Subversion Repositories QNX 8.QNX8 LLVM/Clang compiler suite

//===- Lexer.h - C Language Family Lexer ------------------------*- C++ -*-===//

//

// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.

// See https://llvm.org/LICENSE.txt for license information.

// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception

//

//===----------------------------------------------------------------------===//

//

//  This file defines the Lexer interface.

//

//===----------------------------------------------------------------------===//

#ifndef LLVM_CLANG_LEX_LEXER_H

#define LLVM_CLANG_LEX_LEXER_H

#include "clang/Basic/LangOptions.h"

#include "clang/Basic/SourceLocation.h"

#include "clang/Basic/TokenKinds.h"

#include "clang/Lex/DependencyDirectivesScanner.h"

#include "clang/Lex/PreprocessorLexer.h"

#include "clang/Lex/Token.h"

#include "llvm/ADT/SmallVector.h"

#include "llvm/ADT/StringRef.h"

#include <cassert>

#include <cstdint>

#include <optional>

#include <string>

namespace llvm {

class MemoryBufferRef;

} // namespace llvm

namespace clang {

class DiagnosticBuilder;

class Preprocessor;

class SourceManager;

class LangOptions;

/// ConflictMarkerKind - Kinds of conflict marker which the lexer might be

/// recovering from.

enum ConflictMarkerKind {

  /// Not within a conflict marker.

  CMK_None,

  /// A normal or diff3 conflict marker, initiated by at least 7 "<"s,

  /// separated by at least 7 "="s or "|"s, and terminated by at least 7 ">"s.

  CMK_Normal,

  /// A Perforce-style conflict marker, initiated by 4 ">"s,

  /// separated by 4 "="s, and terminated by 4 "<"s.

  CMK_Perforce

};

/// Describes the bounds (start, size) of the preamble and a flag required by

/// PreprocessorOptions::PrecompiledPreambleBytes.

/// The preamble includes the BOM, if any.

struct PreambleBounds {

  /// Size of the preamble in bytes.

  unsigned Size;

  /// Whether the preamble ends at the start of a new line.

  ///

  /// Used to inform the lexer as to whether it's starting at the beginning of

  /// a line after skipping the preamble.

  bool PreambleEndsAtStartOfLine;

  PreambleBounds(unsigned Size, bool PreambleEndsAtStartOfLine)

      : Size(Size), PreambleEndsAtStartOfLine(PreambleEndsAtStartOfLine) {}

};

/// Lexer - This provides a simple interface that turns a text buffer into a

/// stream of tokens.  This provides no support for file reading or buffering,

/// or buffering/seeking of tokens, only forward lexing is supported.  It relies

/// on the specified Preprocessor object to handle preprocessor directives, etc.

class Lexer : public PreprocessorLexer {

  friend class Preprocessor;

  void anchor() override;

  //===--------------------------------------------------------------------===//

  // Constant configuration values for this lexer.

  // Start of the buffer.

  const char *BufferStart;

  // End of the buffer.

  const char *BufferEnd;

  // Location for start of file.

  SourceLocation FileLoc;

  // LangOpts enabled by this language.

  // Storing LangOptions as reference here is important from performance point

  // of view. Lack of reference means that LangOptions copy constructor would be

  // called by Lexer(..., const LangOptions &LangOpts,...). Given that local

  // Lexer objects are created thousands times (in Lexer::getRawToken,

  // Preprocessor::EnterSourceFile and other places) during single module

  // processing in frontend it would make std::vector<std::string> copy

  // constructors surprisingly hot.

  const LangOptions &LangOpts;

  // True if '//' line comments are enabled.

  bool LineComment;

  // True if lexer for _Pragma handling.

  bool Is_PragmaLexer;

  //===--------------------------------------------------------------------===//

  // Context-specific lexing flags set by the preprocessor.

  //

  /// ExtendedTokenMode - The lexer can optionally keep comments and whitespace

  /// and return them as tokens.  This is used for -C and -CC modes, and

  /// whitespace preservation can be useful for some clients that want to lex

  /// the file in raw mode and get every character from the file.

  ///

  /// When this is set to 2 it returns comments and whitespace.  When set to 1

  /// it returns comments, when it is set to 0 it returns normal tokens only.

  unsigned char ExtendedTokenMode;

  //===--------------------------------------------------------------------===//

  // Context that changes as the file is lexed.

  // NOTE: any state that mutates when in raw mode must have save/restore code

  // in Lexer::isNextPPTokenLParen.

  // BufferPtr - Current pointer into the buffer.  This is the next character

  // to be lexed.

  const char *BufferPtr;

  // IsAtStartOfLine - True if the next lexed token should get the "start of

  // line" flag set on it.

  bool IsAtStartOfLine;

  bool IsAtPhysicalStartOfLine;

  bool HasLeadingSpace;

  bool HasLeadingEmptyMacro;

  /// True if this is the first time we're lexing the input file.

  bool IsFirstTimeLexingFile;

  // NewLinePtr - A pointer to new line character '\n' being lexed. For '\r\n',

  // it also points to '\n.'

  const char *NewLinePtr;

  // CurrentConflictMarkerState - The kind of conflict marker we are handling.

  ConflictMarkerKind CurrentConflictMarkerState;

  /// Non-empty if this \p Lexer is \p isDependencyDirectivesLexer().

  ArrayRef<dependency_directives_scan::Directive> DepDirectives;

  /// If this \p Lexer is \p isDependencyDirectivesLexer(), it represents the

  /// next token to use from the current dependency directive.

  unsigned NextDepDirectiveTokenIndex = 0;

  void InitLexer(const char *BufStart, const char *BufPtr, const char *BufEnd);

public:

  /// Lexer constructor - Create a new lexer object for the specified buffer

  /// with the specified preprocessor managing the lexing process.  This lexer

  /// assumes that the associated file buffer and Preprocessor objects will

  /// outlive it, so it doesn't take ownership of either of them.

  Lexer(FileID FID, const llvm::MemoryBufferRef &InputFile, Preprocessor &PP,

        bool IsFirstIncludeOfFile = true);

  /// Lexer constructor - Create a new raw lexer object.  This object is only

  /// suitable for calls to 'LexFromRawLexer'.  This lexer assumes that the

  /// text range will outlive it, so it doesn't take ownership of it.

  Lexer(SourceLocation FileLoc, const LangOptions &LangOpts,

        const char *BufStart, const char *BufPtr, const char *BufEnd,

        bool IsFirstIncludeOfFile = true);

  /// Lexer constructor - Create a new raw lexer object.  This object is only

  /// suitable for calls to 'LexFromRawLexer'.  This lexer assumes that the

  /// text range will outlive it, so it doesn't take ownership of it.

  Lexer(FileID FID, const llvm::MemoryBufferRef &FromFile,

        const SourceManager &SM, const LangOptions &LangOpts,

        bool IsFirstIncludeOfFile = true);

  Lexer(const Lexer &) = delete;

  Lexer &operator=(const Lexer &) = delete;

  /// Create_PragmaLexer: Lexer constructor - Create a new lexer object for

  /// _Pragma expansion.  This has a variety of magic semantics that this method

  /// sets up.  It returns a new'd Lexer that must be delete'd when done.

  static Lexer *Create_PragmaLexer(SourceLocation SpellingLoc,

                                   SourceLocation ExpansionLocStart,

                                   SourceLocation ExpansionLocEnd,

                                   unsigned TokLen, Preprocessor &PP);

  /// getFileLoc - Return the File Location for the file we are lexing out of.

  /// The physical location encodes the location where the characters come from,

  /// the virtual location encodes where we should *claim* the characters came

  /// from.  Currently this is only used by _Pragma handling.

  SourceLocation getFileLoc() const { return FileLoc; }

private:

  /// Lex - Return the next token in the file.  If this is the end of file, it

  /// return the tok::eof token.  This implicitly involves the preprocessor.

  bool Lex(Token &Result);

  /// Called when the preprocessor is in 'dependency scanning lexing mode'.

  bool LexDependencyDirectiveToken(Token &Result);

  /// Called when the preprocessor is in 'dependency scanning lexing mode' and

  /// is skipping a conditional block.

  bool LexDependencyDirectiveTokenWhileSkipping(Token &Result);

  /// True when the preprocessor is in 'dependency scanning lexing mode' and

  /// created this \p Lexer for lexing a set of dependency directive tokens.

  bool isDependencyDirectivesLexer() const { return !DepDirectives.empty(); }

  /// Initializes \p Result with data from \p DDTok and advances \p BufferPtr to

  /// the position just after the token.

  /// \returns the buffer pointer at the beginning of the token.

  const char *convertDependencyDirectiveToken(

      const dependency_directives_scan::Token &DDTok, Token &Result);

public:

  /// isPragmaLexer - Returns true if this Lexer is being used to lex a pragma.

  bool isPragmaLexer() const { return Is_PragmaLexer; }

private:

  /// IndirectLex - An indirect call to 'Lex' that can be invoked via

  ///  the PreprocessorLexer interface.

  void IndirectLex(Token &Result) override { Lex(Result); }

public:

  /// LexFromRawLexer - Lex a token from a designated raw lexer (one with no

  /// associated preprocessor object.  Return true if the 'next character to

  /// read' pointer points at the end of the lexer buffer, false otherwise.

  bool LexFromRawLexer(Token &Result) {

    assert(LexingRawMode && "Not already in raw mode!");

    Lex(Result);

    // Note that lexing to the end of the buffer doesn't implicitly delete the

    // lexer when in raw mode.

    return BufferPtr == BufferEnd;

  }

  /// isKeepWhitespaceMode - Return true if the lexer should return tokens for

  /// every character in the file, including whitespace and comments.  This

  /// should only be used in raw mode, as the preprocessor is not prepared to

  /// deal with the excess tokens.

  bool isKeepWhitespaceMode() const {

    return ExtendedTokenMode > 1;

  }

  /// SetKeepWhitespaceMode - This method lets clients enable or disable

  /// whitespace retention mode.

  void SetKeepWhitespaceMode(bool Val) {

    assert((!Val || LexingRawMode || LangOpts.TraditionalCPP) &&

           "Can only retain whitespace in raw mode or -traditional-cpp");

    ExtendedTokenMode = Val ? 2 : 0;

  }

  /// inKeepCommentMode - Return true if the lexer should return comments as

  /// tokens.

  bool inKeepCommentMode() const {

    return ExtendedTokenMode > 0;

  }

  /// SetCommentRetentionMode - Change the comment retention mode of the lexer

  /// to the specified mode.  This is really only useful when lexing in raw

  /// mode, because otherwise the lexer needs to manage this.

  void SetCommentRetentionState(bool Mode) {

    assert(!isKeepWhitespaceMode() &&

           "Can't play with comment retention state when retaining whitespace");

    ExtendedTokenMode = Mode ? 1 : 0;

  }

  /// Sets the extended token mode back to its initial value, according to the

  /// language options and preprocessor. This controls whether the lexer

  /// produces comment and whitespace tokens.

  ///

  /// This requires the lexer to have an associated preprocessor. A standalone

  /// lexer has nothing to reset to.

  void resetExtendedTokenMode();

  /// Gets source code buffer.

  StringRef getBuffer() const {

    return StringRef(BufferStart, BufferEnd - BufferStart);

  }

  /// ReadToEndOfLine - Read the rest of the current preprocessor line as an

  /// uninterpreted string.  This switches the lexer out of directive mode.

  void ReadToEndOfLine(SmallVectorImpl<char> *Result = nullptr);

  /// Diag - Forwarding function for diagnostics.  This translate a source

  /// position in the current buffer into a SourceLocation object for rendering.

  DiagnosticBuilder Diag(const char *Loc, unsigned DiagID) const;

  /// getSourceLocation - Return a source location identifier for the specified

  /// offset in the current file.

  SourceLocation getSourceLocation(const char *Loc, unsigned TokLen = 1) const;

  /// getSourceLocation - Return a source location for the next character in

  /// the current file.

  SourceLocation getSourceLocation() override {

    return getSourceLocation(BufferPtr);

  }

  /// Return the current location in the buffer.

  const char *getBufferLocation() const { return BufferPtr; }

  /// Returns the current lexing offset.

  unsigned getCurrentBufferOffset() {

    assert(BufferPtr >= BufferStart && "Invalid buffer state");

    return BufferPtr - BufferStart;

  }

  /// Set the lexer's buffer pointer to \p Offset.

  void seek(unsigned Offset, bool IsAtStartOfLine);

  /// Stringify - Convert the specified string into a C string by i) escaping

  /// '\\' and " characters and ii) replacing newline character(s) with "\\n".

  /// If Charify is true, this escapes the ' character instead of ".

  static std::string Stringify(StringRef Str, bool Charify = false);

  /// Stringify - Convert the specified string into a C string by i) escaping

  /// '\\' and " characters and ii) replacing newline character(s) with "\\n".

  static void Stringify(SmallVectorImpl<char> &Str);

  /// getSpelling - This method is used to get the spelling of a token into a

  /// preallocated buffer, instead of as an std::string.  The caller is required

  /// to allocate enough space for the token, which is guaranteed to be at least

  /// Tok.getLength() bytes long.  The length of the actual result is returned.

  ///

  /// Note that this method may do two possible things: it may either fill in

  /// the buffer specified with characters, or it may *change the input pointer*

  /// to point to a constant buffer with the data already in it (avoiding a

  /// copy).  The caller is not allowed to modify the returned buffer pointer

  /// if an internal buffer is returned.

  static unsigned getSpelling(const Token &Tok, const char *&Buffer,

                              const SourceManager &SourceMgr,

                              const LangOptions &LangOpts,

                              bool *Invalid = nullptr);

  /// getSpelling() - Return the 'spelling' of the Tok token.  The spelling of a

  /// token is the characters used to represent the token in the source file

  /// after trigraph expansion and escaped-newline folding.  In particular, this

  /// wants to get the true, uncanonicalized, spelling of things like digraphs

  /// UCNs, etc.

  static std::string getSpelling(const Token &Tok,

                                 const SourceManager &SourceMgr,

                                 const LangOptions &LangOpts,

                                 bool *Invalid = nullptr);

  /// getSpelling - This method is used to get the spelling of the

  /// token at the given source location.  If, as is usually true, it

  /// is not necessary to copy any data, then the returned string may

  /// not point into the provided buffer.

  ///

  /// This method lexes at the expansion depth of the given

  /// location and does not jump to the expansion or spelling

  /// location.

  static StringRef getSpelling(SourceLocation loc,

                               SmallVectorImpl<char> &buffer,

                               const SourceManager &SM,

                               const LangOptions &options,

                               bool *invalid = nullptr);

  /// MeasureTokenLength - Relex the token at the specified location and return

  /// its length in bytes in the input file.  If the token needs cleaning (e.g.

  /// includes a trigraph or an escaped newline) then this count includes bytes

  /// that are part of that.

  static unsigned MeasureTokenLength(SourceLocation Loc,

                                     const SourceManager &SM,

                                     const LangOptions &LangOpts);

  /// Relex the token at the specified location.

  /// \returns true if there was a failure, false on success.

  static bool getRawToken(SourceLocation Loc, Token &Result,

                          const SourceManager &SM,

                          const LangOptions &LangOpts,

                          bool IgnoreWhiteSpace = false);

  /// Given a location any where in a source buffer, find the location

  /// that corresponds to the beginning of the token in which the original

  /// source location lands.

  static SourceLocation GetBeginningOfToken(SourceLocation Loc,

                                            const SourceManager &SM,

                                            const LangOptions &LangOpts);

  /// Get the physical length (including trigraphs and escaped newlines) of the

  /// first \p Characters characters of the token starting at TokStart.

  static unsigned getTokenPrefixLength(SourceLocation TokStart,

                                       unsigned CharNo,

                                       const SourceManager &SM,

                                       const LangOptions &LangOpts);

  /// AdvanceToTokenCharacter - If the current SourceLocation specifies a

  /// location at the start of a token, return a new location that specifies a

  /// character within the token.  This handles trigraphs and escaped newlines.

  static SourceLocation AdvanceToTokenCharacter(SourceLocation TokStart,

                                                unsigned Characters,

                                                const SourceManager &SM,

                                                const LangOptions &LangOpts) {

    return TokStart.getLocWithOffset(

        getTokenPrefixLength(TokStart, Characters, SM, LangOpts));

  }

  /// Computes the source location just past the end of the

  /// token at this source location.

  ///

  /// This routine can be used to produce a source location that

  /// points just past the end of the token referenced by \p Loc, and

  /// is generally used when a diagnostic needs to point just after a

  /// token where it expected something different that it received. If

  /// the returned source location would not be meaningful (e.g., if

  /// it points into a macro), this routine returns an invalid

  /// source location.

  ///

  /// \param Offset an offset from the end of the token, where the source

  /// location should refer to. The default offset (0) produces a source

  /// location pointing just past the end of the token; an offset of 1 produces

  /// a source location pointing to the last character in the token, etc.

  static SourceLocation getLocForEndOfToken(SourceLocation Loc, unsigned Offset,

                                            const SourceManager &SM,

                                            const LangOptions &LangOpts);

  /// Given a token range, produce a corresponding CharSourceRange that

  /// is not a token range. This allows the source range to be used by

  /// components that don't have access to the lexer and thus can't find the

  /// end of the range for themselves.

  static CharSourceRange getAsCharRange(SourceRange Range,

                                        const SourceManager &SM,

                                        const LangOptions &LangOpts) {

    SourceLocation End = getLocForEndOfToken(Range.getEnd(), 0, SM, LangOpts);

    return End.isInvalid() ? CharSourceRange()

                           : CharSourceRange::getCharRange(

                                 Range.getBegin(), End);

  }

  static CharSourceRange getAsCharRange(CharSourceRange Range,

                                        const SourceManager &SM,

                                        const LangOptions &LangOpts) {

    return Range.isTokenRange()

               ? getAsCharRange(Range.getAsRange(), SM, LangOpts)

               : Range;

  }

  /// Returns true if the given MacroID location points at the first

  /// token of the macro expansion.

  ///

  /// \param MacroBegin If non-null and function returns true, it is set to

  /// begin location of the macro.

  static bool isAtStartOfMacroExpansion(SourceLocation loc,

                                        const SourceManager &SM,

                                        const LangOptions &LangOpts,

                                        SourceLocation *MacroBegin = nullptr);

  /// Returns true if the given MacroID location points at the last

  /// token of the macro expansion.

  ///

  /// \param MacroEnd If non-null and function returns true, it is set to

  /// end location of the macro.

  static bool isAtEndOfMacroExpansion(SourceLocation loc,

                                      const SourceManager &SM,

                                      const LangOptions &LangOpts,

                                      SourceLocation *MacroEnd = nullptr);

  /// Accepts a range and returns a character range with file locations.

  ///

  /// Returns a null range if a part of the range resides inside a macro

  /// expansion or the range does not reside on the same FileID.

  ///

  /// This function is trying to deal with macros and return a range based on

  /// file locations. The cases where it can successfully handle macros are:

  ///

  /// -begin or end range lies at the start or end of a macro expansion, in

  ///  which case the location will be set to the expansion point, e.g:

  ///    \#define M 1 2

  ///    a M

  /// If you have a range [a, 2] (where 2 came from the macro), the function

  /// will return a range for "a M"

  /// if you have range [a, 1], the function will fail because the range

  /// overlaps with only a part of the macro

  ///

  /// -The macro is a function macro and the range can be mapped to the macro

  ///  arguments, e.g:

  ///    \#define M 1 2

  ///    \#define FM(x) x

  ///    FM(a b M)

  /// if you have range [b, 2], the function will return the file range "b M"

  /// inside the macro arguments.

  /// if you have range [a, 2], the function will return the file range

  /// "FM(a b M)" since the range includes all of the macro expansion.

  static CharSourceRange makeFileCharRange(CharSourceRange Range,

                                           const SourceManager &SM,

                                           const LangOptions &LangOpts);

  /// Returns a string for the source that the range encompasses.

  static StringRef getSourceText(CharSourceRange Range,

                                 const SourceManager &SM,

                                 const LangOptions &LangOpts,

                                 bool *Invalid = nullptr);

  /// Retrieve the name of the immediate macro expansion.

  ///

  /// This routine starts from a source location, and finds the name of the macro

  /// responsible for its immediate expansion. It looks through any intervening

  /// macro argument expansions to compute this. It returns a StringRef which

  /// refers to the SourceManager-owned buffer of the source where that macro

  /// name is spelled. Thus, the result shouldn't out-live that SourceManager.

  static StringRef getImmediateMacroName(SourceLocation Loc,

                                         const SourceManager &SM,

                                         const LangOptions &LangOpts);

  /// Retrieve the name of the immediate macro expansion.

  ///

  /// This routine starts from a source location, and finds the name of the

  /// macro responsible for its immediate expansion. It looks through any

  /// intervening macro argument expansions to compute this. It returns a

  /// StringRef which refers to the SourceManager-owned buffer of the source

  /// where that macro name is spelled. Thus, the result shouldn't out-live

  /// that SourceManager.

  ///

  /// This differs from Lexer::getImmediateMacroName in that any macro argument

  /// location will result in the topmost function macro that accepted it.

  /// e.g.

  /// \code

  ///   MAC1( MAC2(foo) )

  /// \endcode

  /// for location of 'foo' token, this function will return "MAC1" while

  /// Lexer::getImmediateMacroName will return "MAC2".

  static StringRef getImmediateMacroNameForDiagnostics(

      SourceLocation Loc, const SourceManager &SM, const LangOptions &LangOpts);

  /// Compute the preamble of the given file.

  ///

  /// The preamble of a file contains the initial comments, include directives,

  /// and other preprocessor directives that occur before the code in this

  /// particular file actually begins. The preamble of the main source file is

  /// a potential prefix header.

  ///

  /// \param Buffer The memory buffer containing the file's contents.

  ///

  /// \param MaxLines If non-zero, restrict the length of the preamble

  /// to fewer than this number of lines.

  ///

  /// \returns The offset into the file where the preamble ends and the rest

  /// of the file begins along with a boolean value indicating whether

  /// the preamble ends at the beginning of a new line.

  static PreambleBounds ComputePreamble(StringRef Buffer,

                                        const LangOptions &LangOpts,

                                        unsigned MaxLines = 0);

  /// Finds the token that comes right after the given location.

  ///

  /// Returns the next token, or none if the location is inside a macro.

  static std::optional<Token> findNextToken(SourceLocation Loc,

                                            const SourceManager &SM,

                                            const LangOptions &LangOpts);

  /// Checks that the given token is the first token that occurs after

  /// the given location (this excludes comments and whitespace). Returns the

  /// location immediately after the specified token. If the token is not found

  /// or the location is inside a macro, the returned source location will be

  /// invalid.

  static SourceLocation findLocationAfterToken(SourceLocation loc,

                                         tok::TokenKind TKind,

                                         const SourceManager &SM,

                                         const LangOptions &LangOpts,

                                         bool SkipTrailingWhitespaceAndNewLine);

  /// Returns true if the given character could appear in an identifier.

  static bool isAsciiIdentifierContinueChar(char c,

                                            const LangOptions &LangOpts);

  /// Checks whether new line pointed by Str is preceded by escape

  /// sequence.

  static bool isNewLineEscaped(const char *BufferStart, const char *Str);

  /// getCharAndSizeNoWarn - Like the getCharAndSize method, but does not ever

  /// emit a warning.

  static inline char getCharAndSizeNoWarn(const char *Ptr, unsigned &Size,

                                          const LangOptions &LangOpts) {

    // If this is not a trigraph and not a UCN or escaped newline, return

    // quickly.

    if (isObviouslySimpleCharacter(Ptr[0])) {

      Size = 1;

      return *Ptr;

    }

    Size = 0;

    return getCharAndSizeSlowNoWarn(Ptr, Size, LangOpts);

  }

  /// Returns the leading whitespace for line that corresponds to the given

  /// location \p Loc.

  static StringRef getIndentationForLine(SourceLocation Loc,

                                         const SourceManager &SM);

  /// Check if this is the first time we're lexing the input file.

  bool isFirstTimeLexingFile() const { return IsFirstTimeLexingFile; }

private:

  //===--------------------------------------------------------------------===//

  // Internal implementation interfaces.

  /// LexTokenInternal - Internal interface to lex a preprocessing token. Called

  /// by Lex.

  ///

  bool LexTokenInternal(Token &Result, bool TokAtPhysicalStartOfLine);

  bool CheckUnicodeWhitespace(Token &Result, uint32_t C, const char *CurPtr);

  bool LexUnicodeIdentifierStart(Token &Result, uint32_t C, const char *CurPtr);

  /// FormTokenWithChars - When we lex a token, we have identified a span

  /// starting at BufferPtr, going to TokEnd that forms the token.  This method

  /// takes that range and assigns it to the token as its location and size.  In

  /// addition, since tokens cannot overlap, this also updates BufferPtr to be

  /// TokEnd.

  void FormTokenWithChars(Token &Result, const char *TokEnd,

                          tok::TokenKind Kind) {

    unsigned TokLen = TokEnd-BufferPtr;

    Result.setLength(TokLen);

    Result.setLocation(getSourceLocation(BufferPtr, TokLen));

    Result.setKind(Kind);

    BufferPtr = TokEnd;

  }

  /// isNextPPTokenLParen - Return 1 if the next unexpanded token will return a

  /// tok::l_paren token, 0 if it is something else and 2 if there are no more

  /// tokens in the buffer controlled by this lexer.

  unsigned isNextPPTokenLParen();

  //===--------------------------------------------------------------------===//

  // Lexer character reading interfaces.

  // This lexer is built on two interfaces for reading characters, both of which

  // automatically provide phase 1/2 translation.  getAndAdvanceChar is used

  // when we know that we will be reading a character from the input buffer and

  // that this character will be part of the result token. This occurs in (f.e.)

  // string processing, because we know we need to read until we find the

  // closing '"' character.

  //

  // The second interface is the combination of getCharAndSize with

  // ConsumeChar.  getCharAndSize reads a phase 1/2 translated character,

  // returning it and its size.  If the lexer decides that this character is

  // part of the current token, it calls ConsumeChar on it.  This two stage

  // approach allows us to emit diagnostics for characters (e.g. warnings about

  // trigraphs), knowing that they only are emitted if the character is

  // consumed.

  /// isObviouslySimpleCharacter - Return true if the specified character is

  /// obviously the same in translation phase 1 and translation phase 3.  This

  /// can return false for characters that end up being the same, but it will

  /// never return true for something that needs to be mapped.

  static bool isObviouslySimpleCharacter(char C) {

    return C != '?' && C != '\\';

  }

  /// getAndAdvanceChar - Read a single 'character' from the specified buffer,

  /// advance over it, and return it.  This is tricky in several cases.  Here we

  /// just handle the trivial case and fall-back to the non-inlined

  /// getCharAndSizeSlow method to handle the hard case.

  inline char getAndAdvanceChar(const char *&Ptr, Token &Tok) {

    // If this is not a trigraph and not a UCN or escaped newline, return

    // quickly.

    if (isObviouslySimpleCharacter(Ptr[0])) return *Ptr++;

    unsigned Size = 0;

    char C = getCharAndSizeSlow(Ptr, Size, &Tok);

    Ptr += Size;

    return C;

  }

  /// ConsumeChar - When a character (identified by getCharAndSize) is consumed

  /// and added to a given token, check to see if there are diagnostics that

  /// need to be emitted or flags that need to be set on the token.  If so, do

  /// it.

  const char *ConsumeChar(const char *Ptr, unsigned Size, Token &Tok) {

    // Normal case, we consumed exactly one token.  Just return it.

    if (Size == 1)

      return Ptr+Size;

    // Otherwise, re-lex the character with a current token, allowing

    // diagnostics to be emitted and flags to be set.

    Size = 0;

    getCharAndSizeSlow(Ptr, Size, &Tok);

    return Ptr+Size;

  }

  /// getCharAndSize - Peek a single 'character' from the specified buffer,

  /// get its size, and return it.  This is tricky in several cases.  Here we

  /// just handle the trivial case and fall-back to the non-inlined

  /// getCharAndSizeSlow method to handle the hard case.

  inline char getCharAndSize(const char *Ptr, unsigned &Size) {

    // If this is not a trigraph and not a UCN or escaped newline, return

    // quickly.

    if (isObviouslySimpleCharacter(Ptr[0])) {

      Size = 1;

      return *Ptr;

    }

    Size = 0;

    return getCharAndSizeSlow(Ptr, Size);

  }

  /// getCharAndSizeSlow - Handle the slow/uncommon case of the getCharAndSize

  /// method.

  char getCharAndSizeSlow(const char *Ptr, unsigned &Size,

                          Token *Tok = nullptr);

  /// getEscapedNewLineSize - Return the size of the specified escaped newline,

  /// or 0 if it is not an escaped newline. P[-1] is known to be a "\" on entry

  /// to this function.

  static unsigned getEscapedNewLineSize(const char *P);

  /// SkipEscapedNewLines - If P points to an escaped newline (or a series of

  /// them), skip over them and return the first non-escaped-newline found,

  /// otherwise return P.

  static const char *SkipEscapedNewLines(const char *P);

  /// getCharAndSizeSlowNoWarn - Same as getCharAndSizeSlow, but never emits a

  /// diagnostic.

  static char getCharAndSizeSlowNoWarn(const char *Ptr, unsigned &Size,

                                       const LangOptions &LangOpts);

  //===--------------------------------------------------------------------===//

  // Other lexer functions.

  void SetByteOffset(unsigned Offset, bool StartOfLine);

  void PropagateLineStartLeadingSpaceInfo(Token &Result);

  const char *LexUDSuffix(Token &Result, const char *CurPtr,

                          bool IsStringLiteral);

  // Helper functions to lex the remainder of a token of the specific type.

  // This function handles both ASCII and Unicode identifiers after

  // the first codepoint of the identifyier has been parsed.

  bool LexIdentifierContinue(Token &Result, const char *CurPtr);

  bool LexNumericConstant    (Token &Result, const char *CurPtr);

  bool LexStringLiteral      (Token &Result, const char *CurPtr,

                              tok::TokenKind Kind);

  bool LexRawStringLiteral   (Token &Result, const char *CurPtr,

                              tok::TokenKind Kind);

  bool LexAngledStringLiteral(Token &Result, const char *CurPtr);

  bool LexCharConstant       (Token &Result, const char *CurPtr,

                              tok::TokenKind Kind);

  bool LexEndOfFile          (Token &Result, const char *CurPtr);

  bool SkipWhitespace        (Token &Result, const char *CurPtr,

                              bool &TokAtPhysicalStartOfLine);

  bool SkipLineComment       (Token &Result, const char *CurPtr,

                              bool &TokAtPhysicalStartOfLine);

  bool SkipBlockComment      (Token &Result, const char *CurPtr,

                              bool &TokAtPhysicalStartOfLine);

  bool SaveLineComment       (Token &Result, const char *CurPtr);

  bool IsStartOfConflictMarker(const char *CurPtr);

  bool HandleEndOfConflictMarker(const char *CurPtr);

  bool lexEditorPlaceholder(Token &Result, const char *CurPtr);

  bool isCodeCompletionPoint(const char *CurPtr) const;

  void cutOffLexing() { BufferPtr = BufferEnd; }

  bool isHexaLiteral(const char *Start, const LangOptions &LangOpts);

  void codeCompleteIncludedFile(const char *PathStart,

                                const char *CompletionPoint, bool IsAngled);

  std::optional<uint32_t>

  tryReadNumericUCN(const char *&StartPtr, const char *SlashLoc, Token *Result);

  std::optional<uint32_t> tryReadNamedUCN(const char *&StartPtr,

                                          const char *SlashLoc, Token *Result);

  /// Read a universal character name.

  ///

  /// \param StartPtr The position in the source buffer after the initial '\'.

  ///                 If the UCN is syntactically well-formed (but not

  ///                 necessarily valid), this parameter will be updated to

  ///                 point to the character after the UCN.

  /// \param SlashLoc The position in the source buffer of the '\'.

  /// \param Result   The token being formed. Pass \c nullptr to suppress

  ///                 diagnostics and handle token formation in the caller.

  ///

  /// \return The Unicode codepoint specified by the UCN, or 0 if the UCN is

  ///         invalid.

  uint32_t tryReadUCN(const char *&StartPtr, const char *SlashLoc, Token *Result);

  /// Try to consume a UCN as part of an identifier at the current

  /// location.

  /// \param CurPtr Initially points to the range of characters in the source

  ///               buffer containing the '\'. Updated to point past the end of

  ///               the UCN on success.

  /// \param Size The number of characters occupied by the '\' (including

  ///             trigraphs and escaped newlines).

  /// \param Result The token being produced. Marked as containing a UCN on

  ///               success.

  /// \return \c true if a UCN was lexed and it produced an acceptable

  ///         identifier character, \c false otherwise.

  bool tryConsumeIdentifierUCN(const char *&CurPtr, unsigned Size,

                               Token &Result);

  /// Try to consume an identifier character encoded in UTF-8.

  /// \param CurPtr Points to the start of the (potential) UTF-8 code unit

  ///        sequence. On success, updated to point past the end of it.

  /// \return \c true if a UTF-8 sequence mapping to an acceptable identifier

  ///         character was lexed, \c false otherwise.

  bool tryConsumeIdentifierUTF8Char(const char *&CurPtr);

};

} // namespace clang

#endif // LLVM_CLANG_LEX_LEXER_H