Preprocessor.h revision a9e274c01ebae45629d93aaa07be450fb77dd3cb
1d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org//===--- Preprocessor.h - C Language Family Preprocessor --------*- C++ -*-===//
2d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org//
3d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org//                     The LLVM Compiler Infrastructure
4d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org//
5d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org// This file is distributed under the University of Illinois Open Source
6d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org// License. See LICENSE.TXT for details.
7d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org//
87f9bebe3906fb30f0632f9d010754d12dd6c02d7yusukes@chromium.org//===----------------------------------------------------------------------===//
9d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org//
10cd35d0324e9cb7d46625ea1c0990dc75c69fc835bashi@chromium.org//  This file defines the Preprocessor interface.
11d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org//
12d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org//===----------------------------------------------------------------------===//
13d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
1474343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org#ifndef LLVM_CLANG_LEX_PREPROCESSOR_H
15d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org#define LLVM_CLANG_LEX_PREPROCESSOR_H
16d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
17d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org#include "clang/Lex/Lexer.h"
1874343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org#include "clang/Lex/PPCallbacks.h"
1974343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org#include "clang/Lex/TokenLexer.h"
2074343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org#include "clang/Basic/IdentifierTable.h"
2174343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org#include "clang/Basic/SourceLocation.h"
2274343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org#include "llvm/ADT/DenseMap.h"
23d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
24d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgnamespace clang {
25d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
26d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgclass SourceManager;
27d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgclass FileManager;
28d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgclass FileEntry;
29d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgclass HeaderSearch;
30d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgclass PragmaNamespace;
31d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgclass PragmaHandler;
32d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgclass ScratchBuffer;
33d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgclass TargetInfo;
34d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgclass PPCallbacks;
35d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgclass DirectoryLookup;
36d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
37d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org/// Preprocessor - This object engages in a tight little dance with the lexer to
38d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org/// efficiently preprocess tokens.  Lexers know only about tokens within a
39d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org/// single source file, and don't know anything about preprocessor-level issues
40d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org/// like the #include stack, token expansion, etc.
41d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org///
42d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgclass Preprocessor {
43d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  Diagnostic        &Diags;
44d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  const LangOptions &Features;
45d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  TargetInfo        &Target;
46d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  FileManager       &FileMgr;
47d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  SourceManager     &SourceMgr;
48d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  ScratchBuffer     *ScratchBuf;
49d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  HeaderSearch      &HeaderInfo;
50d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
51d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// Identifiers for builtin macros and other builtins.
52d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  IdentifierInfo *Ident__LINE__, *Ident__FILE__;   // __LINE__, __FILE__
53d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  IdentifierInfo *Ident__DATE__, *Ident__TIME__;   // __DATE__, __TIME__
54d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  IdentifierInfo *Ident__INCLUDE_LEVEL__;          // __INCLUDE_LEVEL__
55d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  IdentifierInfo *Ident__BASE_FILE__;              // __BASE_FILE__
56d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  IdentifierInfo *Ident__TIMESTAMP__;              // __TIMESTAMP__
57d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  IdentifierInfo *Ident_Pragma, *Ident__VA_ARGS__; // _Pragma, __VA_ARGS__
58d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
59d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  SourceLocation DATELoc, TIMELoc;
60d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
61d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  enum {
62d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org    /// MaxIncludeStackDepth - Maximum depth of #includes.
63d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org    MaxAllowedIncludeStackDepth = 200
64d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  };
65d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
66d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  // State that is set before the preprocessor begins.
67d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  bool KeepComments : 1;
68d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  bool KeepMacroComments : 1;
69d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
70d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  // State that changes while the preprocessor runs:
71d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  bool DisableMacroExpansion : 1;  // True if macro expansion is disabled.
72d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  bool InMacroArgs : 1;            // True if parsing fn macro invocation args.
73d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
74d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// CacheTokens - True when the lexed tokens are cached for backtracking.
75d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  bool CacheTokens : 1;
76d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
77d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// Identifiers - This is mapping/lookup information for all identifiers in
78d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// the program, including program keywords.
79d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  IdentifierTable Identifiers;
80d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
81d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// Selectors - This table contains all the selectors in the program. Unlike
82d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// IdentifierTable above, this table *isn't* populated by the preprocessor.
83d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// It is declared/instantiated here because it's role/lifetime is
84d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// conceptually similar the IdentifierTable. In addition, the current control
85d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// flow (in clang::ParseAST()), make it convenient to put here.
86d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// FIXME: Make sure the lifetime of Identifiers/Selectors *isn't* tied to
87d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// the lifetime fo the preprocessor.
88d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  SelectorTable Selectors;
89d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
90d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// PragmaHandlers - This tracks all of the pragmas that the client registered
91d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// with this preprocessor.
92d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  PragmaNamespace *PragmaHandlers;
93d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
94d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// CurLexer - This is the current top of the stack that we're lexing from if
95d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// not expanding a macro.  One of CurLexer and CurTokenLexer must be null.
96d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  Lexer *CurLexer;
97d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
98d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// CurLookup - The DirectoryLookup structure used to find the current
99d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// FileEntry, if CurLexer is non-null and if applicable.  This allows us to
100d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// implement #include_next and find directory-specific properties.
101d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  const DirectoryLookup *CurDirLookup;
102d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
103d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// CurTokenLexer - This is the current macro we are expanding, if we are
104d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// expanding a macro.  One of CurLexer and CurTokenLexer must be null.
105d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  TokenLexer *CurTokenLexer;
106d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
107d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// IncludeMacroStack - This keeps track of the stack of files currently
108d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// #included, and macros currently being expanded from, not counting
109d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// CurLexer/CurTokenLexer.
110d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  struct IncludeStackInfo {
111d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org    Lexer *TheLexer;
11293aedf7d27cb93f93077f9f7f758e830a392bdfcbashi@google.com    const DirectoryLookup *TheDirLookup;
11393aedf7d27cb93f93077f9f7f758e830a392bdfcbashi@google.com    TokenLexer *TheTokenLexer;
11493aedf7d27cb93f93077f9f7f758e830a392bdfcbashi@google.com    IncludeStackInfo(Lexer *L, const DirectoryLookup *D, TokenLexer *TL)
11593aedf7d27cb93f93077f9f7f758e830a392bdfcbashi@google.com      : TheLexer(L), TheDirLookup(D), TheTokenLexer(TL) {
11693aedf7d27cb93f93077f9f7f758e830a392bdfcbashi@google.com    }
11793aedf7d27cb93f93077f9f7f758e830a392bdfcbashi@google.com  };
11893aedf7d27cb93f93077f9f7f758e830a392bdfcbashi@google.com  std::vector<IncludeStackInfo> IncludeMacroStack;
11993aedf7d27cb93f93077f9f7f758e830a392bdfcbashi@google.com
12093aedf7d27cb93f93077f9f7f758e830a392bdfcbashi@google.com  /// Callbacks - These are actions invoked when some preprocessor activity is
12193aedf7d27cb93f93077f9f7f758e830a392bdfcbashi@google.com  /// encountered (e.g. a file is #included, etc).
12293aedf7d27cb93f93077f9f7f758e830a392bdfcbashi@google.com  PPCallbacks *Callbacks;
123d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
124d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// Macros - For each IdentifierInfo with 'HasMacro' set, we keep a mapping
125d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// to the actual definition of the macro.
126d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  llvm::DenseMap<IdentifierInfo*, MacroInfo*> Macros;
127d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
128d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  // Various statistics we track for performance analysis.
129d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  unsigned NumDirectives, NumIncluded, NumDefined, NumUndefined, NumPragma;
130d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  unsigned NumIf, NumElse, NumEndif;
131d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  unsigned NumEnteredSourceFiles, MaxIncludeStackDepth;
132d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  unsigned NumMacroExpanded, NumFnMacroExpanded, NumBuiltinMacroExpanded;
133d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  unsigned NumFastMacroExpanded, NumTokenPaste, NumFastTokenPaste;
134d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  unsigned NumSkipped;
135d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
136d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// Predefines - This string is the predefined macros that preprocessor
137d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// should use from the command line etc.
138d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  std::string Predefines;
139d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
140d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// TokenLexerCache - Cache macro expanders to reduce malloc traffic.
141d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  enum { TokenLexerCacheSize = 8 };
142d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  unsigned NumCachedTokenLexers;
143d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  TokenLexer *TokenLexerCache[TokenLexerCacheSize];
144d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
145d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  // Cached tokens state.
146d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
147d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  typedef std::vector<Token> CachedTokensTy;
148d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
149d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// CachedTokens - Cached tokens are stored here when we do backtracking or
150d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// lookahead. They are "lexed" by the CachingLex() method.
151d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  CachedTokensTy CachedTokens;
152d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
153d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// CachedLexPos - The position of the cached token that CachingLex() should
154d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// "lex" next. If it points beyond the CachedTokens vector, it means that
1556263d06109011f6cff761aed14ded117c8fa1836yusukes@chromium.org  /// a normal Lex() should be invoked.
156d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  CachedTokensTy::size_type CachedLexPos;
157d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
158d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// BacktrackPositions - Stack of backtrack positions, allowing nested
159d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// backtracks. The EnableBacktrackAtThisPos() method pushes a position to
160d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// indicate where CachedLexPos should be set when the BackTrack() method is
161d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// invoked (at which point the last position is popped).
162d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  std::vector<CachedTokensTy::size_type> BacktrackPositions;
163d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
164d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.orgpublic:
165d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  Preprocessor(Diagnostic &diags, const LangOptions &opts, TargetInfo &target,
166d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org               SourceManager &SM, HeaderSearch &Headers);
16774343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org  ~Preprocessor();
16874343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org
16974343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org  Diagnostic &getDiagnostics() const { return Diags; }
17074343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org  const LangOptions &getLangOptions() const { return Features; }
17174343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org  TargetInfo &getTargetInfo() const { return Target; }
17274343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org  FileManager &getFileManager() const { return FileMgr; }
17374343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org  SourceManager &getSourceManager() const { return SourceMgr; }
17474343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org  HeaderSearch &getHeaderSearchInfo() const { return HeaderInfo; }
17574343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org
176a416dc6e594169d1f809a08efbcf6de831dff291ksakamoto@chromium.org  IdentifierTable &getIdentifierTable() { return Identifiers; }
177a416dc6e594169d1f809a08efbcf6de831dff291ksakamoto@chromium.org  SelectorTable &getSelectorTable() { return Selectors; }
178a416dc6e594169d1f809a08efbcf6de831dff291ksakamoto@chromium.org
179a416dc6e594169d1f809a08efbcf6de831dff291ksakamoto@chromium.org  inline FullSourceLoc getFullLoc(SourceLocation Loc) const {
180a416dc6e594169d1f809a08efbcf6de831dff291ksakamoto@chromium.org    return FullSourceLoc(Loc, getSourceManager());
181a416dc6e594169d1f809a08efbcf6de831dff291ksakamoto@chromium.org  }
182a416dc6e594169d1f809a08efbcf6de831dff291ksakamoto@chromium.org
18374343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org  /// SetCommentRetentionState - Control whether or not the preprocessor retains
18474343ba7bd420519a9a4f7b22977d7771286f6f8ksakamoto@chromium.org  /// comments in output.
185d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  void SetCommentRetentionState(bool KeepComments, bool KeepMacroComments) {
186d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org    this->KeepComments = KeepComments | KeepMacroComments;
187d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org    this->KeepMacroComments = KeepMacroComments;
188d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  }
189d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
190d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  bool getCommentRetentionState() const { return KeepComments; }
19100b790a7ad0d65b066a61760f58e7dbfc055cd2dbashi@google.com
192d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// isCurrentLexer - Return true if we are lexing directly from the specified
193ced7112cb41d09cd344bbe4b20459049d6039491bashi@chromium.org  /// lexer.
194a574866c0e4b1539098e64422f59d01c80cea6cfbashi@chromium.org  bool isCurrentLexer(const Lexer *L) const {
195d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org    return CurLexer == L;
196d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  }
197d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
198d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// getCurrentLexer - Return the current file lexer being lexed from.  Note
199f12575f6e5a4b053188d4e205ae6ceb95a60fb2cyusukes@chromium.org  /// that this ignores any potentially active macro expansions and _Pragma
200d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// expansions going on at the time.
201d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  Lexer *getCurrentFileLexer() const;
202eea48361b1ffcc1bded0ba25c8f747e634cd8e51ksakamoto@chromium.org
203d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// getPPCallbacks/setPPCallbacks - Accessors for preprocessor callbacks.
204d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// Note that this class takes ownership of any PPCallbacks object given to
205d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// it.
206d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  PPCallbacks *getPPCallbacks() const { return Callbacks; }
207d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  void setPPCallbacks(PPCallbacks *C) {
208d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org    delete Callbacks;
2094dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org    Callbacks = C;
2104dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org  }
2114dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org
212d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// getMacroInfo - Given an identifier, return the MacroInfo it is #defined to
213d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// or null if it isn't #define'd.
214d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  MacroInfo *getMacroInfo(IdentifierInfo *II) const {
215d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org    return II->hasMacroDefinition() ? Macros.find(II)->second : 0;
216d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  }
217d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
218d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// setMacroInfo - Specify a macro for this identifier.
219d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  ///
220d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  void setMacroInfo(IdentifierInfo *II, MacroInfo *MI);
221d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
222d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  const std::string &getPredefines() const { return Predefines; }
223d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// setPredefines - Set the predefines for this Preprocessor.  These
224d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// predefines are automatically injected when parsing the main file.
225d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  void setPredefines(const char *P) { Predefines = P; }
226d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  void setPredefines(const std::string &P) { Predefines = P; }
227d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
228d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// getIdentifierInfo - Return information about the specified preprocessor
229d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// identifier token.  The version of this method that takes two character
230d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// pointers is preferred unless the identifier is already available as a
231d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// string (this avoids allocation and copying of memory to construct an
232d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// std::string).
233d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  IdentifierInfo *getIdentifierInfo(const char *NameStart,
234d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org                                    const char *NameEnd) {
2354dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org    return &Identifiers.get(NameStart, NameEnd);
2364dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org  }
2374dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org  IdentifierInfo *getIdentifierInfo(const char *NameStr) {
2384dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org    return getIdentifierInfo(NameStr, NameStr+strlen(NameStr));
2394dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org  }
2404dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org
2414dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org  /// AddPragmaHandler - Add the specified pragma handler to the preprocessor.
2424dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org  /// If 'Namespace' is non-null, then it is a token required to exist on the
2434dcad600f99d85201d9db3cb8bee166ec2aaca85bashi@chromium.org  /// pragma line before the pragma string starts, e.g. "STDC" or "GCC".
244d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  void AddPragmaHandler(const char *Namespace, PragmaHandler *Handler);
245d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org
246d257d186ae2a08042a412824678f98241a1a4f3cyusukes@chromium.org  /// EnterMainSourceFile - Enter the specified FileID as the main source file,
247  /// which implicitly adds the builtin defines etc.
248  void EnterMainSourceFile();
249
250  /// EnterSourceFile - Add a source file to the top of the include stack and
251  /// start lexing tokens from it instead of the current buffer.  If isMainFile
252  /// is true, this is the main file for the translation unit.
253  void EnterSourceFile(unsigned CurFileID, const DirectoryLookup *Dir);
254
255  /// EnterMacro - Add a Macro to the top of the include stack and start lexing
256  /// tokens from it instead of the current buffer.  Args specifies the
257  /// tokens input to a function-like macro.
258  void EnterMacro(Token &Identifier, MacroArgs *Args);
259
260  /// EnterTokenStream - Add a "macro" context to the top of the include stack,
261  /// which will cause the lexer to start returning the specified tokens.
262  ///
263  /// If DisableMacroExpansion is true, tokens lexed from the token stream will
264  /// not be subject to further macro expansion.  Otherwise, these tokens will
265  /// be re-macro-expanded when/if expansion is enabled.
266  ///
267  /// If OwnsTokens is false, this method assumes that the specified stream of
268  /// tokens has a permanent owner somewhere, so they do not need to be copied.
269  /// If it is true, it assumes the array of tokens is allocated with new[] and
270  /// must be freed.
271  ///
272  void EnterTokenStream(const Token *Toks, unsigned NumToks,
273                        bool DisableMacroExpansion, bool OwnsTokens);
274
275  /// RemoveTopOfLexerStack - Pop the current lexer/macro exp off the top of the
276  /// lexer stack.  This should only be used in situations where the current
277  /// state of the top-of-stack lexer is known.
278  void RemoveTopOfLexerStack();
279
280  /// EnableBacktrackAtThisPos - From the point that this method is called, and
281  /// until DisableBacktrack() or Backtrack() is called, the Preprocessor keeps
282  /// track of the lexed tokens so that a subsequent Backtrack() call will make
283  /// the Preprocessor re-lex the same tokens.
284  ///
285  /// Nested backtracks are allowed, meaning that EnableBacktrackAtThisPos can
286  /// be called multiple times and DisableBacktrack/Backtrack calls will be
287  /// combined with the EnableBacktrackAtThisPos calls in reverse order.
288  ///
289  /// NOTE: *DO NOT* forget to call either DisableBacktrack() or Backtrack() at
290  /// some point after EnableBacktrackAtThisPos. If you don't, caching of tokens
291  /// will continue indefinitely.
292  ///
293  void EnableBacktrackAtThisPos();
294
295  /// DisableBacktrack - Disable the last EnableBacktrackAtThisPos() call.
296  void DisableBacktrack();
297
298  /// Backtrack - Make Preprocessor re-lex the tokens that were lexed since
299  /// EnableBacktrackAtThisPos() was previously called.
300  void Backtrack();
301
302  /// isBacktrackEnabled - True if EnableBacktrackAtThisPos() was called and
303  /// caching of tokens is on.
304  bool isBacktrackEnabled() const { return CacheTokens; }
305
306  /// Lex - To lex a token from the preprocessor, just pull a token from the
307  /// current lexer or macro object.
308  void Lex(Token &Result) {
309    if (CurLexer)
310      CurLexer->Lex(Result);
311    else if (CurTokenLexer)
312      CurTokenLexer->Lex(Result);
313    else
314      CachingLex(Result);
315  }
316
317  /// LexNonComment - Lex a token.  If it's a comment, keep lexing until we get
318  /// something not a comment.  This is useful in -E -C mode where comments
319  /// would foul up preprocessor directive handling.
320  void LexNonComment(Token &Result) {
321    do
322      Lex(Result);
323    while (Result.getKind() == tok::comment);
324  }
325
326  /// LexUnexpandedToken - This is just like Lex, but this disables macro
327  /// expansion of identifier tokens.
328  void LexUnexpandedToken(Token &Result) {
329    // Disable macro expansion.
330    bool OldVal = DisableMacroExpansion;
331    DisableMacroExpansion = true;
332    // Lex the token.
333    Lex(Result);
334
335    // Reenable it.
336    DisableMacroExpansion = OldVal;
337  }
338
339  /// LookAhead - This peeks ahead N tokens and returns that token without
340  /// consuming any tokens.  LookAhead(0) returns the next token that would be
341  /// returned by Lex(), LookAhead(1) returns the token after it, etc.  This
342  /// returns normal tokens after phase 5.  As such, it is equivalent to using
343  /// 'Lex', not 'LexUnexpandedToken'.
344  const Token &LookAhead(unsigned N) {
345    if (CachedLexPos + N < CachedTokens.size())
346      return CachedTokens[CachedLexPos+N];
347    else
348      return PeekAhead(N+1);
349  }
350
351  /// Diag - Forwarding function for diagnostics.  This emits a diagnostic at
352  /// the specified Token's location, translating the token's start
353  /// position in the current buffer into a SourcePosition object for rendering.
354  void Diag(SourceLocation Loc, unsigned DiagID);
355  void Diag(SourceLocation Loc, unsigned DiagID, const std::string &Msg);
356  void Diag(SourceLocation Loc, unsigned DiagID, const std::string &Msg,
357            const SourceRange &R1, const SourceRange &R2);
358  void Diag(SourceLocation Loc, unsigned DiagID, const SourceRange &R);
359  void Diag(SourceLocation Loc, unsigned DiagID, const SourceRange &R1,
360            const SourceRange &R2);
361  void Diag(const Token &Tok, unsigned DiagID) {
362    Diag(Tok.getLocation(), DiagID);
363  }
364  void Diag(const Token &Tok, unsigned DiagID, const std::string &Msg) {
365    Diag(Tok.getLocation(), DiagID, Msg);
366  }
367
368  /// getSpelling() - Return the 'spelling' of the Tok token.  The spelling of a
369  /// token is the characters used to represent the token in the source file
370  /// after trigraph expansion and escaped-newline folding.  In particular, this
371  /// wants to get the true, uncanonicalized, spelling of things like digraphs
372  /// UCNs, etc.
373  std::string getSpelling(const Token &Tok) const;
374
375  /// getSpelling - This method is used to get the spelling of a token into a
376  /// preallocated buffer, instead of as an std::string.  The caller is required
377  /// to allocate enough space for the token, which is guaranteed to be at least
378  /// Tok.getLength() bytes long.  The length of the actual result is returned.
379  ///
380  /// Note that this method may do two possible things: it may either fill in
381  /// the buffer specified with characters, or it may *change the input pointer*
382  /// to point to a constant buffer with the data already in it (avoiding a
383  /// copy).  The caller is not allowed to modify the returned buffer pointer
384  /// if an internal buffer is returned.
385  unsigned getSpelling(const Token &Tok, const char *&Buffer) const;
386
387
388  /// CreateString - Plop the specified string into a scratch buffer and return
389  /// a location for it.  If specified, the source location provides a source
390  /// location for the token.
391  SourceLocation CreateString(const char *Buf, unsigned Len,
392                              SourceLocation SourceLoc = SourceLocation());
393
394  /// DumpToken - Print the token to stderr, used for debugging.
395  ///
396  void DumpToken(const Token &Tok, bool DumpFlags = false) const;
397  void DumpLocation(SourceLocation Loc) const;
398  void DumpMacro(const MacroInfo &MI) const;
399
400  /// AdvanceToTokenCharacter - Given a location that specifies the start of a
401  /// token, return a new location that specifies a character within the token.
402  SourceLocation AdvanceToTokenCharacter(SourceLocation TokStart,unsigned Char);
403
404  /// IncrementPasteCounter - Increment the counters for the number of token
405  /// paste operations performed.  If fast was specified, this is a 'fast paste'
406  /// case we handled.
407  ///
408  void IncrementPasteCounter(bool isFast) {
409    if (isFast)
410      ++NumFastTokenPaste;
411    else
412      ++NumTokenPaste;
413  }
414
415  void PrintStats();
416
417  /// HandleMicrosoftCommentPaste - When the macro expander pastes together a
418  /// comment (/##/) in microsoft mode, this method handles updating the current
419  /// state, returning the token on the next source line.
420  void HandleMicrosoftCommentPaste(Token &Tok);
421
422  //===--------------------------------------------------------------------===//
423  // Preprocessor callback methods.  These are invoked by a lexer as various
424  // directives and events are found.
425
426  /// LookUpIdentifierInfo - Given a tok::identifier token, look up the
427  /// identifier information for the token and install it into the token.
428  IdentifierInfo *LookUpIdentifierInfo(Token &Identifier,
429                                       const char *BufPtr = 0);
430
431  /// HandleIdentifier - This callback is invoked when the lexer reads an
432  /// identifier and has filled in the tokens IdentifierInfo member.  This
433  /// callback potentially macro expands it or turns it into a named token (like
434  /// 'for').
435  void HandleIdentifier(Token &Identifier);
436
437
438  /// HandleEndOfFile - This callback is invoked when the lexer hits the end of
439  /// the current file.  This either returns the EOF token and returns true, or
440  /// pops a level off the include stack and returns false, at which point the
441  /// client should call lex again.
442  bool HandleEndOfFile(Token &Result, bool isEndOfMacro = false);
443
444  /// HandleEndOfTokenLexer - This callback is invoked when the current
445  /// TokenLexer hits the end of its token stream.
446  bool HandleEndOfTokenLexer(Token &Result);
447
448  /// HandleDirective - This callback is invoked when the lexer sees a # token
449  /// at the start of a line.  This consumes the directive, modifies the
450  /// lexer/preprocessor state, and advances the lexer(s) so that the next token
451  /// read is the correct one.
452  void HandleDirective(Token &Result);
453
454  /// CheckEndOfDirective - Ensure that the next token is a tok::eom token.  If
455  /// not, emit a diagnostic and consume up until the eom.
456  void CheckEndOfDirective(const char *Directive);
457private:
458  /// isInPrimaryFile - Return true if we're in the top-level file, not in a
459  /// #include.
460  bool isInPrimaryFile() const;
461
462  /// isSystemHeader - Return true if F is a system header.
463  bool isSystemHeader(const FileEntry* F) const;
464
465  /// DiscardUntilEndOfDirective - Read and discard all tokens remaining on the
466  /// current line until the tok::eom token is found.
467  void DiscardUntilEndOfDirective();
468
469  /// ReadMacroName - Lex and validate a macro name, which occurs after a
470  /// #define or #undef.  This emits a diagnostic, sets the token kind to eom,
471  /// and discards the rest of the macro line if the macro name is invalid.
472  void ReadMacroName(Token &MacroNameTok, char isDefineUndef = 0);
473
474  /// ReadMacroDefinitionArgList - The ( starting an argument list of a macro
475  /// definition has just been read.  Lex the rest of the arguments and the
476  /// closing ), updating MI with what we learn.  Return true if an error occurs
477  /// parsing the arg list.
478  bool ReadMacroDefinitionArgList(MacroInfo *MI);
479
480  /// SkipExcludedConditionalBlock - We just read a #if or related directive and
481  /// decided that the subsequent tokens are in the #if'd out portion of the
482  /// file.  Lex the rest of the file, until we see an #endif.  If
483  /// FoundNonSkipPortion is true, then we have already emitted code for part of
484  /// this #if directive, so #else/#elif blocks should never be entered. If
485  /// FoundElse is false, then #else directives are ok, if not, then we have
486  /// already seen one so a #else directive is a duplicate.  When this returns,
487  /// the caller can lex the first valid token.
488  void SkipExcludedConditionalBlock(SourceLocation IfTokenLoc,
489                                    bool FoundNonSkipPortion, bool FoundElse);
490
491  /// EvaluateDirectiveExpression - Evaluate an integer constant expression that
492  /// may occur after a #if or #elif directive and return it as a bool.  If the
493  /// expression is equivalent to "!defined(X)" return X in IfNDefMacro.
494  bool EvaluateDirectiveExpression(IdentifierInfo *&IfNDefMacro);
495
496  /// RegisterBuiltinPragmas - Install the standard preprocessor pragmas:
497  /// #pragma GCC poison/system_header/dependency and #pragma once.
498  void RegisterBuiltinPragmas();
499
500  /// RegisterBuiltinMacros - Register builtin macros, such as __LINE__ with the
501  /// identifier table.
502  void RegisterBuiltinMacros();
503  IdentifierInfo *RegisterBuiltinMacro(const char *Name);
504
505  /// HandleMacroExpandedIdentifier - If an identifier token is read that is to
506  /// be expanded as a macro, handle it and return the next token as 'Tok'.  If
507  /// the macro should not be expanded return true, otherwise return false.
508  bool HandleMacroExpandedIdentifier(Token &Tok, MacroInfo *MI);
509
510  /// isNextPPTokenLParen - Determine whether the next preprocessor token to be
511  /// lexed is a '('.  If so, consume the token and return true, if not, this
512  /// method should have no observable side-effect on the lexed tokens.
513  bool isNextPPTokenLParen();
514
515  /// ReadFunctionLikeMacroArgs - After reading "MACRO(", this method is
516  /// invoked to read all of the formal arguments specified for the macro
517  /// invocation.  This returns null on error.
518  MacroArgs *ReadFunctionLikeMacroArgs(Token &MacroName, MacroInfo *MI);
519
520  /// ExpandBuiltinMacro - If an identifier token is read that is to be expanded
521  /// as a builtin macro, handle it and return the next token as 'Tok'.
522  void ExpandBuiltinMacro(Token &Tok);
523
524  /// Handle_Pragma - Read a _Pragma directive, slice it up, process it, then
525  /// return the first token after the directive.  The _Pragma token has just
526  /// been read into 'Tok'.
527  void Handle_Pragma(Token &Tok);
528
529
530  /// EnterSourceFileWithLexer - Add a lexer to the top of the include stack and
531  /// start lexing tokens from it instead of the current buffer.
532  void EnterSourceFileWithLexer(Lexer *TheLexer, const DirectoryLookup *Dir);
533
534  /// GetIncludeFilenameSpelling - Turn the specified lexer token into a fully
535  /// checked and spelled filename, e.g. as an operand of #include. This returns
536  /// true if the input filename was in <>'s or false if it were in ""'s.  The
537  /// caller is expected to provide a buffer that is large enough to hold the
538  /// spelling of the filename, but is also expected to handle the case when
539  /// this method decides to use a different buffer.
540  bool GetIncludeFilenameSpelling(SourceLocation Loc,
541                                  const char *&BufStart, const char *&BufEnd);
542
543  /// LookupFile - Given a "foo" or <foo> reference, look up the indicated file,
544  /// return null on failure.  isAngled indicates whether the file reference is
545  /// for system #include's or not (i.e. using <> instead of "").
546  const FileEntry *LookupFile(const char *FilenameStart,const char *FilenameEnd,
547                              bool isAngled, const DirectoryLookup *FromDir,
548                              const DirectoryLookup *&CurDir);
549
550  //===--------------------------------------------------------------------===//
551  // Caching stuff.
552  void CachingLex(Token &Result);
553  bool InCachingLexMode() const { return CurLexer == 0 && CurTokenLexer == 0; }
554  void EnterCachingLexMode();
555  void ExitCachingLexMode() {
556    if (InCachingLexMode())
557      RemoveTopOfLexerStack();
558  }
559  const Token &PeekAhead(unsigned N);
560
561  //===--------------------------------------------------------------------===//
562  /// Handle*Directive - implement the various preprocessor directives.  These
563  /// should side-effect the current preprocessor object so that the next call
564  /// to Lex() will return the appropriate token next.
565
566  void HandleUserDiagnosticDirective(Token &Tok, bool isWarning);
567  void HandleIdentSCCSDirective(Token &Tok);
568
569  // File inclusion.
570  void HandleIncludeDirective(Token &Tok,
571                              const DirectoryLookup *LookupFrom = 0,
572                              bool isImport = false);
573  void HandleIncludeNextDirective(Token &Tok);
574  void HandleImportDirective(Token &Tok);
575
576  // Macro handling.
577  void HandleDefineDirective(Token &Tok);
578  void HandleUndefDirective(Token &Tok);
579  // HandleAssertDirective(Token &Tok);
580  // HandleUnassertDirective(Token &Tok);
581
582  // Conditional Inclusion.
583  void HandleIfdefDirective(Token &Tok, bool isIfndef,
584                            bool ReadAnyTokensBeforeDirective);
585  void HandleIfDirective(Token &Tok, bool ReadAnyTokensBeforeDirective);
586  void HandleEndifDirective(Token &Tok);
587  void HandleElseDirective(Token &Tok);
588  void HandleElifDirective(Token &Tok);
589
590  // Pragmas.
591  void HandlePragmaDirective();
592public:
593  void HandlePragmaOnce(Token &OnceTok);
594  void HandlePragmaMark();
595  void HandlePragmaPoison(Token &PoisonTok);
596  void HandlePragmaSystemHeader(Token &SysHeaderTok);
597  void HandlePragmaDependency(Token &DependencyTok);
598};
599
600/// PreprocessorFactory - A generic factory interface for lazily creating
601///  Preprocessor objects on-demand when they are needed.
602class PreprocessorFactory {
603public:
604  virtual ~PreprocessorFactory();
605  virtual Preprocessor* CreatePreprocessor() = 0;
606};
607
608}  // end namespace clang
609
610#endif
611