CommentLexer.h

Go to the documentation of this file.

//===--- CommentLexer.h - Lexer for structured comments ---------*- 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 lexer for structured comments and supporting token class.
//
//===----------------------------------------------------------------------===//
 
#ifndef LLVM_CLANG_AST_COMMENTLEXER_H
#define LLVM_CLANG_AST_COMMENTLEXER_H
 
#include "clang/Basic/Diagnostic.h"
#include "clang/Basic/SourceManager.h"
#include "llvm/ADT/SmallString.h"
#include "llvm/ADT/StringRef.h"
#include "llvm/Support/Allocator.h"
#include "llvm/Support/raw_ostream.h"
 
namespace clang {
namespace comments {
 
class Lexer;
class TextTokenRetokenizer;
struct CommandInfo;
class CommandTraits;
 
namespace tok {
enum TokenKind {
  eof,
  newline,
  text,
  unknown_command,   // Command that does not have an ID.
  backslash_command, // Command with an ID, that used backslash marker.
  at_command,        // Command with an ID, that used 'at' marker.
  verbatim_block_begin,
  verbatim_block_line,
  verbatim_block_end,
  verbatim_line_name,
  verbatim_line_text,
  html_start_tag,     // <tag
  html_ident,         // attr
  html_equals,        // =
  html_quoted_string, // "blah\"blah" or 'blah\'blah'
  html_greater,       // >
  html_slash_greater, // />
  html_end_tag        // </tag
};
} // end namespace tok
 
/// Comment token.
class Token {
  friend class Lexer;
  friend class TextTokenRetokenizer;
 
  /// The location of the token.
  SourceLocation Loc;
 
  /// The actual kind of the token.
  tok::TokenKind Kind;
 
  /// Integer value associated with a token.
  ///
  /// If the token is a known command, contains command ID and TextPtr is
  /// unused (command spelling can be found with CommandTraits).  Otherwise,
  /// contains the length of the string that starts at TextPtr.
  unsigned IntVal;
 
  /// Length of the token spelling in comment.  Can be 0 for synthenized
  /// tokens.
  unsigned Length;
 
  /// Contains text value associated with a token.
  const char *TextPtr;
 
public:
  SourceLocation getLocation() const LLVM_READONLY { return Loc; }
  void setLocation(SourceLocation SL) { Loc = SL; }
 
  SourceLocation getEndLocation() const LLVM_READONLY {
    if (Length == 0 || Length == 1)
      return Loc;
    return Loc.getLocWithOffset(Length - 1);
  }
 
  tok::TokenKind getKind() const LLVM_READONLY { return Kind; }
  void setKind(tok::TokenKind K) { Kind = K; }
 
  bool is(tok::TokenKind K) const LLVM_READONLY { return Kind == K; }
  bool isNot(tok::TokenKind K) const LLVM_READONLY { return Kind != K; }
 
  unsigned getLength() const LLVM_READONLY { return Length; }
  void setLength(unsigned L) { Length = L; }
 
  StringRef getText() const LLVM_READONLY {
    assert(is(tok::text));
    return StringRef(TextPtr, IntVal);
  }
 
  void setText(StringRef Text) {
    assert(is(tok::text));
    TextPtr = Text.data();
    IntVal = Text.size();
  }
 
  StringRef getUnknownCommandName() const LLVM_READONLY {
    assert(is(tok::unknown_command));
    return StringRef(TextPtr, IntVal);
  }
 
  void setUnknownCommandName(StringRef Name) {
    assert(is(tok::unknown_command));
    TextPtr = Name.data();
    IntVal = Name.size();
  }
 
  unsigned getCommandID() const LLVM_READONLY {
    assert(is(tok::backslash_command) || is(tok::at_command));
    return IntVal;
  }
 
  void setCommandID(unsigned ID) {
    assert(is(tok::backslash_command) || is(tok::at_command));
    IntVal = ID;
  }
 
  unsigned getVerbatimBlockID() const LLVM_READONLY {
    assert(is(tok::verbatim_block_begin) || is(tok::verbatim_block_end));
    return IntVal;
  }
 
  void setVerbatimBlockID(unsigned ID) {
    assert(is(tok::verbatim_block_begin) || is(tok::verbatim_block_end));
    IntVal = ID;
  }
 
  StringRef getVerbatimBlockText() const LLVM_READONLY {
    assert(is(tok::verbatim_block_line));
    return StringRef(TextPtr, IntVal);
  }
 
  void setVerbatimBlockText(StringRef Text) {
    assert(is(tok::verbatim_block_line));
    TextPtr = Text.data();
    IntVal = Text.size();
  }
 
  unsigned getVerbatimLineID() const LLVM_READONLY {
    assert(is(tok::verbatim_line_name));
    return IntVal;
  }
 
  void setVerbatimLineID(unsigned ID) {
    assert(is(tok::verbatim_line_name));
    IntVal = ID;
  }
 
  StringRef getVerbatimLineText() const LLVM_READONLY {
    assert(is(tok::verbatim_line_text));
    return StringRef(TextPtr, IntVal);
  }
 
  void setVerbatimLineText(StringRef Text) {
    assert(is(tok::verbatim_line_text));
    TextPtr = Text.data();
    IntVal = Text.size();
  }
 
  StringRef getHTMLTagStartName() const LLVM_READONLY {
    assert(is(tok::html_start_tag));
    return StringRef(TextPtr, IntVal);
  }
 
  void setHTMLTagStartName(StringRef Name) {
    assert(is(tok::html_start_tag));
    TextPtr = Name.data();
    IntVal = Name.size();
  }
 
  StringRef getHTMLIdent() const LLVM_READONLY {
    assert(is(