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/SmallPtrSet.h"
  26 #include "llvm/ADT/StringSet.h"
  27 #include "llvm/Support/Regex.h"
  28
  29 #include <stack>
  30
  31 namespace clang {
  32 namespace format {
  33
  34 enum LexerState {
  35   NORMAL,
  36   TEMPLATE_STRING,
  37   TOKEN_STASHED,
  38 };
  39
  40 class FormatTokenLexer {
  41 public:
  42   FormatTokenLexer(const SourceManager &SourceMgr, FileID ID, unsigned Column,
  43                    const FormatStyle &Style, encoding::Encoding Encoding,
  44                    llvm::SpecificBumpPtrAllocator<FormatToken> &Allocator,
  45                    IdentifierTable &IdentTable);
  46
  47   ArrayRef<FormatToken *> lex();
  48
  49   const AdditionalKeywords &getKeywords() { return Keywords; }
  50
  51 private:
  52   void tryMergePreviousTokens();
  53
  54   bool tryMergeLessLess();
  55   bool tryMergeGreaterGreater();
  56   bool tryMergeNSStringLiteral();
  57   bool tryMergeJSPrivateIdentifier();
  58   bool tryMergeCSharpStringLiteral();
  59   bool tryMergeCSharpKeywordVariables();
  60   bool tryMergeNullishCoalescingEqual();
  61   bool tryTransformCSharpForEach();
  62   bool tryMergeForEach();
  63   bool tryTransformTryUsageForC();
  64
  65   // Merge the most recently lexed tokens into a single token if their kinds are
  66   // correct.
  67   bool tryMergeTokens(ArrayRef<tok::TokenKind> Kinds, TokenType NewType);
  68   // Merge without checking their kinds.
  69   bool tryMergeTokens(size_t Count, TokenType NewType);
  70   // Merge if their kinds match any one of Kinds.
  71   bool tryMergeTokensAny(ArrayRef<ArrayRef<tok::TokenKind>> Kinds,
  72                          TokenType NewType);
  73
  74   // Returns \c true if \p Tok can only be followed by an operand in JavaScript.
  75   bool precedesOperand(FormatToken *Tok);
  76
  77   bool canPrecedeRegexLiteral(FormatToken *Prev);
  78
  79   // Tries to parse a JavaScript Regex literal starting at the current token,
  80   // if that begins with a slash and is in a location where JavaScript allows
  81   // regex literals. Changes the current token to a regex literal and updates
  82   // its text if successful.
  83   void tryParseJSRegexLiteral();
  84
  85   // Handles JavaScript template strings.
  86   //
  87   // JavaScript template strings use backticks ('`') as delimiters, and allow
  88   // embedding expressions nested in ${expr-here}. Template strings can be
  89   // nested recursively, i.e. expressions can contain template strings in turn.
  90   //
  91   // The code below parses starting from a backtick, up to a closing backtick or
  92   // an opening ${. It also maintains a stack of lexing contexts to handle
  93   // nested template parts by balancing curly braces.
  94   void handleTemplateStrings();
  95
  96   void handleCSharpVerbatimAndInterpolatedStrings();
  97
  98   void tryParsePythonComment();
  99
 100   bool tryMerge_TMacro();
 101
 102   bool tryMergeConflictMarkers();
 103
 104   void truncateToken(size_t NewLen);
 105
 106   FormatToken *getStashedToken();
 107
 108   FormatToken *getNextToken();
 109
 110   FormatToken *FormatTok;
 111   bool IsFirstToken;
 112   std::stack<LexerState> StateStack;
 113   unsigned Column;
 114   unsigned TrailingWhitespace;
 115   std::unique_ptr<Lexer> Lex;
 116   LangOptions LangOpts;
 117   const SourceManager &SourceMgr;
 118   FileID ID;
 119   const FormatStyle &Style;
 120   IdentifierTable &IdentTable;
 121   AdditionalKeywords Keywords;
 122   encoding::Encoding Encoding;
 123   llvm::SpecificBumpPtrAllocator<FormatToken> &Allocator;
 124   // Index (in 'Tokens') of the last token that starts a new line.
 125   unsigned FirstInLineIndex;
 126   SmallVector<FormatToken *, 16> Tokens;
 127
 128   llvm::SmallMapVector<IdentifierInfo *, TokenType, 8> Macros;
 129
 130   llvm::SmallPtrSet<IdentifierInfo *, 8> TypeNames;
 131
 132   bool FormattingDisabled;
 133
 134   llvm::Regex MacroBlockBeginRegex;
 135   llvm::Regex MacroBlockEndRegex;
 136
 137   // Targets that may appear inside a C# attribute.
 138   static const llvm::StringSet<> CSharpAttributeTargets;
 139
 140   /// Handle Verilog-specific tokens.
 141   bool readRawTokenVerilogSpecific(Token &Tok);
 142
 143   void readRawToken(FormatToken &Tok);
 144
 145   void resetLexer(unsigned Offset);
 146 };
 147
 148 } // namespace format
 149 } // namespace clang
 150
 151 #endif