__divmod__()`` method should be\n the equivalent to using ``__floordiv__()`` and ``__mod__()``; it\n should not be related to ``__truediv__()`` (described below). Note\n that ``__pow__()`` should be defined to accept an optional third\n argument if the ternary version of the built-in ``pow()`` function\n is to be supported.\n\n If one of those methods does not support the operation with the\n supplied arguments, it should return ``NotImplemented``.\n\nobject.__div__(self, other)\nobject.__truediv__(self, other)\n\n The division operator (``/``) is implemented by these methods. The\n ``__truediv__()`` method is used when ``__future__.division`` is in\n effect, otherwise ``__div__()`` is used. If only one of these two\n methods is defined, the object will not support division in the\n alternate context; ``TypeError`` will be raised instead.\n\nobject.__radd__(self, other)\nobject.__rsub__(self, other)\nobject.__rmul__(self, other)\nobject.__rdiv__(self, other)\nobject.__rtruediv__(self, other)\nobject.__rfloordiv__(self, other)\nobject.__rmod__(self, other)\nobject.__rdivmod__(self, other)\nobject.__rpow__(self, other)\nobject.__rlshift__(self, other)\nobject.__rrshift__(self, other)\nobject.__rand__(self, other)\nobject.__rxor__(self, other)\nobject.__ror__(self, other)\n\n These methods are called to implement the binary arithmetic\n operations (``+``, ``-``, ``*``, ``/``, ``%``, ``divmod()``,\n ``pow()``, ``**``, ``<<``, ``>>``, ``&``, ``^``, ``|``) with\n reflected (swapped) operands. These functions are only called if\n the left operand does not support the corresponding operation and\n the operands are of different types. [2] For instance, to evaluate\n the expression ``x - y``, where *y* is an instance of a class that\n has an ``__rsub__()`` method, ``y.__rsub__(x)`` is called if\n ``x.__sub__(y)`` returns *NotImplemented*.\n\n Note that ternary ``pow()`` will not try calling ``__rpow__()``\n (the coercion rules would become too complicated).\n\n Note: If the right operand\'s type is a subclass of the left operand\'s\n type and that subclass provides the reflected method for the\n operation, this method will be called before the left operand\'s\n non-reflected method. This behavior allows subclasses to\n override their ancestors\' operations.\n\nobject.__iadd__(self, other)\nobject.__isub__(self, other)\nobject.__imul__(self, other)\nobject.__idiv__(self, other)\nobject.__itruediv__(self, other)\nobject.__ifloordiv__(self, other)\nobject.__imod__(self, other)\nobject.__ipow__(self, other[, modulo])\nobject.__ilshift__(self, other)\nobject.__irshift__(self, other)\nobject.__iand__(self, other)\nobject.__ixor__(self, other)\nobject.__ior__(self, other)\n\n These methods are called to implement the augmented arithmetic\n assignments (``+=``, ``-=``, ``*=``, ``/=``, ``//=``, ``%=``,\n ``**=``, ``<<=``, ``>>=``, ``&=``, ``^=``, ``|=``). These methods\n should attempt to do the operation in-place (modifying *self*) and\n return the result (which could be, but does not have to be,\n *self*). If a specific method is not defined, the augmented\n assignment falls back to the normal methods. For instance, to\n execute the statement ``x += y``, where *x* is an instance of a\n class that has an ``__iadd__()`` method, ``x.__iadd__(y)`` is\n called. If *x* is an instance of a class that does not define a\n ``__iadd__()`` method, ``x.__add__(y)`` and ``y.__radd__(x)`` are\n considered, as with the evaluation of ``x + y``.\n\nobject.__neg__(self)\nobject.__pos__(self)\nobject.__abs__(self)\nobject.__invert__(self)\n\n Called to implement the unary arithmetic operations (``-``, ``+``,\n ``abs()`` and ``~``).\n\nobject.__complex__(self)\nobject.__int__(self)\nobject.__long__(self)\nobject.__float__(self)\n\n Called to implement the built-in functions ``complex()``,\n ``int()``, ``long()``, and ``float()``. Should return a value of\n the appropriate type.\n\nobject.__oct__(self)\nobject.__hex__(self)\n\n Called to implement the built-in functions ``oct()`` and ``hex()``.\n Should return a string value.\n\nobject.__index__(self)\n\n Called to implement ``operator.index()``. Also called whenever\n Python needs an integer object (such as in slicing). Must return\n an integer (int or long).\n\n New in version 2.5.\n\nobject.__coerce__(self, other)\n\n Called to implement "mixed-mode" numeric arithmetic. Should either\n return a 2-tuple containing *self* and *other* converted to a\n common numeric type, or ``None`` if conversion is impossible. When\n the common type would be the type of ``other``, it is sufficient to\n return ``None``, since the interpreter will also ask the other\n object to attempt a coercion (but sometimes, if the implementation\n of the other type cannot be changed, it is useful to do the\n conversion to the other type here). A return value of\n ``NotImplemented`` is equivalent to returning ``None``.\n\n\nCoercion rules\n==============\n\nThis section used to document the rules for coercion. As the language\nhas evolved, the coercion rules have become hard to document\nprecisely; documenting what one version of one particular\nimplementation does is undesirable. Instead, here are some informal\nguidelines regarding coercion. In Python 3, coercion will not be\nsupported.\n\n* If the left operand of a % operator is a string or Unicode object,\n no coercion takes place and the string formatting operation is\n invoked instead.\n\n* It is no longer recommended to define a coercion operation. Mixed-\n mode operations on types that don\'t define coercion pass the\n original arguments to the operation.\n\n* New-style classes (those derived from ``object``) never invoke the\n ``__coerce__()`` method in response to a binary operator; the only\n time ``__coerce__()`` is invoked is when the built-in function\n ``coerce()`` is called.\n\n* For most intents and purposes, an operator that returns\n ``NotImplemented`` is treated the same as one that is not\n implemented at all.\n\n* Below, ``__op__()`` and ``__rop__()`` are used to signify the\n generic method names corresponding to an operator; ``__iop__()`` is\n used for the corresponding in-place operator. For example, for the\n operator \'``+``\', ``__add__()`` and ``__radd__()`` are used for the\n left and right variant of the binary operator, and ``__iadd__()``\n for the in-place variant.\n\n* For objects *x* and *y*, first ``x.__op__(y)`` is tried. If this is\n not implemented or returns ``NotImplemented``, ``y.__rop__(x)`` is\n tried. If this is also not implemented or returns\n ``NotImplemented``, a ``TypeError`` exception is raised. But see\n the following exception:\n\n* Exception to the previous item: if the left operand is an instance\n of a built-in type or a new-style class, and the right operand is an\n instance of a proper subclass of that type or class and overrides\n the base\'s ``__rop__()`` method, the right operand\'s ``__rop__()``\n method is tried *before* the left operand\'s ``__op__()`` method.\n\n This is done so that a subclass can completely override binary\n operators. Otherwise, the left operand\'s ``__op__()`` method would\n always accept the right operand: when an instance of a given class\n is expected, an instance of a subclass of that class is always\n acceptable.\n\n* When either operand type defines a coercion, this coercion is called\n before that type\'s ``__op__()`` or ``__rop__()`` method is called,\n but no sooner. If the coercion returns an object of a different\n type for the operand whose coercion is invoked, part of the process\n is redone using the new object.\n\n* When an in-place operator (like \'``+=``\') is used, if the left\n operand implements ``__iop__()``, it is invoked without any\n coercion. When the operation falls back to ``__op__()`` and/or\n ``__rop__()``, the normal coercion rules apply.\n\n* In ``x + y``, if *x* is a sequence that implements sequence\n concatenation, sequence concatenation is invoked.\n\n* In ``x * y``, if one operand is a sequence that implements sequence\n repetition, and the other is an integer (``int`` or ``long``),\n sequence repetition is invoked.\n\n* Rich comparisons (implemented by methods ``__eq__()`` and so on)\n never use coercion. Three-way comparison (implemented by\n ``__cmp__()``) does use coercion under the same conditions as other\n binary operations use it.\n\n* In the current implementation, the built-in numeric types ``int``,\n ``long``, ``float``, and ``complex`` do not use coercion. All these\n types implement a ``__coerce__()`` method, for use by the built-in\n ``coerce()`` function.\n\n Changed in version 2.7.\n\n\nWith Statement Context Managers\n===============================\n\nNew in version 2.5.\n\nA *context manager* is an object that defines the runtime context to\nbe established when executing a ``with`` statement. The context\nmanager handles the entry into, and the exit from, the desired runtime\ncontext for the execution of the block of code. Context managers are\nnormally invoked using the ``with`` statement (described in section\n*The with statement*), but can also be used by directly invoking their\nmethods.\n\nTypical uses of context managers include saving and restoring various\nkinds of global state, locking and unlocking resources, closing opened\nfiles, etc.\n\nFor more information on context managers, see *Context Manager Types*.\n\nobject.__enter__(self)\n\n Enter the runtime context related to this object. The ``with``\n statement will bind this method\'s return value to the target(s)\n specified in the ``as`` clause of the statement, if any.\n\nobject.__exit__(self, exc_type, exc_value, traceback)\n\n Exit the runtime context related to this object. The parameters\n describe the exception that caused the context to be exited. If the\n context was exited without an exception, all three arguments will\n be ``None``.\n\n If an exception is supplied, and the method wishes to suppress the\n exception (i.e., prevent it from being propagated), it should\n return a true value. Otherwise, the exception will be processed\n normally upon exit from this method.\n\n Note that ``__exit__()`` methods should not reraise the passed-in\n exception; this is the caller\'s responsibility.\n\nSee also:\n\n **PEP 0343** - The "with" statement\n The specification, background, and examples for the Python\n ``with`` statement.\n\n\nSpecial method lookup for old-style classes\n===========================================\n\nFor old-style classes, special methods are always looked up in exactly\nthe same way as any other method or attribute. This is the case\nregardless of whether the method is being looked up explicitly as in\n``x.__getitem__(i)`` or implicitly as in ``x[i]``.\n\nThis behaviour means that special methods may exhibit different\nbehaviour for different instances of a single old-style class if the\nappropriate special attributes are set differently:\n\n >>> class C:\n ... pass\n ...\n >>> c1 = C()\n >>> c2 = C()\n >>> c1.__len__ = lambda: 5\n >>> c2.__len__ = lambda: 9\n >>> len(c1)\n 5\n >>> len(c2)\n 9\n\n\nSpecial method lookup for new-style classes\n===========================================\n\nFor new-style classes, implicit invocations of special methods are\nonly guaranteed to work correctly if defined on an object\'s type, not\nin the object\'s instance dictionary. That behaviour is the reason why\nthe following code raises an exception (unlike the equivalent example\nwith old-style classes):\n\n >>> class C(object):\n ... pass\n ...\n >>> c = C()\n >>> c.__len__ = lambda: 5\n >>> len(c)\n Traceback (most recent call last):\n File "<stdin>", line 1, in <module>\n TypeError: object of type \'C\' has no len()\n\nThe rationale behind this behaviour lies with a number of special\nmethods such as ``__hash__()`` and ``__repr__()`` that are implemented\nby all objects, including type objects. If the implicit lookup of\nthese methods used the conventional lookup process, they would fail\nwhen invoked on the type object itself:\n\n >>> 1 .__hash__() == hash(1)\n True\n >>> int.__hash__() == hash(int)\n Traceback (most recent call last):\n File "<stdin>", line 1, in <module>\n TypeError: descriptor \'__hash__\' of \'int\' object needs an argument\n\nIncorrectly attempting to invoke an unbound method of a class in this\nway is sometimes referred to as \'metaclass confusion\', and is avoided\nby bypassing the instance when looking up special methods:\n\n >>> type(1).__hash__(1) == hash(1)\n True\n >>> type(int).__hash__(int) == hash(int)\n True\n\nIn addition to bypassing any instance attributes in the interest of\ncorrectness, implicit special method lookup generally also bypasses\nthe ``__getattribute__()`` method even of the object\'s metaclass:\n\n >>> class Meta(type):\n ... def __getattribute__(*args):\n ... print "Metaclass getattribute invoked"\n ... return type.__getattribute__(*args)\n ...\n >>> class C(object):\n ... __metaclass__ = Meta\n ... def __len__(self):\n ... return 10\n ... def __getattribute__(*args):\n ... print "Class getattribute invoked"\n ... return object.__getattribute__(*args)\n ...\n >>> c = C()\n >>> c.__len__() # Explicit lookup via instance\n Class getattribute invoked\n 10\n >>> type(c).__len__(c) # Explicit lookup via type\n Metaclass getattribute invoked\n 10\n >>> len(c) # Implicit lookup\n 10\n\nBypassing the ``__getattribute__()`` machinery in this fashion\nprovides significant scope for speed optimisations within the\ninterpreter, at the cost of some flexibility in the handling of\nspecial methods (the special method *must* be set on the class object\nitself in order to be consistently invoked by the interpreter).\n\n-[ Footnotes ]-\n\n[1] It *is* possible in some cases to change an object\'s type, under\n certain controlled conditions. It generally isn\'t a good idea\n though, since it can lead to some very strange behaviour if it is\n handled incorrectly.\n\n[2] For operands of the same type, it is assumed that if the non-\n reflected method (such as ``__add__()``) fails the operation is\n not supported, which is why the reflected method is not called.\n',
/prebuilts/python/linux-x86/2.7.5/lib/python2.7/pydoc_data/
H A D	topics.py	6 'attribute-access': '\nCustomizing attribute access\n****************************\n\nThe following methods can be defined to customize the meaning of\nattribute access (use of, assignment to, or deletion of ``x.name``)\nfor class instances.\n\nobject.__getattr__(self, name)\n\n Called when an attribute lookup has not found the attribute in the\n usual places (i.e. it is not an instance attribute nor is it found\n in the class tree for ``self``). ``name`` is the attribute name.\n This method should return the (computed) attribute value or raise\n an ``AttributeError`` exception.\n\n Note that if the attribute is found through the normal mechanism,\n ``__getattr__()`` is not called. (This is an intentional asymmetry\n between ``__getattr__()`` and ``__setattr__()``.) This is done both\n for efficiency reasons and because otherwise ``__getattr__()``\n would have no way to access other attributes of the instance. Note\n that at least for instance variables, you can fake total control by\n not inserting any values in the instance attribute dictionary (but\n instead inserting them in another object). See the\n ``__getattribute__()`` method below for a way to actually get total\n control in new-style classes.\n\nobject.__setattr__(self, name, value)\n\n Called when an attribute assignment is attempted. This is called\n instead of the normal mechanism (i.e. store the value in the\n instance dictionary). *name* is the attribute name, *value* is the\n value to be assigned to it.\n\n If ``__setattr__()`` wants to assign to an instance attribute, it\n should not simply execute ``self.name = value`` --- this would\n cause a recursive call to itself. Instead, it should insert the\n value in the dictionary of instance attributes, e.g.,\n ``self.__dict__[name] = value``. For new-style classes, rather\n than accessing the instance dictionary, it should call the base\n class method with the same name, for example,\n ``object.__setattr__(self, name, value)``.\n\nobject.__delattr__(self, name)\n\n Like ``__setattr__()`` but for attribute deletion instead of\n assignment. This should only be implemented if ``del obj.name`` is\n meaningful for the object.\n\n\nMore attribute access for new-style classes\n===========================================\n\nThe following methods only apply to new-style classes.\n\nobject.__getattribute__(self, name)\n\n Called unconditionally to implement attribute accesses for\n instances of the class. If the class also defines\n ``__getattr__()``, the latter will not be called unless\n ``__getattribute__()`` either calls it explicitly or raises an\n ``AttributeError``. This method should return the (computed)\n attribute value or raise an ``AttributeError`` exception. In order\n to avoid infinite recursion in this method, its implementation\n should always call the base class method with the same name to\n access any attributes it needs, for example,\n ``object.__getattribute__(self, name)``.\n\n Note: This method may still be bypassed when looking up special methods\n as the result of implicit invocation via language syntax or\n built-in functions. See *Special method lookup for new-style\n classes*.\n\n\nImplementing Descriptors\n========================\n\nThe following methods only apply when an instance of the class\ncontaining the method (a so-called *descriptor* class) appears in an\n*owner* class (the descriptor must be in either the owner\'s class\ndictionary or in the class dictionary for one of its parents). In the\nexamples below, "the attribute" refers to the attribute whose name is\nthe key of the property in the owner class\' ``__dict__``.\n\nobject.__get__(self, instance, owner)\n\n Called to get the attribute of the owner class (class attribute\n access) or of an instance of that class (instance attribute\n access). *owner* is always the owner class, while *instance* is the\n instance that the attribute was accessed through, or ``None`` when\n the attribute is accessed through the *owner*. This method should\n return the (computed) attribute value or raise an\n ``AttributeError`` exception.\n\nobject.__set__(self, instance, value)\n\n Called to set the attribute on an instance *instance* of the owner\n class to a new value, *value*.\n\nobject.__delete__(self, instance)\n\n Called to delete the attribute on an instance *instance* of the\n owner class.\n\n\nInvoking Descriptors\n====================\n\nIn general, a descriptor is an object attribute with "binding\nbehavior", one whose attribute access has been overridden by methods\nin the descriptor protocol: ``__get__()``, ``__set__()``, and\n``__delete__()``. If any of those methods are defined for an object,\nit is said to be a descriptor.\n\nThe default behavior for attribute access is to get, set, or delete\nthe attribute from an object\'s dictionary. For instance, ``a.x`` has a\nlookup chain starting with ``a.__dict__[\'x\']``, then\n``type(a).__dict__[\'x\']``, and continuing through the base classes of\n``type(a)`` excluding metaclasses.\n\nHowever, if the looked-up value is an object defining one of the\ndescriptor methods, then Python may override the default behavior and\ninvoke the descriptor method instead. Where this occurs in the\nprecedence chain depends on which descriptor methods were defined and\nhow they were called. Note that descriptors are only invoked for new\nstyle objects or classes (ones that subclass ``object()`` or\n``type()``).\n\nThe starting point for descriptor invocation is a binding, ``a.x``.\nHow the arguments are assembled depends on ``a``:\n\nDirect Call\n The simplest and least common call is when user code directly\n invokes a descriptor method: ``x.__get__(a)``.\n\nInstance Binding\n If binding to a new-style object instance, ``a.x`` is transformed\n into the call: ``type(a).__dict__[\'x\'].__get__(a, type(a))``.\n\nClass Binding\n If binding to a new-style class, ``A.x`` is transformed into the\n call: ``A.__dict__[\'x\'].__get__(None, A)``.\n\nSuper Binding\n If ``a`` is an instance of ``super``, then the binding ``super(B,\n obj).m()`` searches ``obj.__class__.__mro__`` for the base class\n ``A`` immediately preceding ``B`` and then invokes the descriptor\n with the call: ``A.__dict__[\'m\'].__get__(obj, obj.__class__)``.\n\nFor instance bindings, the precedence of descriptor invocation depends\non the which descriptor methods are defined. A descriptor can define\nany combination of ``__get__()``, ``__set__()`` and ``__delete__()``.\nIf it does not define ``__get__()``, then accessing the attribute will\nreturn the descriptor object itself unless there is a value in the\nobject\'s instance dictionary. If the descriptor defines ``__set__()``\nand/or ``__delete__()``, it is a data descriptor; if it defines\nneither, it is a non-data descriptor. Normally, data descriptors\ndefine both ``__get__()`` and ``__set__()``, while non-data\ndescriptors have just the ``__get__()`` method. Data descriptors with\n``__set__()`` and ``__get__()`` defined always override a redefinition\nin an instance dictionary. In contrast, non-data descriptors can be\noverridden by instances.\n\nPython methods (including ``staticmethod()`` and ``classmethod()``)\nare implemented as non-data descriptors. Accordingly, instances can\nredefine and override methods. This allows individual instances to\nacquire behaviors that differ from other instances of the same class.\n\nThe ``property()`` function is implemented as a data descriptor.\nAccordingly, instances cannot override the behavior of a property.\n\n\n__slots__\n=========\n\nBy default, instances of both old and new-style classes have a\ndictionary for attribute storage. This wastes space for objects\nhaving very few instance variables. The space consumption can become\nacute when creating large numbers of instances.\n\nThe default can be overridden by defining *__slots__* in a new-style\nclass definition. The *__slots__* declaration takes a sequence of\ninstance variables and reserves just enough space in each instance to\nhold a value for each variable. Space is saved because *__dict__* is\nnot created for each instance.\n\n__slots__\n\n This class variable can be assigned a string, iterable, or sequence\n of strings with variable names used by instances. If defined in a\n new-style class, *__slots__* reserves space for the declared\n variables and prevents the automatic creation of *__dict__* and\n *__weakref__* for each instance.\n\n New in version 2.2.\n\nNotes on using *__slots__*\n\n* When inheriting from a class without *__slots__*, the *__dict__*\n attribute of that class will always be accessible, so a *__slots__*\n definition in the subclass is meaningless.\n\n* Without a *__dict__* variable, instances cannot be assigned new\n variables not listed in the *__slots__* definition. Attempts to\n assign to an unlisted variable name raises ``AttributeError``. If\n dynamic assignment of new variables is desired, then add\n ``\'__dict__\'`` to the sequence of strings in the *__slots__*\n declaration.\n\n Changed in version 2.3: Previously, adding ``\'__dict__\'`` to the\n *__slots__* declaration would not enable the assignment of new\n attributes not specifically listed in the sequence of instance\n variable names.\n\n* Without a *__weakref__* variable for each instance, classes defining\n *__slots__* do not support weak references to its instances. If weak\n reference support is needed, then add ``\'__weakref__\'`` to the\n sequence of strings in the *__slots__* declaration.\n\n Changed in version 2.3: Previously, adding ``\'__weakref__\'`` to the\n *__slots__* declaration would not enable support for weak\n references.\n\n* *__slots__* are implemented at the class level by creating\n descriptors (*Implementing Descriptors*) for each variable name. As\n a result, class attributes cannot be used to set default values for\n instance variables defined by *__slots__*; otherwise, the class\n attribute would overwrite the descriptor assignment.\n\n* The action of a *__slots__* declaration is limited to the class\n where it is defined. As a result, subclasses will have a *__dict__*\n unless they also define *__slots__* (which must only contain names\n of any *additional* slots).\n\n* If a class defines a slot also defined in a base class, the instance\n variable defined by the base class slot is inaccessible (except by\n retrieving its descriptor directly from the base class). This\n renders the meaning of the program undefined. In the future, a\n check may be added to prevent this.\n\n* Nonempty *__slots__* does not work for classes derived from\n "variable-length" built-in types such as ``long``, ``str`` and\n ``tuple``.\n\n* Any non-string iterable may be assigned to *__slots__*. Mappings may\n also be used; however, in the future, special meaning may be\n assigned to the values corresponding to each key.\n\n* *__class__* assignment works only if both classes have the same\n *__slots__*.\n\n Changed in version 2.6: Previously, *__class__* assignment raised an\n error if either new or old class had *__slots__*.\n', 18 'calls': '\nCalls\n*****\n\nA call calls a callable object (e.g., a *function*) with a possibly\nempty series of *arguments*:\n\n call ::= primary "(" [argument_list [","]\n | expression genexpr_for] ")"\n argument_list ::= positional_arguments ["," keyword_arguments]\n ["," "*" expression] ["," keyword_arguments]\n ["," "**" expression]\n | keyword_arguments ["," "*" expression]\n ["," "**" expression]\n | "*" expression ["," "*" expression] ["," "**" expression]\n | "**" expression\n positional_arguments ::= expression ("," expression)*\n keyword_arguments ::= keyword_item ("," keyword_item)*\n keyword_item ::= identifier "=" expression\n\nA trailing comma may be present after the positional and keyword\narguments but does not affect the semantics.\n\nThe primary must evaluate to a callable object (user-defined\nfunctions, built-in functions, methods of built-in objects, class\nobjects, methods of class instances, and certain class instances\nthemselves are callable; extensions may define additional callable\nobject types). All argument expressions are evaluated before the call\nis attempted. Please refer to section *Function definitions* for the\nsyntax of formal *parameter* lists.\n\nIf keyword arguments are present, they are first converted to\npositional arguments, as follows. First, a list of unfilled slots is\ncreated for the formal parameters. If there are N positional\narguments, they are placed in the first N slots. Next, for each\nkeyword argument, the identifier is used to determine the\ncorresponding slot (if the identifier is the same as the first formal\nparameter name, the first slot is used, and so on). If the slot is\nalready filled, a ``TypeError`` exception is raised. Otherwise, the\nvalue of the argument is placed in the slot, filling it (even if the\nexpression is ``None``, it fills the slot). When all arguments have\nbeen processed, the slots that are still unfilled are filled with the\ncorresponding default value from the function definition. (Default\nvalues are calculated, once, when the function is defined; thus, a\nmutable object such as a list or dictionary used as default value will\nbe shared by all calls that don\'t specify an argument value for the\ncorresponding slot; this should usually be avoided.) If there are any\nunfilled slots for which no default value is specified, a\n``TypeError`` exception is raised. Otherwise, the list of filled\nslots is used as the argument list for the call.\n\n**CPython implementation detail:** An implementation may provide\nbuilt-in functions whose positional parameters do not have names, even\nif they are \'named\' for the purpose of documentation, and which\ntherefore cannot be supplied by keyword. In CPython, this is the case\nfor functions implemented in C that use ``PyArg_ParseTuple()`` to\nparse their arguments.\n\nIf there are more positional arguments than there are formal parameter\nslots, a ``TypeError`` exception is raised, unless a formal parameter\nusing the syntax ``*identifier`` is present; in this case, that formal\nparameter receives a tuple containing the excess positional arguments\n(or an empty tuple if there were no excess positional arguments).\n\nIf any keyword argument does not correspond to a formal parameter\nname, a ``TypeError`` exception is raised, unless a formal parameter\nusing the syntax ``**identifier`` is present; in this case, that\nformal parameter receives a dictionary containing the excess keyword\narguments (using the keywords as keys and the argument values as\ncorresponding values), or a (new) empty dictionary if there were no\nexcess keyword arguments.\n\nIf the syntax ``*expression`` appears in the function call,\n``expression`` must evaluate to an iterable. Elements from this\niterable are treated as if they were additional positional arguments;\nif there are positional arguments *x1*, ..., *xN*, and ``expression``\nevaluates to a sequence *y1*, ..., *yM*, this is equivalent to a call\nwith M+N positional arguments *x1*, ..., *xN*, *y1*, ..., *yM*.\n\nA consequence of this is that although the ``*expression`` syntax may\nappear *after* some keyword arguments, it is processed *before* the\nkeyword arguments (and the ``**expression`` argument, if any -- see\nbelow). So:\n\n >>> def f(a, b):\n ... print a, b\n ...\n >>> f(b=1, *(2,))\n 2 1\n >>> f(a=1, *(2,))\n Traceback (most recent call last):\n File "<stdin>", line 1, in ?\n TypeError: f() got multiple values for keyword argument \'a\'\n >>> f(1, *(2,))\n 1 2\n\nIt is unusual for both keyword arguments and the ``*expression``\nsyntax to be used in the same call, so in practice this confusion does\nnot arise.\n\nIf the syntax ``**expression`` appears in the function call,\n``expression`` must evaluate to a mapping, the contents of which are\ntreated as additional keyword arguments. In the case of a keyword\nappearing in both ``expression`` and as an explicit keyword argument,\na ``TypeError`` exception is raised.\n\nFormal parameters using the syntax ``*identifier`` or ``**identifier``\ncannot be used as positional argument slots or as keyword argument\nnames. Formal parameters using the syntax ``(sublist)`` cannot be\nused as keyword argument names; the outermost sublist corresponds to a\nsingle unnamed argument slot, and the argument value is assigned to\nthe sublist using the usual tuple assignment rules after all other\nparameter processing is done.\n\nA call always returns some value, possibly ``None``, unless it raises\nan exception. How this value is computed depends on the type of the\ncallable object.\n\nIf it is---\n\na user-defined function:\n The code block for the function is executed, passing it the\n argument list. The first thing the code block will do is bind the\n formal parameters to the arguments; this is described in section\n *Function definitions*. When the code block executes a ``return``\n statement, this specifies the return value of the function call.\n\na built-in function or method:\n The result is up to the interpreter; see *Built-in Functions* for\n the descriptions of built-in functions and methods.\n\na class object:\n A new instance of that class is returned.\n\na class instance method:\n The corresponding user-defined function is called, with an argument\n list that is one longer than the argument list of the call: the\n instance becomes the first argument.\n\na class instance:\n The class must define a ``__call__()`` method; the effect is then\n the same as if that method was called.\n', 61 'specialnames': '\nSpecial method names\n********************\n\nA class can implement certain operations that are invoked by special\nsyntax (such as arithmetic operations or subscripting and slicing) by\ndefining methods with special names. This is Python\'s approach to\n*operator overloading*, allowing classes to define their own behavior\nwith respect to language operators. For instance, if a class defines\na method named ``__getitem__()``, and ``x`` is an instance of this\nclass, then ``x[i]`` is roughly equivalent to ``x.__getitem__(i)`` for\nold-style classes and ``type(x).__getitem__(x, i)`` for new-style\nclasses. Except where mentioned, attempts to execute an operation\nraise an exception when no appropriate method is defined (typically\n``AttributeError`` or ``TypeError``).\n\nWhen implementing a class that emulates any built-in type, it is\nimportant that the emulation only be implemented to the degree that it\nmakes sense for the object being modelled. For example, some\nsequences may work well with retrieval of individual elements, but\nextracting a slice may not make sense. (One example of this is the\n``NodeList`` interface in the W3C\'s Document Object Model.)\n\n\nBasic customization\n===================\n\nobject.__new__(cls[, ...])\n\n Called to create a new instance of class *cls*. ``__new__()`` is a\n static method (special-cased so you need not declare it as such)\n that takes the class of which an instance was requested as its\n first argument. The remaining arguments are those passed to the\n object constructor expression (the call to the class). The return\n value of ``__new__()`` should be the new object instance (usually\n an instance of *cls*).\n\n Typical implementations create a new instance of the class by\n invoking the superclass\'s ``__new__()`` method using\n ``super(currentclass, cls).__new__(cls[, ...])`` with appropriate\n arguments and then modifying the newly-created instance as\n necessary before returning it.\n\n If ``__new__()`` returns an instance of *cls*, then the new\n instance\'s ``__init__()`` method will be invoked like\n ``__init__(self[, ...])``, where *self* is the new instance and the\n remaining arguments are the same as were passed to ``__new__()``.\n\n If ``__new__()`` does not return an instance of *cls*, then the new\n instance\'s ``__init__()`` method will not be invoked.\n\n ``__new__()`` is intended mainly to allow subclasses of immutable\n types (like int, str, or tuple) to customize instance creation. It\n is also commonly overridden in custom metaclasses in order to\n customize class creation.\n\nobject.__init__(self[, ...])\n\n Called when the instance is created. The arguments are those\n passed to the class constructor expression. If a base class has an\n ``__init__()`` method, the derived class\'s ``__init__()`` method,\n if any, must explicitly call it to ensure proper initialization of\n the base class part of the instance; for example:\n ``BaseClass.__init__(self, [args...])``. As a special constraint\n on constructors, no value may be returned; doing so will cause a\n ``TypeError`` to be raised at runtime.\n\nobject.__del__(self)\n\n Called when the instance is about to be destroyed. This is also\n called a destructor. If a base class has a ``__del__()`` method,\n the derived class\'s ``__del__()`` method, if any, must explicitly\n call it to ensure proper deletion of the base class part of the\n instance. Note that it is possible (though not recommended!) for\n the ``__del__()`` method to postpone destruction of the instance by\n creating a new reference to it. It may then be called at a later\n time when this new reference is deleted. It is not guaranteed that\n ``__del__()`` methods are called for objects that still exist when\n the interpreter exits.\n\n Note: ``del x`` doesn\'t directly call ``x.__del__()`` --- the former\n decrements the reference count for ``x`` by one, and the latter\n is only called when ``x``\'s reference count reaches zero. Some\n common situations that may prevent the reference count of an\n object from going to zero include: circular references between\n objects (e.g., a doubly-linked list or a tree data structure with\n parent and child pointers); a reference to the object on the\n stack frame of a function that caught an exception (the traceback\n stored in ``sys.exc_traceback`` keeps the stack frame alive); or\n a reference to the object on the stack frame that raised an\n unhandled exception in interactive mode (the traceback stored in\n ``sys.last_traceback`` keeps the stack frame alive). The first\n situation can only be remedied by explicitly breaking the cycles;\n the latter two situations can be resolved by storing ``None`` in\n ``sys.exc_traceback`` or ``sys.last_traceback``. Circular\n references which are garbage are detected when the option cycle\n detector is enabled (it\'s on by default), but can only be cleaned\n up if there are no Python-level ``__del__()`` methods involved.\n Refer to the documentation for the ``gc`` module for more\n information about how ``__del__()`` methods are handled by the\n cycle detector, particularly the description of the ``garbage``\n value.\n\n Warning: Due to the precarious circumstances under which ``__del__()``\n methods are invoked, exceptions that occur during their execution\n are ignored, and a warning is printed to ``sys.stderr`` instead.\n Also, when ``__del__()`` is invoked in response to a module being\n deleted (e.g., when execution of the program is done), other\n globals referenced by the ``__del__()`` method may already have\n been deleted or in the process of being torn down (e.g. the\n import machinery shutting down). For this reason, ``__del__()``\n methods should do the absolute minimum needed to maintain\n external invariants. Starting with version 1.5, Python\n guarantees that globals whose name begins with a single\n underscore are deleted from their module before other globals are\n deleted; if no other references to such globals exist, this may\n help in assuring that imported modules are still available at the\n time when the ``__del__()`` method is called.\n\n See also the *-R* command-line option.\n\nobject.__repr__(self)\n\n Called by the ``repr()`` built-in function and by string\n conversions (reverse quotes) to compute the "official" string\n representation of an object. If at all possible, this should look\n like a valid Python expression that could be used to recreate an\n object with the same value (given an appropriate environment). If\n this is not possible, a string of the form ``<...some useful\n description...>`` should be returned. The return value must be a\n string object. If a class defines ``__repr__()`` but not\n ``__str__()``, then ``__repr__()`` is also used when an "informal"\n string representation of instances of that class is required.\n\n This is typically used for debugging, so it is important that the\n representation is information-rich and unambiguous.\n\nobject.__str__(self)\n\n Called by the ``str()`` built-in function and by the ``print``\n statement to compute the "informal" string representation of an\n object. This differs from ``__repr__()`` in that it does not have\n to be a valid Python expression: a more convenient or concise\n representation may be used instead. The return value must be a\n string object.\n\nobject.__lt__(self, other)\nobject.__le__(self, other)\nobject.__eq__(self, other)\nobject.__ne__(self, other)\nobject.__gt__(self, other)\nobject.__ge__(self, other)\n\n New in version 2.1.\n\n These are the so-called "rich comparison" methods, and are called\n for comparison operators in preference to ``__cmp__()`` below. The\n correspondence between operator symbols and method names is as\n follows: ``x<y`` calls ``x.__lt__(y)``, ``x<=y`` calls\n ``x.__le__(y)``, ``x==y`` calls ``x.__eq__(y)``, ``x!=y`` and\n ``x<>y`` call ``x.__ne__(y)``, ``x>y`` calls ``x.__gt__(y)``, and\n ``x>=y`` calls ``x.__ge__(y)``.\n\n A rich comparison method may return the singleton\n ``NotImplemented`` if it does not implement the operation for a\n given pair of arguments. By convention, ``False`` and ``True`` are\n returned for a successful comparison. However, these methods can\n return any value, so if the comparison operator is used in a\n Boolean context (e.g., in the condition of an ``if`` statement),\n Python will call ``bool()`` on the value to determine if the result\n is true or false.\n\n There are no implied relationships among the comparison operators.\n The truth of ``x==y`` does not imply that ``x!=y`` is false.\n Accordingly, when defining ``__eq__()``, one should also define\n ``__ne__()`` so that the operators will behave as expected. See\n the paragraph on ``__hash__()`` for some important notes on\n creating *hashable* objects which support custom comparison\n operations and are usable as dictionary keys.\n\n There are no swapped-argument versions of these methods (to be used\n when the left argument does not support the operation but the right\n argument does); rather, ``__lt__()`` and ``__gt__()`` are each\n other\'s reflection, ``__le__()`` and ``__ge__()`` are each other\'s\n reflection, and ``__eq__()`` and ``__ne__()`` are their own\n reflection.\n\n Arguments to rich comparison methods are never coerced.\n\n To automatically generate ordering operations from a single root\n operation, see ``functools.total_ordering()``.\n\nobject.__cmp__(self, other)\n\n Called by comparison operations if rich comparison (see above) is\n not defined. Should return a negative integer if ``self < other``,\n zero if ``self == other``, a positive integer if ``self > other``.\n If no ``__cmp__()``, ``__eq__()`` or ``__ne__()`` operation is\n defined, class instances are compared by object identity\n ("address"). See also the description of ``__hash__()`` for some\n important notes on creating *hashable* objects which support custom\n comparison operations and are usable as dictionary keys. (Note: the\n restriction that exceptions are not propagated by ``__cmp__()`` has\n been removed since Python 1.5.)\n\nobject.__rcmp__(self, other)\n\n Changed in version 2.1: No longer supported.\n\nobject.__hash__(self)\n\n Called by built-in function ``hash()`` and for operations on\n members of hashed collections including ``set``, ``frozenset``, and\n ``dict``. ``__hash__()`` should return an integer. The only\n required property is that objects which compare equal have the same\n hash value; it is advised to somehow mix together (e.g. using\n exclusive or) the hash values for the components of the object that\n also play a part in comparison of objects.\n\n If a class does not define a ``__cmp__()`` or ``__eq__()`` method\n it should not define a ``__hash__()`` operation either; if it\n defines ``__cmp__()`` or ``__eq__()`` but not ``__hash__()``, its\n instances will not be usable in hashed collections. If a class\n defines mutable objects and implements a ``__cmp__()`` or\n ``__eq__()`` method, it should not implement ``__hash__()``, since\n hashable collection implementations require that a object\'s hash\n value is immutable (if the object\'s hash value changes, it will be\n in the wrong hash bucket).\n\n User-defined classes have ``__cmp__()`` and ``__hash__()`` methods\n by default; with them, all objects compare unequal (except with\n themselves) and ``x.__hash__()`` returns ``id(x)``.\n\n Classes which inherit a ``__hash__()`` method from a parent class\n but change the meaning of ``__cmp__()`` or ``__eq__()`` such that\n the hash value returned is no longer appropriate (e.g. by switching\n to a value-based concept of equality instead of the default\n identity based equality) can explicitly flag themselves as being\n unhashable by setting ``__hash__ = None`` in the class definition.\n Doing so means that not only will instances of the class raise an\n appropriate ``TypeError`` when a program attempts to retrieve their\n hash value, but they will also be correctly identified as\n unhashable when checking ``isinstance(obj, collections.Hashable)``\n (unlike classes which define their own ``__hash__()`` to explicitly\n raise ``TypeError``).\n\n Changed in version 2.5: ``__hash__()`` may now also return a long\n integer object; the 32-bit integer is then derived from the hash of\n that object.\n\n Changed in version 2.6: ``__hash__`` may now be set to ``None`` to\n explicitly flag instances of a class as unhashable.\n\nobject.__nonzero__(self)\n\n Called to implement truth value testing and the built-in operation\n ``bool()``; should return ``False`` or ``True``, or their integer\n equivalents ``0`` or ``1``. When this method is not defined,\n ``__len__()`` is called, if it is defined, and the object is\n considered true if its result is nonzero. If a class defines\n neither ``__len__()`` nor ``__nonzero__()``, all its instances are\n considered true.\n\nobject.__unicode__(self)\n\n Called to implement ``unicode()`` built-in; should return a Unicode\n object. When this method is not defined, string conversion is\n attempted, and the result of string conversion is converted to\n Unicode using the system default encoding.\n\n\nCustomizing attribute access\n============================\n\nThe following methods can be defined to customize the meaning of\nattribute access (use of, assignment to, or deletion of ``x.name``)\nfor class instances.\n\nobject.__getattr__(self, name)\n\n Called when an attribute lookup has not found the attribute in the\n usual places (i.e. it is not an instance attribute nor is it found\n in the class tree for ``self``). ``name`` is the attribute name.\n This method should return the (computed) attribute value or raise\n an ``AttributeError`` exception.\n\n Note that if the attribute is found through the normal mechanism,\n ``__getattr__()`` is not called. (This is an intentional asymmetry\n between ``__getattr__()`` and ``__setattr__()``.) This is done both\n for efficiency reasons and because otherwise ``__getattr__()``\n would have no way to access other attributes of the instance. Note\n that at least for instance variables, you can fake total control by\n not inserting any values in the instance attribute dictionary (but\n instead inserting them in another object). See the\n ``__getattribute__()`` method below for a way to actually get total\n control in new-style classes.\n\nobject.__setattr__(self, name, value)\n\n Called when an attribute assignment is attempted. This is called\n instead of the normal mechanism (i.e. store the value in the\n instance dictionary). *name* is the attribute name, *value* is the\n value to be assigned to it.\n\n If ``__setattr__()`` wants to assign to an instance attribute, it\n should not simply execute ``self.name = value`` --- this would\n cause a recursive call to itself. Instead, it should insert the\n value in the dictionary of instance attributes, e.g.,\n ``self.__dict__[name] = value``. For new-style classes, rather\n than accessing the instance dictionary, it should call the base\n class method with the same name, for example,\n ``object.__setattr__(self, name, value)``.\n\nobject.__delattr__(self, name)\n\n Like ``__setattr__()`` but for attribute deletion instead of\n assignment. This should only be implemented if ``del obj.name`` is\n meaningful for the object.\n\n\nMore attribute access for new-style classes\n-------------------------------------------\n\nThe following methods only apply to new-style classes.\n\nobject.__getattribute__(self, name)\n\n Called unconditionally to implement attribute accesses for\n instances of the class. If the class also defines\n ``__getattr__()``, the latter will not be called unless\n ``__getattribute__()`` either calls it explicitly or raises an\n ``AttributeError``. This method should return the (computed)\n attribute value or raise an ``AttributeError`` exception. In order\n to avoid infinite recursion in this method, its implementation\n should always call the base class method with the same name to\n access any attributes it needs, for example,\n ``object.__getattribute__(self, name)``.\n\n Note: This method may still be bypassed when looking up special methods\n as the result of implicit invocation via language syntax or\n built-in functions. See *Special method lookup for new-style\n classes*.\n\n\nImplementing Descriptors\n------------------------\n\nThe following methods only apply when an instance of the class\ncontaining the method (a so-called *descriptor* class) appears in an\n*owner* class (the descriptor must be in either the owner\'s class\ndictionary or in the class dictionary for one of its parents). In the\nexamples below, "the attribute" refers to the attribute whose name is\nthe key of the property in the owner class\' ``__dict__``.\n\nobject.__get__(self, instance, owner)\n\n Called to get the attribute of the owner class (class attribute\n access) or of an instance of that class (instance attribute\n access). *owner* is always the owner class, while *instance* is the\n instance that the attribute was accessed through, or ``None`` when\n the attribute is accessed through the *owner*. This method should\n return the (computed) attribute value or raise an\n ``AttributeError`` exception.\n\nobject.__set__(self, instance, value)\n\n Called to set the attribute on an instance *instance* of the owner\n class to a new value, *value*.\n\nobject.__delete__(self, instance)\n\n Called to delete the attribute on an instance *instance* of the\n owner class.\n\n\nInvoking Descriptors\n--------------------\n\nIn general, a descriptor is an object attribute with "binding\nbehavior", one whose attribute access has been overridden by methods\nin the descriptor protocol: ``__get__()``, ``__set__()``, and\n``__delete__()``. If any of those methods are defined for an object,\nit is said to be a descriptor.\n\nThe default behavior for attribute access is to get, set, or delete\nthe attribute from an object\'s dictionary. For instance, ``a.x`` has a\nlookup chain starting with ``a.__dict__[\'x\']``, then\n``type(a).__dict__[\'x\']``, and continuing through the base classes of\n``type(a)`` excluding metaclasses.\n\nHowever, if the looked-up value is an object defining one of the\ndescriptor methods, then Python may override the default behavior and\ninvoke the descriptor method instead. Where this occurs in the\nprecedence chain depends on which descriptor methods were defined and\nhow they were called. Note that descriptors are only invoked for new\nstyle objects or classes (ones that subclass ``object()`` or\n``type()``).\n\nThe starting point for descriptor invocation is a binding, ``a.x``.\nHow the arguments are assembled depends on ``a``:\n\nDirect Call\n The simplest and least common call is when user code directly\n invokes a descriptor method: ``x.__get__(a)``.\n\nInstance Binding\n If binding to a new-style object instance, ``a.x`` is transformed\n into the call: ``type(a).__dict__[\'x\'].__get__(a, type(a))``.\n\nClass Binding\n If binding to a new-style class, ``A.x`` is transformed into the\n call: ``A.__dict__[\'x\'].__get__(None, A)``.\n\nSuper Binding\n If ``a`` is an instance of ``super``, then the binding ``super(B,\n obj).m()`` searches ``obj.__class__.__mro__`` for the base class\n ``A`` immediately preceding ``B`` and then invokes the descriptor\n with the call: ``A.__dict__[\'m\'].__get__(obj, obj.__class__)``.\n\nFor instance bindings, the precedence of descriptor invocation depends\non the which descriptor methods are defined. A descriptor can define\nany combination of ``__get__()``, ``__set__()`` and ``__delete__()``.\nIf it does not define ``__get__()``, then accessing the attribute will\nreturn the descriptor object itself unless there is a value in the\nobject\'s instance dictionary. If the descriptor defines ``__set__()``\nand/or ``__delete__()``, it is a data descriptor; if it defines\nneither, it is a non-data descriptor. Normally, data descriptors\ndefine both ``__get__()`` and ``__set__()``, while non-data\ndescriptors have just the ``__get__()`` method. Data descriptors with\n``__set__()`` and ``__get__()`` defined always override a redefinition\nin an instance dictionary. In contrast, non-data descriptors can be\noverridden by instances.\n\nPython methods (including ``staticmethod()`` and ``classmethod()``)\nare implemented as non-data descriptors. Accordingly, instances can\nredefine and override methods. This allows individual instances to\nacquire behaviors that differ from other instances of the same class.\n\nThe ``property()`` function is implemented as a data descriptor.\nAccordingly, instances cannot override the behavior of a property.\n\n\n__slots__\n---------\n\nBy default, instances of both old and new-style classes have a\ndictionary for attribute storage. This wastes space for objects\nhaving very few instance variables. The space consumption can become\nacute when creating large numbers of instances.\n\nThe default can be overridden by defining *__slots__* in a new-style\nclass definition. The *__slots__* declaration takes a sequence of\ninstance variables and reserves just enough space in each instance to\nhold a value for each variable. Space is saved because *__dict__* is\nnot created for each instance.\n\n__slots__\n\n This class variable can be assigned a string, iterable, or sequence\n of strings with variable names used by instances. If defined in a\n new-style class, *__slots__* reserves space for the declared\n variables and prevents the automatic creation of *__dict__* and\n *__weakref__* for each instance.\n\n New in version 2.2.\n\nNotes on using *__slots__*\n\n* When inheriting from a class without *__slots__*, the *__dict__*\n attribute of that class will always be accessible, so a *__slots__*\n definition in the subclass is meaningless.\n\n* Without a *__dict__* variable, instances cannot be assigned new\n variables not listed in the *__slots__* definition. Attempts to\n assign to an unlisted variable name raises ``AttributeError``. If\n dynamic assignment of new variables is desired, then add\n ``\'__dict__\'`` to the sequence of strings in the *__slots__*\n declaration.\n\n Changed in version 2.3: Previously, adding ``\'__dict__\'`` to the\n *__slots__* declaration would not enable the assignment of new\n attributes not specifically listed in the sequence of instance\n variable names.\n\n* Without a *__weakref__* variable for each instance, classes defining\n *__slots__* do not support weak references to its instances. If weak\n reference support is needed, then add ``\'__weakref__\'`` to the\n sequence of strings in the *__slots__* declaration.\n\n Changed in version 2.3: Previously, adding ``\'__weakref__\'`` to the\n *__slots__* declaration would not enable support for weak\n references.\n\n* *__slots__* are implemented at the class level by creating\n descriptors (*Implementing Descriptors*) for each variable name. As\n a result, class attributes cannot be used to set default values for\n instance variables defined by *__slots__*; otherwise, the class\n attribute would overwrite the descriptor assignment.\n\n* The action of a *__slots__* declaration is limited to the class\n where it is defined. As a result, subclasses will have a *__dict__*\n unless they also define *__slots__* (which must only contain names\n of any *additional* slots).\n\n* If a class defines a slot also defined in a base class, the instance\n variable defined by the base class slot is inaccessible (except by\n retrieving its descriptor directly from the base class). This\n renders the meaning of the program undefined. In the future, a\n check may be added to prevent this.\n\n* Nonempty *__slots__* does not work for classes derived from\n "variable-length" built-in types such as ``long``, ``str`` and\n ``tuple``.\n\n* Any non-string iterable may be assigned to *__slots__*. Mappings may\n also be used; however, in the future, special meaning may be\n assigned to the values corresponding to each key.\n\n* *__class__* assignment works only if both classes have the same\n *__slots__*.\n\n Changed in version 2.6: Previously, *__class__* assignment raised an\n error if either new or old class had *__slots__*.\n\n\nCustomizing class creation\n==========================\n\nBy default, new-style classes are constructed using ``type()``. A\nclass definition is read into a separate namespace and the value of\nclass name is bound to the result of ``type(name, bases, dict)``.\n\nWhen the class definition is read, if *__metaclass__* is defined then\nthe callable assigned to it will be called instead of ``type()``. This\nallows classes or functions to be written which monitor or alter the\nclass creation process:\n\n* Modifying the class dictionary prior to the class being created.\n\n* Returning an instance of another class -- essentially performing the\n role of a factory function.\n\nThese steps will have to be performed in the metaclass\'s ``__new__()``\nmethod -- ``type.__new__()`` can then be called from this method to\ncreate a class with different properties. This example adds a new\nelement to the class dictionary before creating the class:\n\n class metacls(type):\n def __new__(mcs, name, bases, dict):\n dict[\'foo\'] = \'metacls was here\'\n return type.__new__(mcs, name, bases, dict)\n\nYou can of course also override other class methods (or add new\nmethods); for example defining a custom ``__call__()`` method in the\nmetaclass allows custom behavior when the class is called, e.g. not\nalways creating a new instance.\n\n__metaclass__\n\n This variable can be any callable accepting arguments for ``name``,\n ``bases``, and ``dict``. Upon class creation, the callable is used\n instead of the built-in ``type()``.\n\n New in version 2.2.\n\nThe appropriate metaclass is determined by the following precedence\nrules:\n\n* If ``dict[\'__metaclass__\']`` exists, it is used.\n\n* Otherwise, if there is at least one base class, its metaclass is\n used (this looks for a *__class__* attribute first and if not found,\n uses its type).\n\n* Otherwise, if a global variable named __metaclass__ exists, it is\n used.\n\n* Otherwise, the old-style, classic metaclass (types.ClassType) is\n used.\n\nThe potential uses for metaclasses are boundless. Some ideas that have\nbeen explored including logging, interface checking, automatic\ndelegation, automatic property creation, proxies, frameworks, and\nautomatic resource locking/synchronization.\n\n\nCustomizing instance and subclass checks\n========================================\n\nNew in version 2.6.\n\nThe following methods are used to override the default behavior of the\n``isinstance()`` and ``issubclass()`` built-in functions.\n\nIn particular, the metaclass ``abc.ABCMeta`` implements these methods\nin order to allow the addition of Abstract Base Classes (ABCs) as\n"virtual base classes" to any class or type (including built-in\ntypes), including other ABCs.\n\nclass.__instancecheck__(self, instance)\n\n Return true if *instance* should be considered a (direct or\n indirect) instance of *class*. If defined, called to implement\n ``isinstance(instance, class)``.\n\nclass.__subclasscheck__(self, subclass)\n\n Return true if *subclass* should be considered a (direct or\n indirect) subclass of *class*. If defined, called to implement\n ``issubclass(subclass, class)``.\n\nNote that these methods are looked up on the type (metaclass) of a\nclass. They cannot be defined as class methods in the actual class.\nThis is consistent with the lookup of special methods that are called\non instances, only in this case the instance is itself a class.\n\nSee also:\n\n **PEP 3119** - Introducing Abstract Base Classes\n Includes the specification for customizing ``isinstance()`` and\n ``issubclass()`` behavior through ``__instancecheck__()`` and\n ``__subclasscheck__()``, with motivation for this functionality\n in the context of adding Abstract Base Classes (see the ``abc``\n module) to the language.\n\n\nEmulating callable objects\n==========================\n\nobject.__call__(self[, args...])\n\n Called when the instance is "called" as a function; if this method\n is defined, ``x(arg1, arg2, ...)`` is a shorthand for\n ``x.__call__(arg1, arg2, ...)``.\n\n\nEmulating container types\n=========================\n\nThe following methods can be defined to implement container objects.\nContainers usually are sequences (such as lists or tuples) or mappings\n(like dictionaries), but can represent other containers as well. The\nfirst set of methods is used either to emulate a sequence or to\nemulate a mapping; the difference is that for a sequence, the\nallowable keys should be the integers *k* for which ``0 <= k < N``\nwhere *N* is the length of the sequence, or slice objects, which\ndefine a range of items. (For backwards compatibility, the method\n``__getslice__()`` (see below) can also be defined to handle simple,\nbut not extended slices.) It is also recommended that mappings provide\nthe methods ``keys()``, ``values()``, ``items()``, ``has_key()``,\n``get()``, ``clear()``, ``setdefault()``, ``iterkeys()``,\n``itervalues()``, ``iteritems()``, ``pop()``, ``popitem()``,\n``copy()``, and ``update()`` behaving similar to those for Python\'s\nstandard dictionary objects. The ``UserDict`` module provides a\n``DictMixin`` class to help create those methods from a base set of\n``__getitem__()``, ``__setitem__()``, ``__delitem__()``, and\n``keys()``. Mutable sequences should provide methods ``append()``,\n``count()``, ``index()``, ``extend()``, ``insert()``, ``pop()``,\n``remove()``, ``reverse()`` and ``sort()``, like Python standard list\nobjects. Finally, sequence types should implement addition (meaning\nconcatenation) and multiplication (meaning repetition) by defining the\nmethods ``__add__()``, ``__radd__()``, ``__iadd__()``, ``__mul__()``,\n``__rmul__()`` and ``__imul__()`` described below; they should not\ndefine ``__coerce__()`` or other numerical operators. It is\nrecommended that both mappings and sequences implement the\n``__contains__()`` method to allow efficient use of the ``in``\noperator; for mappings, ``in`` should be equivalent of ``has_key()``;\nfor sequences, it should search through the values. It is further\nrecommended that both mappings and sequences implement the\n``__iter__()`` method to allow efficient iteration through the\ncontainer; for mappings, ``__iter__()`` should be the same as\n``iterkeys()``; for sequences, it should iterate through the values.\n\nobject.__len__(self)\n\n Called to implement the built-in function ``len()``. Should return\n the length of the object, an integer ``>=`` 0. Also, an object\n that doesn\'t define a ``__nonzero__()`` method and whose\n ``__len__()`` method returns zero is considered to be false in a\n Boolean context.\n\nobject.__getitem__(self, key)\n\n Called to implement evaluation of ``self[key]``. For sequence\n types, the accepted keys should be integers and slice objects.\n Note that the special interpretation of negative indexes (if the\n class wishes to emulate a sequence type) is up to the\n ``__getitem__()`` method. If *key* is of an inappropriate type,\n ``TypeError`` may be raised; if of a value outside the set of\n indexes for the sequence (after any special interpretation of\n negative values), ``IndexError`` should be raised. For mapping\n types, if *key* is missing (not in the container), ``KeyError``\n should be raised.\n\n Note: ``for`` loops expect that an ``IndexError`` will be raised for\n illegal indexes to allow proper detection of the end of the\n sequence.\n\nobject.__setitem__(self, key, value)\n\n Called to implement assignment to ``self[key]``. Same note as for\n ``__getitem__()``. This should only be implemented for mappings if\n the objects support changes to the values for keys, or if new keys\n can be added, or for sequences if elements can be replaced. The\n same exceptions should be raised for improper *key* values as for\n the ``__getitem__()`` method.\n\nobject.__delitem__(self, key)\n\n Called to implement deletion of ``self[key]``. Same note as for\n ``__getitem__()``. This should only be implemented for mappings if\n the objects support removal of keys, or for sequences if elements\n can be removed from the sequence. The same exceptions should be\n raised for improper *key* values as for the ``__getitem__()``\n method.\n\nobject.__iter__(self)\n\n This method is called when an iterator is required for a container.\n This method should return a new iterator object that can iterate\n over all the objects in the container. For mappings, it should\n iterate over the keys of the container, and should also be made\n available as the method ``iterkeys()``.\n\n Iterator objects also need to implement this method; they are\n required to return themselves. For more information on iterator\n objects, see *Iterator Types*.\n\nobject.__reversed__(self)\n\n Called (if present) by the ``reversed()`` built-in to implement\n reverse iteration. It should return a new iterator object that\n iterates over all the objects in the container in reverse order.\n\n If the ``__reversed__()`` method is not provided, the\n ``reversed()`` built-in will fall back to using the sequence\n protocol (``__len__()`` and ``__getitem__()``). Objects that\n support the sequence protocol should only provide\n ``__reversed__()`` if they can provide an implementation that is\n more efficient than the one provided by ``reversed()``.\n\n New in version 2.6.\n\nThe membership test operators (``in`` and ``not in``) are normally\nimplemented as an iteration through a sequence. However, container\nobjects can supply the following special method with a more efficient\nimplementation, which also does not require the object be a sequence.\n\nobject.__contains__(self, item)\n\n Called to implement membership test operators. Should return true\n if *item* is in *self*, false otherwise. For mapping objects, this\n should consider the keys of the mapping rather than the values or\n the key-item pairs.\n\n For objects that don\'t define ``__contains__()``, the membership\n test first tries iteration via ``__iter__()``, then the old\n sequence iteration protocol via ``__getitem__()``, see *this\n section in the language reference*.\n\n\nAdditional methods for emulation of sequence types\n==================================================\n\nThe following optional methods can be defined to further emulate\nsequence objects. Immutable sequences methods should at most only\ndefine ``__getslice__()``; mutable sequences might define all three\nmethods.\n\nobject.__getslice__(self, i, j)\n\n Deprecated since version 2.0: Support slice objects as parameters\n to the ``__getitem__()`` method. (However, built-in types in\n CPython currently still implement ``__getslice__()``. Therefore,\n you have to override it in derived classes when implementing\n slicing.)\n\n Called to implement evaluation of ``self[i:j]``. The returned\n object should be of the same type as *self*. Note that missing *i*\n or *j* in the slice expression are replaced by zero or\n ``sys.maxint``, respectively. If negative indexes are used in the\n slice, the length of the sequence is added to that index. If the\n instance does not implement the ``__len__()`` method, an\n ``AttributeError`` is raised. No guarantee is made that indexes\n adjusted this way are not still negative. Indexes which are\n greater than the length of the sequence are not modified. If no\n ``__getslice__()`` is found, a slice object is created instead, and\n passed to ``__getitem__()`` instead.\n\nobject.__setslice__(self, i, j, sequence)\n\n Called to implement assignment to ``self[i:j]``. Same notes for *i*\n and *j* as for ``__getslice__()``.\n\n This method is deprecated. If no ``__setslice__()`` is found, or\n for extended slicing of the form ``self[i:j:k]``, a slice object is\n created, and passed to ``__setitem__()``, instead of\n ``__setslice__()`` being called.\n\nobject.__delslice__(self, i, j)\n\n Called to implement deletion of ``self[i:j]``. Same notes for *i*\n and *j* as for ``__getslice__()``. This method is deprecated. If no\n ``__delslice__()`` is found, or for extended slicing of the form\n ``self[i:j:k]``, a slice object is created, and passed to\n ``__delitem__()``, instead of ``__delslice__()`` being called.\n\nNotice that these methods are only invoked when a single slice with a\nsingle colon is used, and the slice method is available. For slice\noperations involving extended slice notation, or in absence of the\nslice methods, ``__getitem__()``, ``__setitem__()`` or\n``__delitem__()`` is called with a slice object as argument.\n\nThe following example demonstrate how to make your program or module\ncompatible with earlier versions of Python (assuming that methods\n``__getitem__()``, ``__setitem__()`` and ``__delitem__()`` support\nslice objects as arguments):\n\n class MyClass:\n ...\n def __getitem__(self, index):\n ...\n def __setitem__(self, index, value):\n ...\n def __delitem__(self, index):\n ...\n\n if sys.version_info < (2, 0):\n # They won\'t be defined if version is at least 2.0 final\n\n def __getslice__(self, i, j):\n return self[max(0, i):max(0, j):]\n def __setslice__(self, i, j, seq):\n self[max(0, i):max(0, j):] = seq\n def __delslice__(self, i, j):\n del self[max(0, i):max(0, j):]\n ...\n\nNote the calls to ``max()``; these are necessary because of the\nhandling of negative indices before the ``__*slice__()`` methods are\ncalled. When negative indexes are used, the ``__*item__()`` methods\nreceive them as provided, but the ``__*slice__()`` methods get a\n"cooked" form of the index values. For each negative index value, the\nlength of the sequence is added to the index before calling the method\n(which may still result in a negative index); this is the customary\nhandling of negative indexes by the built-in sequence types, and the\n``__*item__()`` methods are expected to do this as well. However,\nsince they should already be doing that, negative indexes cannot be\npassed in; they must be constrained to the bounds of the sequence\nbefore being passed to the ``__*item__()`` methods. Calling ``max(0,\ni)`` conveniently returns the proper value.\n\n\nEmulating numeric types\n=======================\n\nThe following methods can be defined to emulate numeric objects.\nMethods corresponding to operations that are not supported by the\nparticular kind of number implemented (e.g., bitwise operations for\nnon-integral numbers) should be left undefined.\n\nobject.__add__(self, other)\nobject.__sub__(self, other)\nobject.__mul__(self, other)\nobject.__floordiv__(self, other)\nobject.__mod__(self, other)\nobject.__divmod__(self, other)\nobject.__pow__(self, other[, modulo])\nobject.__lshift__(self, other)\nobject.__rshift__(self, other)\nobject.__and__(self, other)\nobject.__xor__(self, other)\nobject.__or__(self, other)\n\n These methods are called to implement the binary arithmetic\n operations (``+``, ``-``, ``*``, ``//``, ``%``, ``divmod()``,\n ``pow()``, ``**``, ``<<``, ``>>``, ``&``, ``^``, ``|``). For\n instance, to evaluate the expression ``x + y``, where *x* is an\n instance of a class that has an ``__add__()`` method,\n ``x.__add__(y)`` is called. The ``__divmod__()`` method should be\n the equivalent to using ``__floordiv__()`` and ``__mod__()``; it\n should not be related to ``__truediv__()`` (described below). Note\n that ``__pow__()`` should be defined to accept an optional third\n argument if the ternary version of the built-in ``pow()`` function\n is to be supported.\n\n If one of those methods does not support the operation with the\n supplied arguments, it should return ``NotImplemented``.\n\nobject.__div__(self, other)\nobject.__truediv__(self, other)\n\n The division operator (``/``) is implemented by these methods. The\n ``__truediv__()`` method is used when ``__future__.division`` is in\n effect, otherwise ``__div__()`` is used. If only one of these two\n methods is defined, the object will not support division in the\n alternate context; ``TypeError`` will be raised instead.\n\nobject.__radd__(self, other)\nobject.__rsub__(self, other)\nobject.__rmul__(self, other)\nobject.__rdiv__(self, other)\nobject.__rtruediv__(self, other)\nobject.__rfloordiv__(self, other)\nobject.__rmod__(self, other)\nobject.__rdivmod__(self, other)\nobject.__rpow__(self, other)\nobject.__rlshift__(self, other)\nobject.__rrshift__(self, other)\nobject.__rand__(self, other)\nobject.__rxor__(self, other)\nobject.__ror__(self, other)\n\n These methods are called to implement the binary arithmetic\n operations (``+``, ``-``, ``*``, ``/``, ``%``, ``divmod()``,\n ``pow()``, ``**``, ``<<``, ``>>``, ``&``, ``^``, ``|``) with\n reflected (swapped) operands. These functions are only called if\n the left operand does not support the corresponding operation and\n the operands are of different types. [2] For instance, to evaluate\n the expression ``x - y``, where *y* is an instance of a class that\n has an ``__rsub__()`` method, ``y.__rsub__(x)`` is called if\n ``x.__sub__(y)`` returns *NotImplemented*.\n\n Note that ternary ``pow()`` will not try calling ``__rpow__()``\n (the coercion rules would become too complicated).\n\n Note: If the right operand\'s type is a subclass of the left operand\'s\n type and that subclass provides the reflected method for the\n operation, this method will be called before the left operand\'s\n non-reflected method. This behavior allows subclasses to\n override their ancestors\' operations.\n\nobject.__iadd__(self, other)\nobject.__isub__(self, other)\nobject.__imul__(self, other)\nobject.__idiv__(self, other)\nobject.__itruediv__(self, other)\nobject.__ifloordiv__(self, other)\nobject.__imod__(self, other)\nobject.__ipow__(self, other[, modulo])\nobject.__ilshift__(self, other)\nobject.__irshift__(self, other)\nobject.__iand__(self, other)\nobject.__ixor__(self, other)\nobject.__ior__(self, other)\n\n These methods are called to implement the augmented arithmetic\n assignments (``+=``, ``-=``, ``*=``, ``/=``, ``//=``, ``%=``,\n ``**=``, ``<<=``, ``>>=``, ``&=``, ``^=``, ``|=``). These methods\n should attempt to do the operation in-place (modifying *self*) and\n return the result (which could be, but does not have to be,\n *self*). If a specific method is not defined, the augmented\n assignment falls back to the normal methods. For instance, to\n execute the statement ``x += y``, where *x* is an instance of a\n class that has an ``__iadd__()`` method, ``x.__iadd__(y)`` is\n called. If *x* is an instance of a class that does not define a\n ``__iadd__()`` method, ``x.__add__(y)`` and ``y.__radd__(x)`` are\n considered, as with the evaluation of ``x + y``.\n\nobject.__neg__(self)\nobject.__pos__(self)\nobject.__abs__(self)\nobject.__invert__(self)\n\n Called to implement the unary arithmetic operations (``-``, ``+``,\n ``abs()`` and ``~``).\n\nobject.__complex__(self)\nobject.__int__(self)\nobject.__long__(self)\nobject.__float__(self)\n\n Called to implement the built-in functions ``complex()``,\n ``int()``, ``long()``, and ``float()``. Should return a value of\n the appropriate type.\n\nobject.__oct__(self)\nobject.__hex__(self)\n\n Called to implement the built-in functions ``oct()`` and ``hex()``.\n Should return a string value.\n\nobject.__index__(self)\n\n Called to implement ``operator.index()``. Also called whenever\n Python needs an integer object (such as in slicing). Must return\n an integer (int or long).\n\n New in version 2.5.\n\nobject.__coerce__(self, other)\n\n Called to implement "mixed-mode" numeric arithmetic. Should either\n return a 2-tuple containing *self* and *other* converted to a\n common numeric type, or ``None`` if conversion is impossible. When\n the common type would be the type of ``other``, it is sufficient to\n return ``None``, since the interpreter will also ask the other\n object to attempt a coercion (but sometimes, if the implementation\n of the other type cannot be changed, it is useful to do the\n conversion to the other type here). A return value of\n ``NotImplemented`` is equivalent to returning ``None``.\n\n\nCoercion rules\n==============\n\nThis section used to document the rules for coercion. As the language\nhas evolved, the coercion rules have become hard to document\nprecisely; documenting what one version of one particular\nimplementation does is undesirable. Instead, here are some informal\nguidelines regarding coercion. In Python 3, coercion will not be\nsupported.\n\n* If the left operand of a % operator is a string or Unicode object,\n no coercion takes place and the string formatting operation is\n invoked instead.\n\n* It is no longer recommended to define a coercion operation. Mixed-\n mode operations on types that don\'t define coercion pass the\n original arguments to the operation.\n\n* New-style classes (those derived from ``object``) never invoke the\n ``__coerce__()`` method in response to a binary operator; the only\n time ``__coerce__()`` is invoked is when the built-in function\n ``coerce()`` is called.\n\n* For most intents and purposes, an operator that returns\n ``NotImplemented`` is treated the same as one that is not\n implemented at all.\n\n* Below, ``__op__()`` and ``__rop__()`` are used to signify the\n generic method names corresponding to an operator; ``__iop__()`` is\n used for the corresponding in-place operator. For example, for the\n operator \'``+``\', ``__add__()`` and ``__radd__()`` are used for the\n left and right variant of the binary operator, and ``__iadd__()``\n for the in-place variant.\n\n* For objects *x* and *y*, first ``x.__op__(y)`` is tried. If this is\n not implemented or returns ``NotImplemented``, ``y.__rop__(x)`` is\n tried. If this is also not implemented or returns\n ``NotImplemented``, a ``TypeError`` exception is raised. But see\n the following exception:\n\n* Exception to the previous item: if the left operand is an instance\n of a built-in type or a new-style class, and the right operand is an\n instance of a proper subclass of that type or class and overrides\n the base\'s ``__rop__()`` method, the right operand\'s ``__rop__()``\n method is tried *before* the left operand\'s ``__op__()`` method.\n\n This is done so that a subclass can completely override binary\n operators. Otherwise, the left operand\'s ``__op__()`` method would\n always accept the right operand: when an instance of a given class\n is expected, an instance of a subclass of that class is always\n acceptable.\n\n* When either operand type defines a coercion, this coercion is called\n before that type\'s ``__op__()`` or ``__rop__()`` method is called,\n but no sooner. If the coercion returns an object of a different\n type for the operand whose coercion is invoked, part of the process\n is redone using the new object.\n\n* When an in-place operator (like \'``+=``\') is used, if the left\n operand implements ``__iop__()``, it is invoked without any\n coercion. When the operation falls back to ``__op__()`` and/or\n ``__rop__()``, the normal coercion rules apply.\n\n* In ``x + y``, if *x* is a sequence that implements sequence\n concatenation, sequence concatenation is invoked.\n\n* In ``x * y``, if one operand is a sequence that implements sequence\n repetition, and the other is an integer (``int`` or ``long``),\n sequence repetition is invoked.\n\n* Rich comparisons (implemented by methods ``__eq__()`` and so on)\n never use coercion. Three-way comparison (implemented by\n ``__cmp__()``) does use coercion under the same conditions as other\n binary operations use it.\n\n* In the current implementation, the built-in numeric types ``int``,\n ``long``, ``float``, and ``complex`` do not use coercion. All these\n types implement a ``__coerce__()`` method, for use by the built-in\n ``coerce()`` function.\n\n Changed in version 2.7.\n\n\nWith Statement Context Managers\n===============================\n\nNew in version 2.5.\n\nA *context manager* is an object that defines the runtime context to\nbe established when executing a ``with`` statement. The context\nmanager handles the entry into, and the exit from, the desired runtime\ncontext for the execution of the block of code. Context managers are\nnormally invoked using the ``with`` statement (described in section\n*The with statement*), but can also be used by directly invoking their\nmethods.\n\nTypical uses of context managers include saving and restoring various\nkinds of global state, locking and unlocking resources, closing opened\nfiles, etc.\n\nFor more information on context managers, see *Context Manager Types*.\n\nobject.__enter__(self)\n\n Enter the runtime context related to this object. The ``with``\n statement will bind this method\'s return value to the target(s)\n specified in the ``as`` clause of the statement, if any.\n\nobject.__exit__(self, exc_type, exc_value, traceback)\n\n Exit the runtime context related to this object. The parameters\n describe the exception that caused the context to be exited. If the\n context was exited without an exception, all three arguments will\n be ``None``.\n\n If an exception is supplied, and the method wishes to suppress the\n exception (i.e., prevent it from being propagated), it should\n return a true value. Otherwise, the exception will be processed\n normally upon exit from this method.\n\n Note that ``__exit__()`` methods should not reraise the passed-in\n exception; this is the caller\'s responsibility.\n\nSee also:\n\n **PEP 0343** - The "with" statement\n The specification, background, and examples for the Python\n ``with`` statement.\n\n\nSpecial method lookup for old-style classes\n===========================================\n\nFor old-style classes, special methods are always looked up in exactly\nthe same way as any other method or attribute. This is the case\nregardless of whether the method is being looked up explicitly as in\n``x.__getitem__(i)`` or implicitly as in ``x[i]``.\n\nThis behaviour means that special methods may exhibit different\nbehaviour for different instances of a single old-style class if the\nappropriate special attributes are set differently:\n\n >>> class C:\n ... pass\n ...\n >>> c1 = C()\n >>> c2 = C()\n >>> c1.__len__ = lambda: 5\n >>> c2.__len__ = lambda: 9\n >>> len(c1)\n 5\n >>> len(c2)\n 9\n\n\nSpecial method lookup for new-style classes\n===========================================\n\nFor new-style classes, implicit invocations of special methods are\nonly guaranteed to work correctly if defined on an object\'s type, not\nin the object\'s instance dictionary. That behaviour is the reason why\nthe following code raises an exception (unlike the equivalent example\nwith old-style classes):\n\n >>> class C(object):\n ... pass\n ...\n >>> c = C()\n >>> c.__len__ = lambda: 5\n >>> len(c)\n Traceback (most recent call last):\n File "<stdin>", line 1, in <module>\n TypeError: object of type \'C\' has no len()\n\nThe rationale behind this behaviour lies with a number of special\nmethods such as ``__hash__()`` and ``__repr__()`` that are implemented\nby all objects, including type objects. If the implicit lookup of\nthese methods used the conventional lookup process, they would fail\nwhen invoked on the type object itself:\n\n >>> 1 .__hash__() == hash(1)\n True\n >>> int.__hash__() == hash(int)\n Traceback (most recent call last):\n File "<stdin>", line 1, in <module>\n TypeError: descriptor \'__hash__\' of \'int\' object needs an argument\n\nIncorrectly attempting to invoke an unbound method of a class in this\nway is sometimes referred to as \'metaclass confusion\', and is avoided\nby bypassing the instance when looking up special methods:\n\n >>> type(1).__hash__(1) == hash(1)\n True\n >>> type(int).__hash__(int) == hash(int)\n True\n\nIn addition to bypassing any instance attributes in the interest of\ncorrectness, implicit special method lookup generally also bypasses\nthe ``__getattribute__()`` method even of the object\'s metaclass:\n\n >>> class Meta(type):\n ... def __getattribute__(*args):\n ... print "Metaclass getattribute invoked"\n ... return type.__getattribute__(*args)\n ...\n >>> class C(object):\n ... __metaclass__ = Meta\n ... def __len__(self):\n ... return 10\n ... def __getattribute__(*args):\n ... print "Class getattribute invoked"\n ... return object.__getattribute__(*args)\n ...\n >>> c = C()\n >>> c.__len__() # Explicit lookup via instance\n Class getattribute invoked\n 10\n >>> type(c).__len__(c) # Explicit lookup via type\n Metaclass getattribute invoked\n 10\n >>> len(c) # Implicit lookup\n 10\n\nBypassing the ``__getattribute__()`` machinery in this fashion\nprovides significant scope for speed optimisations within the\ninterpreter, at the cost of some flexibility in the handling of\nspecial methods (the special method *must* be set on the class object\nitself in order to be consistently invoked by the interpreter).\n\n-[ Footnotes ]-\n\n[1] It *is* possible in some cases to change an object\'s type, under\n certain controlled conditions. It generally isn\'t a good idea\n though, since it can lead to some very strange behaviour if it is\n handled incorrectly.\n\n[2] For operands of the same type, it is assumed that if the non-\n reflected method (such as ``__add__()``) fails the operation is\n not supported, which is why the reflected method is not called.\n',

Completed in 980 milliseconds

123

Indexes created Sun Aug 12 02:23:12 CEST 2018