1 //===--- TokenAnalyzer.h - Analyze Token Streams ----------------*- C++ -*-===//
3 // The LLVM Compiler Infrastructure
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
8 //===----------------------------------------------------------------------===//
11 /// This file declares an abstract TokenAnalyzer, and associated helper
12 /// classes. TokenAnalyzer can be extended to generate replacements based on
13 /// an annotated and pre-processed token stream.
15 //===----------------------------------------------------------------------===//
17 #ifndef LLVM_CLANG_LIB_FORMAT_TOKENANALYZER_H
18 #define LLVM_CLANG_LIB_FORMAT_TOKENANALYZER_H
20 #include "AffectedRangeManager.h"
22 #include "FormatToken.h"
23 #include "FormatTokenLexer.h"
24 #include "TokenAnnotator.h"
25 #include "UnwrappedLineParser.h"
26 #include "clang/Basic/Diagnostic.h"
27 #include "clang/Basic/DiagnosticOptions.h"
28 #include "clang/Basic/FileManager.h"
29 #include "clang/Basic/SourceManager.h"
30 #include "clang/Format/Format.h"
31 #include "llvm/ADT/STLExtras.h"
32 #include "llvm/Support/Debug.h"
39 Environment(SourceManager &SM, FileID ID, ArrayRef<CharSourceRange> Ranges)
40 : SM(SM), ID(ID), CharRanges(Ranges.begin(), Ranges.end()),
41 FirstStartColumn(0), NextStartColumn(0), LastStartColumn(0) {}
43 // This sets up an virtual file system with file \p FileName containing the
44 // fragment \p Code. Assumes that \p Code starts at \p FirstStartColumn,
45 // that the next lines of \p Code should start at \p NextStartColumn, and
46 // that \p Code should end at \p LastStartColumn if it ends in newline.
47 // See also the documentation of clang::format::internal::reformat.
48 Environment(StringRef Code, StringRef FileName,
49 ArrayRef<tooling::Range> Ranges, unsigned FirstStartColumn = 0,
50 unsigned NextStartColumn = 0, unsigned LastStartColumn = 0);
52 FileID getFileID() const { return ID; }
54 const SourceManager &getSourceManager() const { return SM; }
56 ArrayRef<CharSourceRange> getCharRanges() const { return CharRanges; }
58 // Returns the column at which the fragment of code managed by this
59 // environment starts.
60 unsigned getFirstStartColumn() const { return FirstStartColumn; }
62 // Returns the column at which subsequent lines of the fragment of code
63 // managed by this environment should start.
64 unsigned getNextStartColumn() const { return NextStartColumn; }
66 // Returns the column at which the fragment of code managed by this
67 // environment should end if it ends in a newline.
68 unsigned getLastStartColumn() const { return LastStartColumn; }
71 // This is only set if constructed from string.
72 std::unique_ptr<SourceManagerForFile> VirtualSM;
74 // This refers to either a SourceManager provided by users or VirtualSM
75 // created for a single file.
79 SmallVector<CharSourceRange, 8> CharRanges;
80 unsigned FirstStartColumn;
81 unsigned NextStartColumn;
82 unsigned LastStartColumn;
85 class TokenAnalyzer : public UnwrappedLineConsumer {
87 TokenAnalyzer(const Environment &Env, const FormatStyle &Style);
89 std::pair<tooling::Replacements, unsigned> process();
92 virtual std::pair<tooling::Replacements, unsigned>
93 analyze(TokenAnnotator &Annotator,
94 SmallVectorImpl<AnnotatedLine *> &AnnotatedLines,
95 FormatTokenLexer &Tokens) = 0;
97 void consumeUnwrappedLine(const UnwrappedLine &TheLine) override;
99 void finishRun() override;
102 // Stores Style, FileID and SourceManager etc.
103 const Environment &Env;
104 // AffectedRangeMgr stores ranges to be fixed.
105 AffectedRangeManager AffectedRangeMgr;
106 SmallVector<SmallVector<UnwrappedLine, 16>, 2> UnwrappedLines;
107 encoding::Encoding Encoding;
110 } // end namespace format
111 } // end namespace clang