#include "clang/Lex/Lexer.h"
#include "UnicodeCharSets.h"
#include "clang/Basic/CharInfo.h"
#include "clang/Basic/IdentifierTable.h"
#include "clang/Basic/LangOptions.h"
#include "clang/Basic/SourceLocation.h"
#include "clang/Basic/SourceManager.h"
#include "clang/Basic/TokenKinds.h"
#include "clang/Lex/LexDiagnostic.h"
#include "clang/Lex/LiteralSupport.h"
#include "clang/Lex/MultipleIncludeOpt.h"
#include "clang/Lex/Preprocessor.h"
#include "clang/Lex/PreprocessorOptions.h"
#include "clang/Lex/Token.h"
#include "clang/Basic/Diagnostic.h"
#include "clang/Basic/LLVM.h"
#include "llvm/ADT/None.h"
#include "llvm/ADT/Optional.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/ADT/StringSwitch.h"
#include "llvm/ADT/StringRef.h"
#include "llvm/Support/Compiler.h"
#include "llvm/Support/ConvertUTF.h"
#include "llvm/Support/MathExtras.h"
#include "llvm/Support/MemoryBuffer.h"
#include "llvm/Support/NativeFormatting.h"
#include "llvm/Support/UnicodeCharRanges.h"
#include <algorithm>
#include <cassert>
#include <cstddef>
#include <cstdint>
#include <cstring>
#include <string>
#include <tuple>
#include <utility>

Include dependency graph for Lexer.cpp:

Enumerations
enum	PreambleDirectiveKind

Functions
template<typename T >
static void	StringifyImpl (T &Str, char Quote)

static size_t	getSpellingSlow (const Token &Tok, const char BufPtr, const LangOptions &LangOpts, char Spelling)
	Slow case of getSpelling. More...

static const char *	findBeginningOfLine (StringRef Buffer, unsigned Offset)
	Returns the pointer that points to the beginning of line that contains the given offset, or null if the offset if invalid. More...

static SourceLocation	getBeginningOfFileToken (SourceLocation Loc, const SourceManager &SM, const LangOptions &LangOpts)

static CharSourceRange	makeRangeFromFileLocs (CharSourceRange Range, const SourceManager &SM, const LangOptions &LangOpts)

static LLVM_ATTRIBUTE_NOINLINE SourceLocation	GetMappedTokenLoc (Preprocessor &PP, SourceLocation FileLoc, unsigned CharNo, unsigned TokLen)
	GetMappedTokenLoc - If lexing out of a 'mapped buffer', where we pretend the lexer buffer was all expanded at a single point, perform the mapping. More...

static char	GetTrigraphCharForLetter (char Letter)
	GetTrigraphCharForLetter - Given a character that occurs after a ?? pair, return the decoded trigraph letter it corresponds to, or '\0' if nothing. More...

static char	DecodeTrigraphChar (const char CP, Lexer L)
	DecodeTrigraphChar - If the specified character is a legal trigraph when prefixed with ??, emit a trigraph warning. More...

static bool	isAllowedIDChar (uint32_t C, const LangOptions &LangOpts)

static bool	isAllowedInitiallyIDChar (uint32_t C, const LangOptions &LangOpts)

static CharSourceRange	makeCharRange (Lexer &L, const char Begin, const char End)

static void	maybeDiagnoseIDCharCompat (DiagnosticsEngine &Diags, uint32_t C, CharSourceRange Range, bool IsFirst)

static void	maybeDiagnoseUTF8Homoglyph (DiagnosticsEngine &Diags, uint32_t C, CharSourceRange Range)
	After encountering UTF-8 character C and interpreting it as an identifier character, check whether it's a homoglyph for a common non-identifier source character that is unlikely to be an intentional identifier character and warn if so. More...

static bool	isEndOfBlockCommentWithEscapedNewLine (const char CurPtr, Lexer L)
	isBlockCommentEndOfEscapedNewLine - Return true if the specified newline character (either \n or \r) is part of an escaped newline sequence. More...

static const char *	FindConflictEnd (const char CurPtr, const char BufferEnd, ConflictMarkerKind CMK)
	Find the end of a version control conflict marker. More...

static const char *	findPlaceholderEnd (const char CurPtr, const char BufferEnd)

Enumeration Type Documentation

◆ PreambleDirectiveKind

enum PreambleDirectiveKind

Definition at line 566 of file Lexer.cpp.

Function Documentation

◆ DecodeTrigraphChar()

static char DecodeTrigraphChar	(	const char *	CP,
		Lexer *	L
	)

static

DecodeTrigraphChar - If the specified character is a legal trigraph when prefixed with ??, emit a trigraph warning.

If trigraphs are enabled, return the result character. Finally, emit a warning about trigraph use whether trigraphs are enabled or not.

Definition at line 1166 of file Lexer.cpp.

References clang::Lexer::Diag(), clang::Lexer::getLangOpts(), GetTrigraphCharForLetter(), clang::PreprocessorLexer::isLexingRawMode(), clang::isWhitespace(), and P.

Referenced by clang::Lexer::findLocationAfterToken().

◆ findBeginningOfLine()

static const char* findBeginningOfLine	(	StringRef	Buffer,
		unsigned	Offset
	)

static

Returns the pointer that points to the beginning of line that contains the given offset, or null if the offset if invalid.

Definition at line 480 of file Lexer.cpp.

References clang::Lexer::isNewLineEscaped(), clang::isVerticalWhitespace(), and Offset.

Referenced by getBeginningOfFileToken(), and clang::Lexer::getIndentationForLine().

◆ FindConflictEnd()

static const char* FindConflictEnd	(	const char *	CurPtr,
		const char *	BufferEnd,
		ConflictMarkerKind	CMK
	)

static

Find the end of a version control conflict marker.

Definition at line 2810 of file Lexer.cpp.

References clang::CMK_None, clang::CMK_Normal, clang::CMK_Perforce, clang::Lexer::Diag(), End, and clang::PreprocessorLexer::isLexingRawMode().

◆ findPlaceholderEnd()

static const char* findPlaceholderEnd	(	const char *	CurPtr,
		const char *	BufferEnd
	)

static

Definition at line 2914 of file Lexer.cpp.

References clang::frontend::After, clang::C, clang::Token::clearFlag(), clang::frontend::CPlusPlus, clang::frontend::CPlusPlus2a, clang::FixItHint::CreateInsertion(), clang::FixItHint::CreateRemoval(), clang::FixItHint::CreateReplacement(), clang::Lexer::Diag(), End, clang::Preprocessor::getCodeCompletionLoc(), clang::Preprocessor::getDiagnostics(), clang::Lexer::getLangOpts(), clang::SourceLocation::getLocWithOffset(), clang::Preprocessor::getPreprocessorOpts(), clang::Lexer::getSourceLocation(), clang::Token::hasLeadingEmptyMacro(), clang::Token::hasLeadingSpace(), clang::Token::HasUCN, clang::Lexer::inKeepCommentMode(), isAllowedIDChar(), isAllowedInitiallyIDChar(), clang::isASCII(), clang::Token::isAtStartOfLine(), clang::Preprocessor::isCodeCompletionEnabled(), clang::Token::IsEditorPlaceholder, clang::isHorizontalWhitespace(), clang::Lexer::isKeepWhitespaceMode(), clang::PreprocessorLexer::isLexingRawMode(), clang::Preprocessor::isPreprocessedOutput(), clang::Token::LeadingEmptyMacro, clang::Token::LeadingSpace, clang::PreprocessorOptions::LexEditorPlaceholders, clang::PreprocessorLexer::LexingRawMode, clang::Preprocessor::LookUpIdentifierInfo(), makeCharRange(), maybeDiagnoseIDCharCompat(), maybeDiagnoseUTF8Homoglyph(), clang::PreprocessorLexer::MIOpt, clang::Token::NeedsCleaning, clang::PreprocessorLexer::ParsingFilename, clang::PreprocessorLexer::ParsingPreprocessorDirective, clang::MultipleIncludeOpt::ReadToken(), clang::Lexer::resetExtendedTokenMode(), clang::Token::setFlag(), clang::Token::setIdentifierInfo(), clang::Token::setRawIdentifierData(), clang::Token::StartOfLine, clang::Token::startToken(), and UnicodeWhitespaceCharRanges.

◆ getBeginningOfFileToken()

static SourceLocation getBeginningOfFileToken	(	SourceLocation	Loc,
		const SourceManager &	SM,
		const LangOptions &	LangOpts
	)

static

Definition at line 497 of file Lexer.cpp.

References clang::comments::tok::eof, findBeginningOfLine(), clang::SourceManager::getBufferData(), clang::SourceManager::getDecomposedLoc(), clang::Token::getKind(), clang::Token::getLength(), clang::Token::getLocation(), clang::SourceLocation::getLocWithOffset(), clang::SourceLocation::isFileID(), and clang::Lexer::SetCommentRetentionState().

Referenced by clang::Lexer::GetBeginningOfToken().

◆ GetMappedTokenLoc()

static SourceLocation GetMappedTokenLoc	(	Preprocessor &	PP,
		SourceLocation	FileLoc,
		unsigned	CharNo,
		unsigned	TokLen
	)

static

GetMappedTokenLoc - If lexing out of a 'mapped buffer', where we pretend the lexer buffer was all expanded at a single point, perform the mapping.

This is currently only used for _Pragma implementation, so it is the slow path of the hot getSourceLocation method. Do not allow it to be inlined.

Definition at line 1094 of file Lexer.cpp.

References clang::CharSourceRange::getBegin(), clang::CharSourceRange::getEnd(), clang::SourceLocation::getLocWithOffset(), clang::Preprocessor::getSourceManager(), clang::SourceLocation::isMacroID(), and SM.

Referenced by clang::Lexer::getIndentationForLine(), and clang::Lexer::getSourceLocation().

◆ getSpellingSlow()

static size_t getSpellingSlow	(	const Token &	Tok,
		const char *	BufPtr,
		const LangOptions &	LangOpts,
		char *	Spelling
	)

static

Slow case of getSpelling.

Extract the characters comprising the spelling of this token from the provided input buffer.

Definition at line 261 of file Lexer.cpp.

References clang::Lexer::getCharAndSizeNoWarn(), clang::Token::getKind(), clang::Token::getLength(), clang::tok::isStringLiteral(), memcpy(), and clang::Token::needsCleaning().

Referenced by clang::Lexer::getSpelling().

◆ GetTrigraphCharForLetter()

static char GetTrigraphCharForLetter ( char Letter )

static

GetTrigraphCharForLetter - Given a character that occurs after a ?? pair, return the decoded trigraph letter it corresponds to, or '\0' if nothing.

Definition at line 1147 of file Lexer.cpp.

Referenced by DecodeTrigraphChar(), and clang::Lexer::findLocationAfterToken().

◆ isAllowedIDChar()

static bool isAllowedIDChar	(	uint32_t	C,
		const LangOptions &	LangOpts
	)

static

Definition at line 1423 of file Lexer.cpp.

References C11AllowedIDCharRanges, C99AllowedIDCharRanges, and CXX03AllowedIDCharRanges.

Referenced by findPlaceholderEnd(), isAllowedInitiallyIDChar(), and maybeDiagnoseUTF8Homoglyph().

◆ isAllowedInitiallyIDChar()

static bool isAllowedInitiallyIDChar	(	uint32_t	C,
		const LangOptions &	LangOpts
	)

static

Definition at line 1441 of file Lexer.cpp.

References C11DisallowedInitialIDCharRanges, C99DisallowedInitialIDCharRanges, and isAllowedIDChar().

Referenced by findPlaceholderEnd().

◆ isEndOfBlockCommentWithEscapedNewLine()

static bool isEndOfBlockCommentWithEscapedNewLine	(	const char *	CurPtr,
		Lexer *	L
	)

static

isBlockCommentEndOfEscapedNewLine - Return true if the specified newline character (either \n or \r) is part of an escaped newline sequence.

Issue a diagnostic if so. We know that the newline is inside of a block comment.

Definition at line 2405 of file Lexer.cpp.

References _mm_cmpeq_epi8(), _mm_movemask_epi8(), _mm_set1_epi8(), clang::C, clang::Preprocessor::CodeCompleteNaturalLanguage(), clang::Lexer::Diag(), clang::Preprocessor::getCodeCompletionFileLoc(), clang::Lexer::getLangOpts(), clang::Lexer::getSourceLocation(), clang::Preprocessor::HandleComment(), clang::Lexer::inKeepCommentMode(), clang::isHorizontalWhitespace(), clang::Lexer::isKeepWhitespaceMode(), clang::PreprocessorLexer::isLexingRawMode(), clang::Token::LeadingSpace, clang::Token::setFlag(), and vec_any_eq().

◆ makeCharRange()

static CharSourceRange makeCharRange	(	Lexer &	L,
		const char *	Begin,
		const char *	End
	)

inlinestatic

Definition at line 1458 of file Lexer.cpp.

References clang::CharSourceRange::getCharRange(), and clang::Lexer::getSourceLocation().

Referenced by findPlaceholderEnd(), and maybeDiagnoseUTF8Homoglyph().

◆ makeRangeFromFileLocs()

static CharSourceRange makeRangeFromFileLocs	(	CharSourceRange	Range,
		const SourceManager &	SM,
		const LangOptions &	LangOpts
	)

static

Definition at line 840 of file Lexer.cpp.

References Begin, End, clang::PreprocessorLexer::FID, clang::CharSourceRange::getBegin(), clang::CharSourceRange::getCharRange(), clang::SourceManager::getDecomposedLoc(), clang::CharSourceRange::getEnd(), clang::Lexer::getLocForEndOfToken(), clang::SourceLocation::isFileID(), clang::SourceManager::isInFileID(), clang::FileID::isInvalid(), clang::SourceLocation::isInvalid(), and clang::CharSourceRange::isTokenRange().

Referenced by clang::Lexer::makeFileCharRange().

◆ maybeDiagnoseIDCharCompat()

static void maybeDiagnoseIDCharCompat	(	DiagnosticsEngine &	Diags,
		uint32_t	C,
		CharSourceRange	Range,
		bool	IsFirst
	)

static

Definition at line 1464 of file Lexer.cpp.

References C99AllowedIDCharRanges, C99DisallowedInitialIDCharRanges, CXX03AllowedIDCharRanges, clang::CharSourceRange::getBegin(), clang::DiagnosticsEngine::isIgnored(), and clang::DiagnosticsEngine::Report().

Referenced by findPlaceholderEnd(), and maybeDiagnoseUTF8Homoglyph().

◆ maybeDiagnoseUTF8Homoglyph()

static void maybeDiagnoseUTF8Homoglyph	(	DiagnosticsEngine &	Diags,
		uint32_t	C,
		CharSourceRange	Range
	)

static

After encountering UTF-8 character C and interpreting it as an identifier character, check whether it's a homoglyph for a common non-identifier source character that is unlikely to be an intentional identifier character and warn if so.

Definition at line 1503 of file Lexer.cpp.

References clang::C, clang::frontend::CPlusPlus14, clang::frontend::CPlusPlus17, clang::Lexer::Diag(), clang::CharSourceRange::getBegin(), clang::Lexer::getCharAndSizeNoWarn(), clang::Preprocessor::getDiagnostics(), clang::Lexer::getLangOpts(), clang::Preprocessor::HandleIdentifier(), clang::Token::HasUCN, isAllowedIDChar(), clang::isASCII(), clang::IdentifierInfo::isHandleIdentifierCase(), clang::isIdentifierBody(), clang::PreprocessorLexer::isLexingRawMode(), clang::isPreprocessingNumberBody(), clang::PreprocessorLexer::LexingRawMode, clang::Preprocessor::LookUpIdentifierInfo(), makeCharRange(), maybeDiagnoseIDCharCompat(), clang::operator<(), clang::DiagnosticsEngine::Report(), clang::Result, clang::Token::setFlag(), clang::Token::setKind(), and clang::Token::setRawIdentifierData().

Referenced by findPlaceholderEnd().

◆ StringifyImpl()

template<typename T >

static void StringifyImpl	(	T &	Str,
		char	Quote
	)

static

Definition at line 221 of file Lexer.cpp.

Referenced by clang::Lexer::Stringify().

Enumerations

Functions

Enumeration Type Documentation

◆ PreambleDirectiveKind

Function Documentation

◆ DecodeTrigraphChar()

◆ findBeginningOfLine()

◆ FindConflictEnd()

◆ findPlaceholderEnd()

◆ getBeginningOfFileToken()

◆ GetMappedTokenLoc()

◆ getSpellingSlow()

◆ GetTrigraphCharForLetter()

◆ isAllowedIDChar()

◆ isAllowedInitiallyIDChar()

◆ isEndOfBlockCommentWithEscapedNewLine()

◆ makeCharRange()

◆ makeRangeFromFileLocs()

◆ maybeDiagnoseIDCharCompat()

◆ maybeDiagnoseUTF8Homoglyph()

◆ StringifyImpl()