Lexer.h revision 1fa495304c81e03f07f278a47b5efe9317104aab
1ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang//===--- Lexer.h - C Language Family Lexer ----------------------*- C++ -*-===// 2ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang// 3ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang// The LLVM Compiler Infrastructure 4ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang// 5ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang// This file is distributed under the University of Illinois Open Source 6ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang// License. See LICENSE.TXT for details. 7ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang// 8ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang//===----------------------------------------------------------------------===// 9ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang// 10ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang// This file defines the Lexer interface. 11ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang// 12ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang//===----------------------------------------------------------------------===// 13ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 14ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang#ifndef LLVM_CLANG_LEXER_H 15ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang#define LLVM_CLANG_LEXER_H 16ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 17ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang#include "clang/Lex/PreprocessorLexer.h" 18ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang#include "clang/Basic/LangOptions.h" 19b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian#include "llvm/ADT/SmallVector.h" 20b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian#include <string> 21b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian#include <vector> 22b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian#include <cassert> 23ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 24ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuangnamespace clang { 25ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuangclass Diagnostic; 26ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuangclass SourceManager; 27ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuangclass Preprocessor; 28ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuangclass DiagnosticBuilder; 29ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 30ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang/// Lexer - This provides a simple interface that turns a text buffer into a 31ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang/// stream of tokens. This provides no support for file reading or buffering, 325ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang/// or buffering/seeking of tokens, only forward lexing is supported. It relies 339b35249446b07f40ac5fcc3205f2c048616efacchkuang/// on the specified Preprocessor object to handle preprocessor directives, etc. 349b35249446b07f40ac5fcc3205f2c048616efacchkuangclass Lexer : public PreprocessorLexer { 359b35249446b07f40ac5fcc3205f2c048616efacchkuang //===--------------------------------------------------------------------===// 369b35249446b07f40ac5fcc3205f2c048616efacchkuang // Constant configuration values for this lexer. 379b35249446b07f40ac5fcc3205f2c048616efacchkuang const char *BufferStart; // Start of the buffer. 389b35249446b07f40ac5fcc3205f2c048616efacchkuang const char *BufferEnd; // End of the buffer. 399b35249446b07f40ac5fcc3205f2c048616efacchkuang SourceLocation FileLoc; // Location for start of file. 409b35249446b07f40ac5fcc3205f2c048616efacchkuang LangOptions Features; // Features enabled by this language (cache). 419b35249446b07f40ac5fcc3205f2c048616efacchkuang bool Is_PragmaLexer; // True if lexer for _Pragma handling. 429b35249446b07f40ac5fcc3205f2c048616efacchkuang 439b35249446b07f40ac5fcc3205f2c048616efacchkuang //===--------------------------------------------------------------------===// 449b35249446b07f40ac5fcc3205f2c048616efacchkuang // Context-specific lexing flags set by the preprocessor. 455ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang // 46ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 47b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian /// ExtendedTokenMode - The lexer can optionally keep comments and whitespace 48ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// and return them as tokens. This is used for -C and -CC modes, and 49ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// whitespace preservation can be useful for some clients that want to lex 50ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// the file in raw mode and get every character from the file. 51ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// 52ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// When this is set to 2 it returns comments and whitespace. When set to 1 53ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// it returns comments, when it is set to 0 it returns normal tokens only. 54ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang unsigned char ExtendedTokenMode; 55ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 56ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang //===--------------------------------------------------------------------===// 57f3bed9137f66ef693bd406e43b17e9a1114f1e14hkuang // Context that changes as the file is lexed. 585ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang // NOTE: any state that mutates when in raw mode must have save/restore code 591184aebb761cbeac9124c37189a80a1a58f04b6bhkuang // in Lexer::isNextPPTokenLParen. 601184aebb761cbeac9124c37189a80a1a58f04b6bhkuang 611184aebb761cbeac9124c37189a80a1a58f04b6bhkuang // BufferPtr - Current pointer into the buffer. This is the next character 621184aebb761cbeac9124c37189a80a1a58f04b6bhkuang // to be lexed. 63b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian const char *BufferPtr; 64ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 65ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang // IsAtStartOfLine - True if the next lexed token should get the "start of 66ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang // line" flag set on it. 675ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang bool IsAtStartOfLine; 68b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian 699b35249446b07f40ac5fcc3205f2c048616efacchkuang Lexer(const Lexer&); // DO NOT IMPLEMENT 70b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian void operator=(const Lexer&); // DO NOT IMPLEMENT 71ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang friend class Preprocessor; 72ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 73ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang void InitLexer(const char *BufStart, const char *BufPtr, const char *BufEnd); 74ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuangpublic: 7591037db265ecdd914a26e056cf69207b4f50924ehkuang 76ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// Lexer constructor - Create a new lexer object for the specified buffer 77ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// with the specified preprocessor managing the lexing process. This lexer 78ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// assumes that the associated file buffer and Preprocessor objects will 79ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// outlive it, so it doesn't take ownership of either of them. 80ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang Lexer(FileID FID, Preprocessor &PP); 81ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 82ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// Lexer constructor - Create a new raw lexer object. This object is only 83f3bed9137f66ef693bd406e43b17e9a1114f1e14hkuang /// suitable for calls to 'LexRawToken'. This lexer assumes that the text 84f3bed9137f66ef693bd406e43b17e9a1114f1e14hkuang /// range will outlive it, so it doesn't take ownership of it. 85b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian Lexer(SourceLocation FileLoc, const LangOptions &Features, 86b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian const char *BufStart, const char *BufPtr, const char *BufEnd); 87f3bed9137f66ef693bd406e43b17e9a1114f1e14hkuang 88ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// Lexer constructor - Create a new raw lexer object. This object is only 89ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// suitable for calls to 'LexRawToken'. This lexer assumes that the text 90ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// range will outlive it, so it doesn't take ownership of it. 91ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang Lexer(FileID FID, const SourceManager &SM, const LangOptions &Features); 92ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 93ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// Create_PragmaLexer: Lexer constructor - Create a new lexer object for 949b35249446b07f40ac5fcc3205f2c048616efacchkuang /// _Pragma expansion. This has a variety of magic semantics that this method 959b35249446b07f40ac5fcc3205f2c048616efacchkuang /// sets up. It returns a new'd Lexer that must be delete'd when done. 969b35249446b07f40ac5fcc3205f2c048616efacchkuang static Lexer *Create_PragmaLexer(SourceLocation SpellingLoc, 979b35249446b07f40ac5fcc3205f2c048616efacchkuang SourceLocation InstantiationLocStart, 98ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang SourceLocation InstantiationLocEnd, 99ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang unsigned TokLen, Preprocessor &PP); 100ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 101ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 102ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// getFeatures - Return the language features currently enabled. NOTE: this 103ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// lexer modifies features as a file is parsed! 104ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang const LangOptions &getFeatures() const { return Features; } 105ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 106ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// getFileLoc - Return the File Location for the file we are lexing out of. 107ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// The physical location encodes the location where the characters come from, 1085ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// the virtual location encodes where we should *claim* the characters came 109ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// from. Currently this is only used by _Pragma handling. 110ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang SourceLocation getFileLoc() const { return FileLoc; } 111ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 112ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// Lex - Return the next token in the file. If this is the end of file, it 113ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// return the tok::eof token. Return true if an error occurred and 11491037db265ecdd914a26e056cf69207b4f50924ehkuang /// compilation should terminate, false if normal. This implicitly involves 115f3bed9137f66ef693bd406e43b17e9a1114f1e14hkuang /// the preprocessor. 116b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian void Lex(Token &Result) { 117b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian // Start a new token. 118b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian Result.startToken(); 119ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 120ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang // NOTE, any changes here should also change code after calls to 121ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang // Preprocessor::HandleDirective 122ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang if (IsAtStartOfLine) { 123ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang Result.setFlag(Token::StartOfLine); 124ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang IsAtStartOfLine = false; 125ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang } 126ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 127ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang // Get a token. Note that this may delete the current lexer if the end of 128ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang // file is reached. 129ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang LexTokenInternal(Result); 130ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang } 131ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 132ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// isPragmaLexer - Returns true if this Lexer is being used to lex a pragma. 133ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang bool isPragmaLexer() const { return Is_PragmaLexer; } 134b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian 1355ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// IndirectLex - An indirect call to 'Lex' that can be invoked via 136b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian /// the PreprocessorLexer interface. 1371184aebb761cbeac9124c37189a80a1a58f04b6bhkuang void IndirectLex(Token &Result) { Lex(Result); } 1385ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang 139ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// LexFromRawLexer - Lex a token from a designated raw lexer (one with no 1409b35249446b07f40ac5fcc3205f2c048616efacchkuang /// associated preprocessor object. Return true if the 'next character to 1419b35249446b07f40ac5fcc3205f2c048616efacchkuang /// read' pointer points at the end of the lexer buffer, false otherwise. 1429b35249446b07f40ac5fcc3205f2c048616efacchkuang bool LexFromRawLexer(Token &Result) { 1439b35249446b07f40ac5fcc3205f2c048616efacchkuang assert(LexingRawMode && "Not already in raw mode!"); 1449b35249446b07f40ac5fcc3205f2c048616efacchkuang Lex(Result); 145ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang // Note that lexing to the end of the buffer doesn't implicitly delete the 146ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang // lexer when in raw mode. 147ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang return BufferPtr == BufferEnd; 148ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang } 149ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 150ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// isKeepWhitespaceMode - Return true if the lexer should return tokens for 151ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// every character in the file, including whitespace and comments. This 1525ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// should only be used in raw mode, as the preprocessor is not prepared to 153ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// deal with the excess tokens. 154ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang bool isKeepWhitespaceMode() const { 155ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang return ExtendedTokenMode > 1; 156ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang } 157ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 158ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// SetKeepWhitespaceMode - This method lets clients enable or disable 159ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// whitespace retention mode. 160f3bed9137f66ef693bd406e43b17e9a1114f1e14hkuang void SetKeepWhitespaceMode(bool Val) { 1615ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang assert((!Val || LexingRawMode) && 162ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang "Can only enable whitespace retention in raw mode"); 163ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang ExtendedTokenMode = Val ? 2 : 0; 164ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang } 165ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 1661184aebb761cbeac9124c37189a80a1a58f04b6bhkuang /// inKeepCommentMode - Return true if the lexer should return comments as 16791037db265ecdd914a26e056cf69207b4f50924ehkuang /// tokens. 16891037db265ecdd914a26e056cf69207b4f50924ehkuang bool inKeepCommentMode() const { 16991037db265ecdd914a26e056cf69207b4f50924ehkuang return ExtendedTokenMode > 0; 1701184aebb761cbeac9124c37189a80a1a58f04b6bhkuang } 171ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 172ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// SetCommentRetentionMode - Change the comment retention mode of the lexer 173ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// to the specified mode. This is really only useful when lexing in raw 174ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// mode, because otherwise the lexer needs to manage this. 175ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang void SetCommentRetentionState(bool Mode) { 176ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang assert(!isKeepWhitespaceMode() && 177ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang "Can't play with comment retention state when retaining whitespace"); 178ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang ExtendedTokenMode = Mode ? 1 : 0; 179ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang } 180ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 181ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang const char *getBufferStart() const { return BufferStart; } 182ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 183ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// ReadToEndOfLine - Read the rest of the current preprocessor line as an 184ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// uninterpreted string. This switches the lexer out of directive mode. 185ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang std::string ReadToEndOfLine(); 186ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 187ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 1885ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// Diag - Forwarding function for diagnostics. This translate a source 189ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// position in the current buffer into a SourceLocation object for rendering. 1901184aebb761cbeac9124c37189a80a1a58f04b6bhkuang DiagnosticBuilder Diag(const char *Loc, unsigned DiagID) const; 1911184aebb761cbeac9124c37189a80a1a58f04b6bhkuang 1921184aebb761cbeac9124c37189a80a1a58f04b6bhkuang /// getSourceLocation - Return a source location identifier for the specified 1931184aebb761cbeac9124c37189a80a1a58f04b6bhkuang /// offset in the current file. 194ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang SourceLocation getSourceLocation(const char *Loc, unsigned TokLen = 1) const; 1955ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang 1965ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// getSourceLocation - Return a source location for the next character in 1975ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// the current file. 1985ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang SourceLocation getSourceLocation() { return getSourceLocation(BufferPtr); } 1995ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang 2005ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// Stringify - Convert the specified string into a C string by escaping '\' 2015ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// and " characters. This does not add surrounding ""'s to the string. 2025ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// If Charify is true, this escapes the ' character instead of ". 2035ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang static std::string Stringify(const std::string &Str, bool Charify = false); 2045ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang 2055ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// Stringify - Convert the specified string into a C string by escaping '\' 2069b35249446b07f40ac5fcc3205f2c048616efacchkuang /// and " characters. This does not add surrounding ""'s to the string. 2075ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang static void Stringify(llvm::SmallVectorImpl<char> &Str); 2089b35249446b07f40ac5fcc3205f2c048616efacchkuang 2095ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// MeasureTokenLength - Relex the token at the specified location and return 2109b35249446b07f40ac5fcc3205f2c048616efacchkuang /// its length in bytes in the input file. If the token needs cleaning (e.g. 2115ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// includes a trigraph or an escaped newline) then this count includes bytes 2129b35249446b07f40ac5fcc3205f2c048616efacchkuang /// that are part of that. 2135ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang static unsigned MeasureTokenLength(SourceLocation Loc, 2149b35249446b07f40ac5fcc3205f2c048616efacchkuang const SourceManager &SM); 2155ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang 2169b35249446b07f40ac5fcc3205f2c048616efacchkuang //===--------------------------------------------------------------------===// 2175ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang // Internal implementation interfaces. 2189b35249446b07f40ac5fcc3205f2c048616efacchkuangprivate: 2195ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang 2209b35249446b07f40ac5fcc3205f2c048616efacchkuang /// LexTokenInternal - Internal interface to lex a preprocessing token. Called 2215ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// by Lex. 2229b35249446b07f40ac5fcc3205f2c048616efacchkuang /// 2235ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang void LexTokenInternal(Token &Result); 2249b35249446b07f40ac5fcc3205f2c048616efacchkuang 2255ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// FormTokenWithChars - When we lex a token, we have identified a span 2269b35249446b07f40ac5fcc3205f2c048616efacchkuang /// starting at BufferPtr, going to TokEnd that forms the token. This method 2275ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// takes that range and assigns it to the token as its location and size. In 2289b35249446b07f40ac5fcc3205f2c048616efacchkuang /// addition, since tokens cannot overlap, this also updates BufferPtr to be 2295ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang /// TokEnd. 2305ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang void FormTokenWithChars(Token &Result, const char *TokEnd, 2315ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang tok::TokenKind Kind) { 2325ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang unsigned TokLen = TokEnd-BufferPtr; 2335ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang Result.setLength(TokLen); 2345ae7ac49f08a179e4f054d99fcfc9dce78d26e58hkuang Result.setLocation(getSourceLocation(BufferPtr, TokLen)); 235b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian Result.setKind(Kind); 236b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian BufferPtr = TokEnd; 237b08e2e23eec181e9951df33cd704ac294c5407b6Vignesh Venkatasubramanian } 238ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang 239ba164dffc5a6795bce97fae02b51ccf3330e15e4hkuang /// isNextPPTokenLParen - Return 1 if the next unexpanded token will return a 240 /// tok::l_paren token, 0 if it is something else and 2 if there are no more 241 /// tokens in the buffer controlled by this lexer. 242 unsigned isNextPPTokenLParen(); 243 244 //===--------------------------------------------------------------------===// 245 // Lexer character reading interfaces. 246public: 247 248 // This lexer is built on two interfaces for reading characters, both of which 249 // automatically provide phase 1/2 translation. getAndAdvanceChar is used 250 // when we know that we will be reading a character from the input buffer and 251 // that this character will be part of the result token. This occurs in (f.e.) 252 // string processing, because we know we need to read until we find the 253 // closing '"' character. 254 // 255 // The second interface is the combination of PeekCharAndSize with 256 // ConsumeChar. PeekCharAndSize reads a phase 1/2 translated character, 257 // returning it and its size. If the lexer decides that this character is 258 // part of the current token, it calls ConsumeChar on it. This two stage 259 // approach allows us to emit diagnostics for characters (e.g. warnings about 260 // trigraphs), knowing that they only are emitted if the character is 261 // consumed. 262 263 /// isObviouslySimpleCharacter - Return true if the specified character is 264 /// obviously the same in translation phase 1 and translation phase 3. This 265 /// can return false for characters that end up being the same, but it will 266 /// never return true for something that needs to be mapped. 267 static bool isObviouslySimpleCharacter(char C) { 268 return C != '?' && C != '\\'; 269 } 270 271 /// getAndAdvanceChar - Read a single 'character' from the specified buffer, 272 /// advance over it, and return it. This is tricky in several cases. Here we 273 /// just handle the trivial case and fall-back to the non-inlined 274 /// getCharAndSizeSlow method to handle the hard case. 275 inline char getAndAdvanceChar(const char *&Ptr, Token &Tok) { 276 // If this is not a trigraph and not a UCN or escaped newline, return 277 // quickly. 278 if (isObviouslySimpleCharacter(Ptr[0])) return *Ptr++; 279 280 unsigned Size = 0; 281 char C = getCharAndSizeSlow(Ptr, Size, &Tok); 282 Ptr += Size; 283 return C; 284 } 285 286private: 287 /// ConsumeChar - When a character (identified by PeekCharAndSize) is consumed 288 /// and added to a given token, check to see if there are diagnostics that 289 /// need to be emitted or flags that need to be set on the token. If so, do 290 /// it. 291 const char *ConsumeChar(const char *Ptr, unsigned Size, Token &Tok) { 292 // Normal case, we consumed exactly one token. Just return it. 293 if (Size == 1) 294 return Ptr+Size; 295 296 // Otherwise, re-lex the character with a current token, allowing 297 // diagnostics to be emitted and flags to be set. 298 Size = 0; 299 getCharAndSizeSlow(Ptr, Size, &Tok); 300 return Ptr+Size; 301 } 302 303 /// getCharAndSize - Peek a single 'character' from the specified buffer, 304 /// get its size, and return it. This is tricky in several cases. Here we 305 /// just handle the trivial case and fall-back to the non-inlined 306 /// getCharAndSizeSlow method to handle the hard case. 307 inline char getCharAndSize(const char *Ptr, unsigned &Size) { 308 // If this is not a trigraph and not a UCN or escaped newline, return 309 // quickly. 310 if (isObviouslySimpleCharacter(Ptr[0])) { 311 Size = 1; 312 return *Ptr; 313 } 314 315 Size = 0; 316 return getCharAndSizeSlow(Ptr, Size); 317 } 318 319 /// getCharAndSizeSlow - Handle the slow/uncommon case of the getCharAndSize 320 /// method. 321 char getCharAndSizeSlow(const char *Ptr, unsigned &Size, Token *Tok = 0); 322public: 323 324 /// getCharAndSizeNoWarn - Like the getCharAndSize method, but does not ever 325 /// emit a warning. 326 static inline char getCharAndSizeNoWarn(const char *Ptr, unsigned &Size, 327 const LangOptions &Features) { 328 // If this is not a trigraph and not a UCN or escaped newline, return 329 // quickly. 330 if (isObviouslySimpleCharacter(Ptr[0])) { 331 Size = 1; 332 return *Ptr; 333 } 334 335 Size = 0; 336 return getCharAndSizeSlowNoWarn(Ptr, Size, Features); 337 } 338private: 339 340 /// getCharAndSizeSlowNoWarn - Same as getCharAndSizeSlow, but never emits a 341 /// diagnostic. 342 static char getCharAndSizeSlowNoWarn(const char *Ptr, unsigned &Size, 343 const LangOptions &Features); 344 345 //===--------------------------------------------------------------------===// 346 // Other lexer functions. 347 348 // Helper functions to lex the remainder of a token of the specific type. 349 void LexIdentifier (Token &Result, const char *CurPtr); 350 void LexNumericConstant (Token &Result, const char *CurPtr); 351 void LexStringLiteral (Token &Result, const char *CurPtr,bool Wide); 352 void LexAngledStringLiteral(Token &Result, const char *CurPtr); 353 void LexCharConstant (Token &Result, const char *CurPtr); 354 bool LexEndOfFile (Token &Result, const char *CurPtr); 355 356 bool SkipWhitespace (Token &Result, const char *CurPtr); 357 bool SkipBCPLComment (Token &Result, const char *CurPtr); 358 bool SkipBlockComment (Token &Result, const char *CurPtr); 359 bool SaveBCPLComment (Token &Result, const char *CurPtr); 360}; 361 362 363} // end namespace clang 364 365#endif 366