clang/lib/Format/FormatTokenLexer.h

   1 //===--- FormatTokenLexer.h - Format C++ code ----------------*- C++ ----*-===//
   2 //
   3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
   4 // See https://llvm.org/LICENSE.txt for license information.
   5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
   6 //
   7 //===----------------------------------------------------------------------===//
   8 ///
   9 /// \file
  10 /// This file contains FormatTokenLexer, which tokenizes a source file
  11 /// into a token stream suitable for ClangFormat.
  12 ///
  13 //===----------------------------------------------------------------------===//
  14
  15 #ifndef LLVM_CLANG_LIB_FORMAT_FORMATTOKENLEXER_H
  16 #define LLVM_CLANG_LIB_FORMAT_FORMATTOKENLEXER_H
  17
  18 #include "Encoding.h"
  19 #include "FormatToken.h"
  20 #include "clang/Basic/LangOptions.h"
  21 #include "clang/Basic/SourceLocation.h"
  22 #include "clang/Basic/SourceManager.h"
  23 #include "clang/Format/Format.h"
  24 #include "llvm/ADT/MapVector.h"
  25 #include "llvm/ADT/StringSet.h"
  26 #include "llvm/Support/Regex.h"
  27
  28 #include <stack>
  29
  30 namespace clang {
  31 namespace format {
  32
  33 enum LexerState {
  34   NORMAL,
  35   TEMPLATE_STRING,
  36   TOKEN_STASHED,
  37 };
  38
  39 class FormatTokenLexer {
  40 public:
  41   FormatTokenLexer(const SourceManager &SourceMgr, FileID ID, unsigned Column,
  42                    const FormatStyle &Style, encoding::Encoding Encoding,
  43                    llvm::SpecificBumpPtrAllocator<FormatToken> &Allocator,
  44                    IdentifierTable &IdentTable);
  45
  46   ArrayRef<FormatToken *> lex();
  47
  48   const AdditionalKeywords &getKeywords() { return Keywords; }
  49
  50 private:
  51   void tryMergePreviousTokens();
  52
  53   bool tryMergeLessLess();
  54   bool tryMergeGreaterGreater();
  55   bool tryMergeNSStringLiteral();
  56   bool tryMergeJSPrivateIdentifier();
  57   bool tryMergeCSharpStringLiteral();
  58   bool tryMergeCSharpKeywordVariables();
  59   bool tryMergeNullishCoalescingEqual();
  60   bool tryTransformCSharpForEach();
  61   bool tryMergeForEach();
  62   bool tryTransformTryUsageForC();
  63
  64   // Merge the most recently lexed tokens into a single token if their kinds are
  65   // correct.
  66   bool tryMergeTokens(ArrayRef<tok::TokenKind> Kinds, TokenType NewType);
  67   // Merge without checking their kinds.
  68   bool tryMergeTokens(size_t Count, TokenType NewType);
  69   // Merge if their kinds match any one of Kinds.
  70   bool tryMergeTokensAny(ArrayRef<ArrayRef<tok::TokenKind>> Kinds,
  71                          TokenType NewType);
  72
  73   // Returns \c true if \p Tok can only be followed by an operand in JavaScript.
  74   bool precedesOperand(FormatToken *Tok);
  75
  76   bool canPrecedeRegexLiteral(FormatToken *Prev);
  77
  78   // Tries to parse a JavaScript Regex literal starting at the current token,
  79   // if that begins with a slash and is in a location where JavaScript allows
  80   // regex literals. Changes the current token to a regex literal and updates
  81   // its text if successful.
  82   void tryParseJSRegexLiteral();
  83
  84   // Handles JavaScript template strings.
  85   //
  86   // JavaScript template strings use backticks ('`') as delimiters, and allow
  87   // embedding expressions nested in ${expr-here}. Template strings can be
  88   // nested recursively, i.e. expressions can contain template strings in turn.
  89   //
  90   // The code below parses starting from a backtick, up to a closing backtick or
  91   // an opening ${. It also maintains a stack of lexing contexts to handle
  92   // nested template parts by balancing curly braces.
  93   void handleTemplateStrings();
  94
  95   void handleCSharpVerbatimAndInterpolatedStrings();
  96
  97   void tryParsePythonComment();
  98
  99   bool tryMerge_TMacro();
 100
 101   bool tryMergeConflictMarkers();
 102
 103   void truncateToken(size_t NewLen);
 104
 105   FormatToken *getStashedToken();
 106
 107   FormatToken *getNextToken();
 108
 109   FormatToken *FormatTok;
 110   bool IsFirstToken;
 111   std::stack<LexerState> StateStack;
 112   unsigned Column;
 113   unsigned TrailingWhitespace;
 114   std::unique_ptr<Lexer> Lex;
 115   LangOptions LangOpts;
 116   const SourceManager &SourceMgr;
 117   FileID ID;
 118   const FormatStyle &Style;
 119   IdentifierTable &IdentTable;
 120   AdditionalKeywords Keywords;
 121   encoding::Encoding Encoding;
 122   llvm::SpecificBumpPtrAllocator<FormatToken> &Allocator;
 123   // Index (in 'Tokens') of the last token that starts a new line.
 124   unsigned FirstInLineIndex;
 125   SmallVector<FormatToken *, 16> Tokens;
 126
 127   llvm::SmallMapVector<IdentifierInfo *, TokenType, 8> Macros;
 128
 129   bool FormattingDisabled;
 130
 131   llvm::Regex MacroBlockBeginRegex;
 132   llvm::Regex MacroBlockEndRegex;
 133
 134   // Targets that may appear inside a C# attribute.
 135   static const llvm::StringSet<> CSharpAttributeTargets;
 136
 137   /// Handle Verilog-specific tokens.
 138   bool readRawTokenVerilogSpecific(Token &Tok);
 139
 140   void readRawToken(FormatToken &Tok);
 141
 142   void resetLexer(unsigned Offset);
 143 };
 144
 145 } // namespace format
 146 } // namespace clang
 147
 148 #endif