#include "clang/Lex/Lexer.h"
#include "UnicodeCharSets.h"
#include "clang/Basic/CharInfo.h"
#include "clang/Basic/Diagnostic.h"
#include "clang/Basic/IdentifierTable.h"
#include "clang/Basic/LLVM.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 "llvm/ADT/STLExtras.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/ADT/StringRef.h"
#include "llvm/ADT/StringSwitch.h"
#include "llvm/Support/Compiler.h"
#include "llvm/Support/ConvertUTF.h"
#include "llvm/Support/MathExtras.h"
#include "llvm/Support/MemoryBufferRef.h"
#include "llvm/Support/NativeFormatting.h"
#include "llvm/Support/Unicode.h"
#include "llvm/Support/UnicodeCharRanges.h"
#include <algorithm>
#include <cassert>
#include <cstddef>
#include <cstdint>
#include <cstring>
#include <optional>
#include <string>
#include <tuple>
#include <utility>

Include dependency graph for Lexer.cpp:

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.

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.

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

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

static bool	isInExpansionTokenRange (const SourceLocation Loc, const SourceManager &SM)

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.

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.

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

static bool	isUnicodeWhitespace (uint32_t Codepoint)

static llvm::SmallString< 5 >	codepointAsHexString (uint32_t C)

static bool	isMathematicalExtensionID (uint32_t C, const LangOptions &LangOpts, bool IsStart, bool &IsExtension)

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

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

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

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.

static void	diagnoseInvalidUnicodeCodepointInIdentifier (DiagnosticsEngine &Diags, const LangOptions &LangOpts, uint32_t CodePoint, CharSourceRange Range, bool IsFirst)

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

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

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

Function Documentation

◆ codepointAsHexString()

static llvm::SmallString< 5 > codepointAsHexString ( uint32_t C )

static

Definition at line 1462 of file Lexer.cpp.

References clang::C.

Referenced by diagnoseExtensionInIdentifier(), diagnoseInvalidUnicodeCodepointInIdentifier(), and maybeDiagnoseUTF8Homoglyph().

◆ DecodeTrigraphChar()

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

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 1197 of file Lexer.cpp.

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

◆ diagnoseExtensionInIdentifier()

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

static

Definition at line 1543 of file Lexer.cpp.

References clang::C, codepointAsHexString(), MathematicalNotationProfileIDContinueRanges, MathematicalNotationProfileIDStartRanges, and clang::DiagnosticsEngine::Report().

◆ diagnoseInvalidUnicodeCodepointInIdentifier()

static void diagnoseInvalidUnicodeCodepointInIdentifier	(	DiagnosticsEngine &	Diags,
		const LangOptions &	LangOpts,
		uint32_t	CodePoint,
		CharSourceRange	Range,
		bool	IsFirst
	)

static

Definition at line 1669 of file Lexer.cpp.

References codepointAsHexString(), clang::FixItHint::CreateRemoval(), int, isAllowedIDChar(), isAllowedInitiallyIDChar(), clang::isASCII(), and clang::DiagnosticsEngine::Report().

◆ 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 495 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 3115 of file Lexer.cpp.

References clang::CMK_Perforce.

◆ findPlaceholderEnd()

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

static

Definition at line 3219 of file Lexer.cpp.

◆ getBeginningOfFileToken()

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

static

Definition at line 512 of file Lexer.cpp.

References findBeginningOfLine(), clang::Lexer::getBufferLocation(), clang::Token::getKind(), clang::Token::getLength(), clang::Token::getLocation(), clang::SourceLocation::getLocWithOffset(), clang::Invalid, clang::SourceLocation::isFileID(), clang::Lexer::LexFromRawLexer(), clang::Lexer::SetCommentRetentionState(), and SM.

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 1125 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::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 276 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 1178 of file Lexer.cpp.

Referenced by DecodeTrigraphChar().

◆ isAllowedIDChar()

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

static

Definition at line 1489 of file Lexer.cpp.

References clang::C, C11AllowedIDCharRanges, C99AllowedIDCharRanges, isMathematicalExtensionID(), XIDContinueRanges, and XIDStartRanges.

Referenced by diagnoseInvalidUnicodeCodepointInIdentifier(), and isAllowedInitiallyIDChar().

◆ isAllowedInitiallyIDChar()

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

static

Definition at line 1517 of file Lexer.cpp.

References clang::C, C11DisallowedInitialIDCharRanges, C99DisallowedInitialIDCharRanges, isAllowedIDChar(), isMathematicalExtensionID(), and XIDStartRanges.

Referenced by diagnoseInvalidUnicodeCodepointInIdentifier().

◆ isEndOfBlockCommentWithEscapedNewLine()

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

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 2631 of file Lexer.cpp.

References clang::Lexer::Diag(), clang::isHorizontalWhitespace(), and clang::PreprocessorLexer::isLexingRawMode().

◆ isInExpansionTokenRange()

static bool isInExpansionTokenRange	(	const SourceLocation	Loc,
		const SourceManager &	SM
	)

static

Definition at line 884 of file Lexer.cpp.

References SM.

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

◆ isMathematicalExtensionID()

static bool isMathematicalExtensionID	(	uint32_t	C,
		const LangOptions &	LangOpts,
		bool	IsStart,
		bool &	IsExtension
	)

static

Definition at line 1475 of file Lexer.cpp.

References clang::C, MathematicalNotationProfileIDContinueRanges, and MathematicalNotationProfileIDStartRanges.

Referenced by isAllowedIDChar(), and isAllowedInitiallyIDChar().

◆ isUnicodeWhitespace()

static bool isUnicodeWhitespace ( uint32_t Codepoint )

static

Definition at line 1456 of file Lexer.cpp.

References UnicodeWhitespaceCharRanges.

◆ makeCharRange()

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

inlinestatic

Definition at line 1559 of file Lexer.cpp.

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

◆ makeRangeFromFileLocs()

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

static

Definition at line 856 of file Lexer.cpp.

References Begin, clang::CharSourceRange::getCharRange(), clang::Lexer::getLocForEndOfToken(), clang::SourceLocation::isFileID(), clang::FileID::isInvalid(), and SM.

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

◆ maybeDiagnoseIDCharCompat()

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

static

Definition at line 1565 of file Lexer.cpp.

References clang::C, C99AllowedIDCharRanges, C99DisallowedInitialIDCharRanges, clang::DiagnosticsEngine::isIgnored(), and clang::DiagnosticsEngine::Report().

◆ 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 1594 of file Lexer.cpp.

References clang::C, codepointAsHexString(), clang::operator<(), and clang::DiagnosticsEngine::Report().

◆ StringifyImpl()

template<typename T >

static void StringifyImpl	(	T &	Str,
		char	Quote
	)

static

Definition at line 236 of file Lexer.cpp.

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

Functions

Function Documentation

◆ codepointAsHexString()

◆ DecodeTrigraphChar()

◆ diagnoseExtensionInIdentifier()

◆ diagnoseInvalidUnicodeCodepointInIdentifier()

◆ findBeginningOfLine()

◆ FindConflictEnd()

◆ findPlaceholderEnd()

◆ getBeginningOfFileToken()

◆ GetMappedTokenLoc()

◆ getSpellingSlow()

◆ GetTrigraphCharForLetter()

◆ isAllowedIDChar()

◆ isAllowedInitiallyIDChar()

◆ isEndOfBlockCommentWithEscapedNewLine()

◆ isInExpansionTokenRange()

◆ isMathematicalExtensionID()

◆ isUnicodeWhitespace()

◆ makeCharRange()

◆ makeRangeFromFileLocs()

◆ maybeDiagnoseIDCharCompat()

◆ maybeDiagnoseUTF8Homoglyph()

◆ StringifyImpl()