Token.h revision 44aa1f397855f130e88e62ffc1029f7f83bb5d2e 
1//===--- Token.h - Token interface ------------------------------*- C++ -*-===//
2//
3//                     The LLVM Compiler Infrastructure
4//
5// This file is distributed under the University of Illinois Open Source
6// License. See LICENSE.TXT for details.
7//
8//===----------------------------------------------------------------------===//
9//
10//  This file defines the Token interface.
11//
12//===----------------------------------------------------------------------===//
13
14#ifndef LLVM_CLANG_TOKEN_H
15#define LLVM_CLANG_TOKEN_H
16
17#include "clang/Basic/TemplateKinds.h"
18#include "clang/Basic/TokenKinds.h"
19#include "clang/Basic/SourceLocation.h"
20#include "clang/Basic/OperatorKinds.h"
21#include <cstdlib>
22
23namespace clang {
24
25class IdentifierInfo;
26
27/// Token - This structure provides full information about a lexed token.
28/// It is not intended to be space efficient, it is intended to return as much
29/// information as possible about each returned token.  This is expected to be
30/// compressed into a smaller form if memory footprint is important.
31///
32/// The parser can create a special "annotation token" representing a stream of
33/// tokens that were parsed and semantically resolved, e.g.: "foo::MyClass<int>"
34/// can be represented by a single typename annotation token that carries
35/// information about the SourceRange of the tokens and the type object.
36class Token {
37  /// The location of the token.
38  SourceLocation Loc;
39
40  // Conceptually these next two fields could be in a union.  However, this
41  // causes gcc 4.2 to pessimize LexTokenInternal, a very performance critical
42  // routine. Keeping as separate members with casts until a more beautiful fix
43  // presents itself.
44
45  /// UintData - This holds either the length of the token text, when
46  /// a normal token, or the end of the SourceRange when an annotation
47  /// token.
48  unsigned UintData;
49
50  /// PtrData - This is a union of four different pointer types, which depends
51  /// on what type of token this is:
52  ///  Identifiers, keywords, etc:
53  ///    This is an IdentifierInfo*, which contains the uniqued identifier
54  ///    spelling.
55  ///  Literals:  isLiteral() returns true.
56  ///    This is a pointer to the start of the token in a text buffer, which
57  ///    may be dirty (have trigraphs / escaped newlines).
58  ///  Annotations (resolved type names, C++ scopes, etc): isAnnotation().
59  ///    This is a pointer to sema-specific data for the annotation token.
60  ///  Other:
61  ///    This is null.
62  void *PtrData;
63
64  /// Kind - The actual flavor of token this is.
65  ///
66  unsigned char Kind; // DON'T make Kind a 'tok::TokenKind';
67                      // MSVC will treat it as a signed char and
68                      // TokenKinds > 127 won't be handled correctly.
69
70  /// Flags - Bits we track about this token, members of the TokenFlags enum.
71  unsigned char Flags;
72public:
73
74  // Various flags set per token:
75  enum TokenFlags {
76    StartOfLine   = 0x01,  // At start of line or only after whitespace.
77    LeadingSpace  = 0x02,  // Whitespace exists before this token.
78    DisableExpand = 0x04,  // This identifier may never be macro expanded.
79    NeedsCleaning = 0x08,   // Contained an escaped newline or trigraph.
80    LeadingEmptyMacro = 0x10 // Empty macro exists before this token.
81  };
82
83  tok::TokenKind getKind() const { return (tok::TokenKind)Kind; }
84  void setKind(tok::TokenKind K) { Kind = K; }
85
86  /// is/isNot - Predicates to check if this token is a specific kind, as in
87  /// "if (Tok.is(tok::l_brace)) {...}".
88  bool is(tok::TokenKind K) const { return Kind == (unsigned) K; }
89  bool isNot(tok::TokenKind K) const { return Kind != (unsigned) K; }
90
91  /// isLiteral - Return true if this is a "literal", like a numeric
92  /// constant, string, etc.
93  bool isLiteral() const {
94    return is(tok::numeric_constant) || is(tok::char_constant) ||
95           is(tok::string_literal) || is(tok::wide_string_literal) ||
96           is(tok::angle_string_literal);
97  }
98
99  bool isAnnotation() const {
100    return is(tok::annot_typename) ||
101           is(tok::annot_cxxscope) ||
102           is(tok::annot_template_id);
103  }
104
105  /// getLocation - Return a source location identifier for the specified
106  /// offset in the current file.
107  SourceLocation getLocation() const { return Loc; }
108  unsigned getLength() const {
109    assert(!isAnnotation() && "Annotation tokens have no length field");
110    return UintData;
111  }
112
113  void setLocation(SourceLocation L) { Loc = L; }
114  void setLength(unsigned Len) {
115    assert(!isAnnotation() && "Annotation tokens have no length field");
116    UintData = Len;
117  }
118
119  SourceLocation getAnnotationEndLoc() const {
120    assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token");
121    return SourceLocation::getFromRawEncoding(UintData);
122  }
123  void setAnnotationEndLoc(SourceLocation L) {
124    assert(isAnnotation() && "Used AnnotEndLocID on non-annotation token");
125    UintData = L.getRawEncoding();
126  }
127
128  SourceLocation getLastLoc() const {
129    return isAnnotation() ? getAnnotationEndLoc() : getLocation();
130  }
131
132  /// getAnnotationRange - SourceRange of the group of tokens that this
133  /// annotation token represents.
134  SourceRange getAnnotationRange() const {
135    return SourceRange(getLocation(), getAnnotationEndLoc());
136  }
137  void setAnnotationRange(SourceRange R) {
138    setLocation(R.getBegin());
139    setAnnotationEndLoc(R.getEnd());
140  }
141
142  const char *getName() const {
143    return tok::getTokenName( (tok::TokenKind) Kind);
144  }
145
146  /// startToken - Reset all flags to cleared.
147  ///
148  void startToken() {
149    Kind = tok::unknown;
150    Flags = 0;
151    PtrData = 0;
152    UintData = 0;
153    Loc = SourceLocation();
154  }
155
156  IdentifierInfo *getIdentifierInfo() const {
157    assert(!isAnnotation() && "Used IdentInfo on annotation token!");
158    if (isLiteral()) return 0;
159    return (IdentifierInfo*) PtrData;
160  }
161  void setIdentifierInfo(IdentifierInfo *II) {
162    PtrData = (void*) II;
163  }
164
165  /// getLiteralData - For a literal token (numeric constant, string, etc), this
166  /// returns a pointer to the start of it in the text buffer if known, null
167  /// otherwise.
168  const char *getLiteralData() const {
169    assert(isLiteral() && "Cannot get literal data of non-literal");
170    return reinterpret_cast<const char*>(PtrData);
171  }
172  void setLiteralData(const char *Ptr) {
173    assert(isLiteral() && "Cannot set literal data of non-literal");
174    PtrData = const_cast<char*>(Ptr);
175  }
176
177  void *getAnnotationValue() const {
178    assert(isAnnotation() && "Used AnnotVal on non-annotation token");
179    return PtrData;
180  }
181  void setAnnotationValue(void *val) {
182    assert(isAnnotation() && "Used AnnotVal on non-annotation token");
183    PtrData = val;
184  }
185
186  /// setFlag - Set the specified flag.
187  void setFlag(TokenFlags Flag) {
188    Flags |= Flag;
189  }
190
191  /// clearFlag - Unset the specified flag.
192  void clearFlag(TokenFlags Flag) {
193    Flags &= ~Flag;
194  }
195
196  /// getFlags - Return the internal represtation of the flags.
197  ///  Only intended for low-level operations such as writing tokens to
198  //   disk.
199  unsigned getFlags() const {
200    return Flags;
201  }
202
203  /// setFlagValue - Set a flag to either true or false.
204  void setFlagValue(TokenFlags Flag, bool Val) {
205    if (Val)
206      setFlag(Flag);
207    else
208      clearFlag(Flag);
209  }
210
211  /// isAtStartOfLine - Return true if this token is at the start of a line.
212  ///
213  bool isAtStartOfLine() const { return (Flags & StartOfLine) ? true : false; }
214
215  /// hasLeadingSpace - Return true if this token has whitespace before it.
216  ///
217  bool hasLeadingSpace() const { return (Flags & LeadingSpace) ? true : false; }
218
219  /// isExpandDisabled - Return true if this identifier token should never
220  /// be expanded in the future, due to C99 6.10.3.4p2.
221  bool isExpandDisabled() const {
222    return (Flags & DisableExpand) ? true : false;
223  }
224
225  /// isObjCAtKeyword - Return true if we have an ObjC keyword identifier.
226  bool isObjCAtKeyword(tok::ObjCKeywordKind objcKey) const;
227
228  /// getObjCKeywordID - Return the ObjC keyword kind.
229  tok::ObjCKeywordKind getObjCKeywordID() const;
230
231  /// needsCleaning - Return true if this token has trigraphs or escaped
232  /// newlines in it.
233  ///
234  bool needsCleaning() const { return (Flags & NeedsCleaning) ? true : false; }
235
236  /// \brief Return true if this token has an empty macro before it.
237  ///
238  bool hasLeadingEmptyMacro() const {
239    return (Flags & LeadingEmptyMacro) ? true : false;
240  }
241
242};
243
244/// PPConditionalInfo - Information about the conditional stack (#if directives)
245/// currently active.
246struct PPConditionalInfo {
247  /// IfLoc - Location where the conditional started.
248  ///
249  SourceLocation IfLoc;
250
251  /// WasSkipping - True if this was contained in a skipping directive, e.g.
252  /// in a "#if 0" block.
253  bool WasSkipping;
254
255  /// FoundNonSkip - True if we have emitted tokens already, and now we're in
256  /// an #else block or something.  Only useful in Skipping blocks.
257  bool FoundNonSkip;
258
259  /// FoundElse - True if we've seen a #else in this block.  If so,
260  /// #elif/#else directives are not allowed.
261  bool FoundElse;
262};
263
264}  // end namespace clang
265
266namespace llvm {
267  template <>
268  struct isPodLike<clang::Token> { static const bool value = true; };
269}  // end namespace llvm
270
271#endif
272