TokenLexer.h revision d60a34a4e514ec0dfddd05ef2744be104e111f45
11543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner//===--- TokenLexer.h - Lex from a token buffer -----------------*- C++ -*-===//
25f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer//
35f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer//                     The LLVM Compiler Infrastructure
45f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer//
50bc735ffcfb223c0186419547abaa5c84482663eChris Lattner// This file is distributed under the University of Illinois Open Source
60bc735ffcfb223c0186419547abaa5c84482663eChris Lattner// License. See LICENSE.TXT for details.
75f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer//
85f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer//===----------------------------------------------------------------------===//
95f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer//
10e5c8ffe09f8fec836b90f23c2a3b996bfc5b09c4Chris Lattner// This file defines the TokenLexer interface.
115f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer//
125f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer//===----------------------------------------------------------------------===//
135f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer
141543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner#ifndef LLVM_CLANG_TOKENLEXER_H
151543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner#define LLVM_CLANG_TOKENLEXER_H
165f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer
175f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer#include "clang/Basic/SourceLocation.h"
185f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer
195f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencernamespace clang {
205f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  class MacroInfo;
215f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  class Preprocessor;
22d217773f106856a11879ec79dc468efefaf2ee75Chris Lattner  class Token;
23e5c8ffe09f8fec836b90f23c2a3b996bfc5b09c4Chris Lattner  class MacroArgs;
241eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
251543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner/// TokenLexer - This implements a lexer that returns token from a macro body
265d75de0f821023f4ed4815825bf3aea8a0b5e40dChris Lattner/// or token stream instead of lexing from a character buffer.  This is used for
275d75de0f821023f4ed4815825bf3aea8a0b5e40dChris Lattner/// macro expansion and _Pragma handling, for example.
285f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer///
291543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattnerclass TokenLexer {
305f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// Macro - The macro we are expanding from.  This is null if expanding a
315f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// token stream.
325f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  ///
335f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  MacroInfo *Macro;
345f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer
355f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// ActualArgs - The actual arguments specified for a function-like macro, or
361543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner  /// null.  The TokenLexer owns the pointed-to object.
375f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  MacroArgs *ActualArgs;
385f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer
395f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// PP - The current preprocessor object we are expanding for.
405f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  ///
415f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  Preprocessor &PP;
425f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer
438d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner  /// Tokens - This is the pointer to an array of tokens that the macro is
445f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// defined to, with arguments expanded for function-like macros.  If this is
4532fca722dd974b8202d0fb9c71b6c185c0767da6Chris Lattner  /// a token stream, these are the tokens we are returning.  This points into
465b3284a9275a27f5c4410e25eb8933be540601d6Argyrios Kyrtzidis  /// the macro definition we are lexing from, a cache buffer that is owned by
475b3284a9275a27f5c4410e25eb8933be540601d6Argyrios Kyrtzidis  /// the preprocessor, or some other buffer that we may or may not own
485b3284a9275a27f5c4410e25eb8933be540601d6Argyrios Kyrtzidis  /// (depending on OwnsTokens).
495b3284a9275a27f5c4410e25eb8933be540601d6Argyrios Kyrtzidis  /// Note that if it points into Preprocessor's cache buffer, the Preprocessor
505b3284a9275a27f5c4410e25eb8933be540601d6Argyrios Kyrtzidis  /// may update the pointer as needed.
518d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner  const Token *Tokens;
525b3284a9275a27f5c4410e25eb8933be540601d6Argyrios Kyrtzidis  friend class Preprocessor;
531eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
548d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner  /// NumTokens - This is the length of the Tokens array.
555f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  ///
568d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner  unsigned NumTokens;
571eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
585f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// CurToken - This is the next token that Lex will return.
595f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  ///
605f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  unsigned CurToken;
611eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
62d0f5e443ceed37e1a72015bd109db8e5fc6206cfChandler Carruth  /// ExpandLocStart/End - The source location range where this macro was
63d0f5e443ceed37e1a72015bd109db8e5fc6206cfChandler Carruth  /// expanded.
64d0f5e443ceed37e1a72015bd109db8e5fc6206cfChandler Carruth  SourceLocation ExpandLocStart, ExpandLocEnd;
651eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
66b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis  /// \brief Source location pointing at the source location entry chunk that
67d0f5e443ceed37e1a72015bd109db8e5fc6206cfChandler Carruth  /// was reserved for the current macro expansion.
68b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis  SourceLocation MacroExpansionStart;
69b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis
70d0f5e443ceed37e1a72015bd109db8e5fc6206cfChandler Carruth  /// \brief The offset of the macro expansion in the
71b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis  /// "source location address space".
72b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis  unsigned MacroStartSLocOffset;
73b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis
74b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis  /// \brief FileID/offset of the start of the macro definition.
75b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis  std::pair<FileID, unsigned> MacroDefStartInfo;
76b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis
775f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// Lexical information about the expansion point of the macro: the identifier
785f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// that the macro expanded from had these properties.
799c683062752a26014197df1c8792a19efa9e93cfChris Lattner  bool AtStartOfLine : 1;
809c683062752a26014197df1c8792a19efa9e93cfChris Lattner  bool HasLeadingSpace : 1;
811eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
821543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner  /// OwnsTokens - This is true if this TokenLexer allocated the Tokens
838d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner  /// array, and thus needs to free it when destroyed.  For simple object-like
848d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner  /// macros (for example) we just point into the token buffer of the macro
858d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner  /// definition, we don't make a copy of it.
868d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner  bool OwnsTokens : 1;
871eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
886b884508c3bc97cc9df9516adb92fbf88dd0a2e4Chris Lattner  /// DisableMacroExpansion - This is true when tokens lexed from the TokenLexer
896b884508c3bc97cc9df9516adb92fbf88dd0a2e4Chris Lattner  /// should not be subject to further macro expansion.
906b884508c3bc97cc9df9516adb92fbf88dd0a2e4Chris Lattner  bool DisableMacroExpansion : 1;
911eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
921543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner  TokenLexer(const TokenLexer&);  // DO NOT IMPLEMENT
931543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner  void operator=(const TokenLexer&); // DO NOT IMPLEMENT
945f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencerpublic:
951543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner  /// Create a TokenLexer for the specified macro with the specified actual
965f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// arguments.  Note that this ctor takes ownership of the ActualArgs pointer.
97e7fb48466afcbf2c4ccdfa658824282fdc3c512cChris Lattner  /// ILEnd specifies the location of the ')' for a function-like macro or the
98e7fb48466afcbf2c4ccdfa658824282fdc3c512cChris Lattner  /// identifier for an object-like macro.
99e7fb48466afcbf2c4ccdfa658824282fdc3c512cChris Lattner  TokenLexer(Token &Tok, SourceLocation ILEnd, MacroArgs *ActualArgs,
100e7fb48466afcbf2c4ccdfa658824282fdc3c512cChris Lattner             Preprocessor &pp)
1018d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner    : Macro(0), ActualArgs(0), PP(pp), OwnsTokens(false) {
102e7fb48466afcbf2c4ccdfa658824282fdc3c512cChris Lattner    Init(Tok, ILEnd, ActualArgs);
1039594acf32de2939b15eafa8fe818607bfc56bf66Chris Lattner  }
1041eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
1051543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner  /// Init - Initialize this TokenLexer to expand from the specified macro
1069594acf32de2939b15eafa8fe818607bfc56bf66Chris Lattner  /// with the specified argument information.  Note that this ctor takes
107e7fb48466afcbf2c4ccdfa658824282fdc3c512cChris Lattner  /// ownership of the ActualArgs pointer.  ILEnd specifies the location of the
108e7fb48466afcbf2c4ccdfa658824282fdc3c512cChris Lattner  /// ')' for a function-like macro or the identifier for an object-like macro.
109e7fb48466afcbf2c4ccdfa658824282fdc3c512cChris Lattner  void Init(Token &Tok, SourceLocation ILEnd, MacroArgs *ActualArgs);
1101eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
11121ec0e4c5db85e45b59cafabca66d640e0ee2dccChris Lattner  /// Create a TokenLexer for the specified token stream.  If 'OwnsTokens' is
11221ec0e4c5db85e45b59cafabca66d640e0ee2dccChris Lattner  /// specified, this takes ownership of the tokens and delete[]'s them when
11321ec0e4c5db85e45b59cafabca66d640e0ee2dccChris Lattner  /// the token lexer is empty.
1146b884508c3bc97cc9df9516adb92fbf88dd0a2e4Chris Lattner  TokenLexer(const Token *TokArray, unsigned NumToks, bool DisableExpansion,
11521ec0e4c5db85e45b59cafabca66d640e0ee2dccChris Lattner             bool ownsTokens, Preprocessor &pp)
1168d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner    : Macro(0), ActualArgs(0), PP(pp), OwnsTokens(false) {
11721ec0e4c5db85e45b59cafabca66d640e0ee2dccChris Lattner    Init(TokArray, NumToks, DisableExpansion, ownsTokens);
1189594acf32de2939b15eafa8fe818607bfc56bf66Chris Lattner  }
1191eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
1201543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner  /// Init - Initialize this TokenLexer with the specified token stream.
1219594acf32de2939b15eafa8fe818607bfc56bf66Chris Lattner  /// This does not take ownership of the specified token vector.
1226b884508c3bc97cc9df9516adb92fbf88dd0a2e4Chris Lattner  ///
1231eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump  /// DisableExpansion is true when macro expansion of tokens lexed from this
1246b884508c3bc97cc9df9516adb92fbf88dd0a2e4Chris Lattner  /// stream should be disabled.
1256b884508c3bc97cc9df9516adb92fbf88dd0a2e4Chris Lattner  void Init(const Token *TokArray, unsigned NumToks,
1266b884508c3bc97cc9df9516adb92fbf88dd0a2e4Chris Lattner            bool DisableMacroExpansion, bool OwnsTokens);
1271eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
1281543e9c69202b8e128c7fe34784ae7aa90964889Chris Lattner  ~TokenLexer() { destroy(); }
1291eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
1305f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// isNextTokenLParen - If the next token lexed will pop this macro off the
1315f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// expansion stack, return 2.  If the next unexpanded token is a '(', return
1325f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// 1, otherwise return 0.
1335f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  unsigned isNextTokenLParen() const;
1341eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
1355f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// Lex - Lex and return a token from this macro stream.
136d217773f106856a11879ec79dc468efefaf2ee75Chris Lattner  void Lex(Token &Tok);
1371eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
138b2eb53d9fd973a1a02e05e67a3307b3efd12eff2Peter Collingbourne  /// isParsingPreprocessorDirective - Return true if we are in the middle of a
139b2eb53d9fd973a1a02e05e67a3307b3efd12eff2Peter Collingbourne  /// preprocessor directive.
140b2eb53d9fd973a1a02e05e67a3307b3efd12eff2Peter Collingbourne  bool isParsingPreprocessorDirective() const;
141b2eb53d9fd973a1a02e05e67a3307b3efd12eff2Peter Collingbourne
1425f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencerprivate:
1439594acf32de2939b15eafa8fe818607bfc56bf66Chris Lattner  void destroy();
1441eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
1455f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// isAtEnd - Return true if the next lex call will pop this macro off the
1465f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// include stack.
1475f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  bool isAtEnd() const {
1488d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner    return CurToken == NumTokens;
1495f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  }
1501eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
1515f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// PasteTokens - Tok is the LHS of a ## operator, and CurToken is the ##
1525f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// operator.  Read the ## and RHS, and paste the LHS/RHS together.  If there
1535f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// are is another ## after it, chomp it iteratively.  Return the result as
1543f1cc838f9caf469990f23fccd0940263c0c61ffChris Lattner  /// Tok.  If this returns true, the caller should immediately return the
1553f1cc838f9caf469990f23fccd0940263c0c61ffChris Lattner  /// token.
1563f1cc838f9caf469990f23fccd0940263c0c61ffChris Lattner  bool PasteTokens(Token &Tok);
1571eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
1585f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  /// Expand the arguments of a function-like macro so that we can quickly
1598d896438031f5990cb42eb95a7bae233db2eb333Chris Lattner  /// return preexpanded tokens from Tokens.
1605f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer  void ExpandFunctionArguments();
1611eb4433ac451dc16f4133a88af2d002ac26c58efMike Stump
1623f1cc838f9caf469990f23fccd0940263c0c61ffChris Lattner  /// HandleMicrosoftCommentPaste - In microsoft compatibility mode, /##/ pastes
1633f1cc838f9caf469990f23fccd0940263c0c61ffChris Lattner  /// together to form a comment that comments out everything in the current
1643f1cc838f9caf469990f23fccd0940263c0c61ffChris Lattner  /// macro, other active macros, and anything left on the current physical
165d0f5e443ceed37e1a72015bd109db8e5fc6206cfChandler Carruth  /// source line of the expanded buffer.  Handle this by returning the
1663f1cc838f9caf469990f23fccd0940263c0c61ffChris Lattner  /// first token on the next line.
1673f1cc838f9caf469990f23fccd0940263c0c61ffChris Lattner  void HandleMicrosoftCommentPaste(Token &Tok);
168b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis
169b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis  /// \brief If \arg loc is a FileID and points inside the current macro
170b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis  /// definition, returns the appropriate source location pointing at the
171b73377eeb3eff76be134203aebb6068244b177f3Argyrios Kyrtzidis  /// macro expansion source location entry.
172ec3b5389a5032433116ec2f2ee266b9666ad2eb4Argyrios Kyrtzidis  SourceLocation getExpansionLocForMacroDefLoc(SourceLocation loc) const;
173d60a34a4e514ec0dfddd05ef2744be104e111f45Argyrios Kyrtzidis
174d60a34a4e514ec0dfddd05ef2744be104e111f45Argyrios Kyrtzidis  /// \brief Creates SLocEntries and updates the locations of macro argument
175d60a34a4e514ec0dfddd05ef2744be104e111f45Argyrios Kyrtzidis  /// tokens to their new expanded locations.
176d60a34a4e514ec0dfddd05ef2744be104e111f45Argyrios Kyrtzidis  ///
177d60a34a4e514ec0dfddd05ef2744be104e111f45Argyrios Kyrtzidis  /// \param ArgIdSpellLoc the location of the macro argument id inside the
178d60a34a4e514ec0dfddd05ef2744be104e111f45Argyrios Kyrtzidis  /// macro definition.
179d60a34a4e514ec0dfddd05ef2744be104e111f45Argyrios Kyrtzidis  void updateLocForMacroArgTokens(SourceLocation ArgIdSpellLoc,
180d60a34a4e514ec0dfddd05ef2744be104e111f45Argyrios Kyrtzidis                                  Token *begin_tokens, Token *end_tokens);
1815f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer};
1825f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer
1835f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer}  // end namespace clang
1845f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer
1855f016e2cb5d11daeb237544de1c5d59f20fe1a6eReid Spencer#endif
186