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