clang 20.0.0git
TokenAnalyzer.cpp
Go to the documentation of this file.
1//===--- TokenAnalyzer.cpp - Analyze Token Streams --------------*- 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 implements an abstract TokenAnalyzer and associated helper
11/// classes. TokenAnalyzer can be extended to generate replacements based on
12/// an annotated and pre-processed token stream.
13///
14//===----------------------------------------------------------------------===//
15
16#include "TokenAnalyzer.h"
18#include "Encoding.h"
19#include "FormatToken.h"
20#include "FormatTokenLexer.h"
21#include "TokenAnnotator.h"
22#include "UnwrappedLineParser.h"
27#include "clang/Format/Format.h"
28#include "llvm/ADT/STLExtras.h"
29#include "llvm/ADT/SmallVector.h"
30#include "llvm/Support/Debug.h"
31#include <type_traits>
32
33#define DEBUG_TYPE "format-formatter"
34
35namespace clang {
36namespace format {
37
38// FIXME: Instead of printing the diagnostic we should store it and have a
39// better way to return errors through the format APIs.
41public:
43 const Diagnostic &Info) override {
44 if (DiagLevel == DiagnosticsEngine::Fatal) {
45 Fatal = true;
47 Info.FormatDiagnostic(Message);
48 llvm::errs() << Message << "\n";
49 }
50 }
51
52 bool fatalError() const { return Fatal; }
53
54private:
55 bool Fatal = false;
56};
57
58std::unique_ptr<Environment>
59Environment::make(StringRef Code, StringRef FileName,
60 ArrayRef<tooling::Range> Ranges, unsigned FirstStartColumn,
61 unsigned NextStartColumn, unsigned LastStartColumn) {
62 auto Env = std::make_unique<Environment>(Code, FileName, FirstStartColumn,
63 NextStartColumn, LastStartColumn);
65 Env->SM.getDiagnostics().setClient(&Diags, /*ShouldOwnClient=*/false);
66 SourceLocation StartOfFile = Env->SM.getLocForStartOfFile(Env->ID);
67 for (const tooling::Range &Range : Ranges) {
68 SourceLocation Start = StartOfFile.getLocWithOffset(Range.getOffset());
69 SourceLocation End = Start.getLocWithOffset(Range.getLength());
70 Env->CharRanges.push_back(CharSourceRange::getCharRange(Start, End));
71 }
72 // Validate that we can get the buffer data without a fatal error.
73 Env->SM.getBufferData(Env->ID);
74 if (Diags.fatalError())
75 return nullptr;
76 return Env;
77}
78
79Environment::Environment(StringRef Code, StringRef FileName,
80 unsigned FirstStartColumn, unsigned NextStartColumn,
81 unsigned LastStartColumn)
82 : VirtualSM(new SourceManagerForFile(FileName, Code)), SM(VirtualSM->get()),
83 ID(VirtualSM->get().getMainFileID()), FirstStartColumn(FirstStartColumn),
84 NextStartColumn(NextStartColumn), LastStartColumn(LastStartColumn) {}
85
87 : Style(Style), LangOpts(getFormattingLangOpts(Style)), Env(Env),
88 AffectedRangeMgr(Env.getSourceManager(), Env.getCharRanges()),
89 UnwrappedLines(1),
90 Encoding(encoding::detectEncoding(
91 Env.getSourceManager().getBufferData(Env.getFileID()))) {
92 LLVM_DEBUG(
93 llvm::dbgs() << "File encoding: "
94 << (Encoding == encoding::Encoding_UTF8 ? "UTF8" : "unknown")
95 << "\n");
96 LLVM_DEBUG(llvm::dbgs() << "Language: " << getLanguageName(Style.Language)
97 << "\n");
98}
99
100std::pair<tooling::Replacements, unsigned>
101TokenAnalyzer::process(bool SkipAnnotation) {
103 llvm::SpecificBumpPtrAllocator<FormatToken> Allocator;
104 IdentifierTable IdentTable(LangOpts);
106 Env.getFirstStartColumn(), Style, Encoding, Allocator,
107 IdentTable);
108 ArrayRef<FormatToken *> Toks(Lex.lex());
111 Env.getFirstStartColumn(), Tokens, *this,
112 Allocator, IdentTable);
113 Parser.parse();
114 assert(UnwrappedLines.back().empty());
115 unsigned Penalty = 0;
116 for (unsigned Run = 0, RunE = UnwrappedLines.size(); Run + 1 != RunE; ++Run) {
117 const auto &Lines = UnwrappedLines[Run];
118 LLVM_DEBUG(llvm::dbgs() << "Run " << Run << "...\n");
120 AnnotatedLines.reserve(Lines.size());
121
122 TokenAnnotator Annotator(Style, Lex.getKeywords());
123 for (const UnwrappedLine &Line : Lines) {
124 AnnotatedLines.push_back(new AnnotatedLine(Line));
125 if (!SkipAnnotation)
126 Annotator.annotate(*AnnotatedLines.back());
127 }
128
129 std::pair<tooling::Replacements, unsigned> RunResult =
130 analyze(Annotator, AnnotatedLines, Lex);
131
132 LLVM_DEBUG({
133 llvm::dbgs() << "Replacements for run " << Run << ":\n";
134 for (const tooling::Replacement &Fix : RunResult.first)
135 llvm::dbgs() << Fix.toString() << "\n";
136 });
137 for (AnnotatedLine *Line : AnnotatedLines)
138 delete Line;
139
140 Penalty += RunResult.second;
141 for (const auto &R : RunResult.first) {
142 auto Err = Result.add(R);
143 // FIXME: better error handling here. For now, simply return an empty
144 // Replacements to indicate failure.
145 if (Err) {
146 llvm::errs() << llvm::toString(std::move(Err)) << "\n";
147 return {tooling::Replacements(), 0};
148 }
149 }
150 }
151 return {Result, Penalty};
152}
153
155 assert(!UnwrappedLines.empty());
156 UnwrappedLines.back().push_back(TheLine);
157}
158
161}
162
163} // end namespace format
164} // end namespace clang
AffectedRangeManager class manages affected ranges in the code.
static char ID
Definition: Arena.cpp:183
#define SM(sm)
Definition: Cuda.cpp:83
Defines the Diagnostic-related interfaces.
Contains functions for text encoding manipulation.
Defines the clang::FileManager interface and associated types.
This file contains FormatTokenLexer, which tokenizes a source file into a token stream suitable for C...
This file contains the declaration of the FormatToken, a wrapper around Token with additional informa...
Various functions to configurably format source code.
const Environment & Env
Definition: HTMLLogger.cpp:148
SourceRange Range
Definition: SemaObjC.cpp:758
Defines the SourceManager interface.
This file declares an abstract TokenAnalyzer, and associated helper classes.
This file implements a token annotator, i.e.
This file contains the declaration of the UnwrappedLineParser, which turns a stream of tokens into Un...
static CharSourceRange getCharRange(SourceRange R)
Abstract interface, implemented by clients of the front-end, which formats and prints fully processed...
Definition: Diagnostic.h:1745
A little helper class (which is basically a smart pointer that forwards info from DiagnosticsEngine) ...
Definition: Diagnostic.h:1571
void FormatDiagnostic(SmallVectorImpl< char > &OutStr) const
Format this diagnostic into a string, substituting the formal arguments into the %0 slots.
Definition: Diagnostic.cpp:792
Level
The level of the diagnostic, after it has been through mapping.
Definition: Diagnostic.h:195
Implements an efficient mapping from strings to IdentifierInfo nodes.
Parser - This implements a parser for the C family of languages.
Definition: Parser.h:58
Encodes a location in the source.
SourceLocation getLocWithOffset(IntTy Offset) const
Return a source location with the specified offset from this SourceLocation.
SourceManager and necessary dependencies (e.g.
SourceManager & getSourceManager() const
Definition: TokenAnalyzer.h:38
Environment(StringRef Code, StringRef FileName, unsigned FirstStartColumn=0, unsigned NextStartColumn=0, unsigned LastStartColumn=0)
static std::unique_ptr< Environment > make(StringRef Code, StringRef FileName, ArrayRef< tooling::Range > Ranges, unsigned FirstStartColumn=0, unsigned NextStartColumn=0, unsigned LastStartColumn=0)
unsigned getFirstStartColumn() const
Definition: TokenAnalyzer.h:44
void HandleDiagnostic(DiagnosticsEngine::Level DiagLevel, const Diagnostic &Info) override
Handle this diagnostic, reporting it to the user or capturing it to a log as needed.
const AdditionalKeywords & getKeywords()
ArrayRef< FormatToken * > lex()
encoding::Encoding Encoding
virtual std::pair< tooling::Replacements, unsigned > analyze(TokenAnnotator &Annotator, SmallVectorImpl< AnnotatedLine * > &AnnotatedLines, FormatTokenLexer &Tokens)=0
const Environment & Env
Definition: TokenAnalyzer.h:97
SmallVector< SmallVector< UnwrappedLine, 16 >, 2 > UnwrappedLines
TokenAnalyzer(const Environment &Env, const FormatStyle &Style)
std::pair< tooling::Replacements, unsigned > process(bool SkipAnnotation=false)
void consumeUnwrappedLine(const UnwrappedLine &TheLine) override
Determines extra information about the tokens comprising an UnwrappedLine.
void annotate(AnnotatedLine &Line)
A source range independent of the SourceManager.
Definition: Replacement.h:44
A text replacement.
Definition: Replacement.h:83
std::string toString() const
Returns a human readable string representation.
Definition: Replacement.cpp:87
Maintains a set of replacements that are conflict-free.
Definition: Replacement.h:212
LangOptions getFormattingLangOpts(const FormatStyle &Style=getLLVMStyle())
Returns the LangOpts that the formatter expects you to set.
Definition: Format.cpp:3873
StringRef getLanguageName(FormatStyle::LanguageKind Language)
Definition: Format.h:5504
The JSON file list parser is used to communicate input to InstallAPI.
@ Result
The result type of a method or function.
The FormatStyle is used to configure the formatting to follow specific guidelines.
Definition: Format.h:55
LanguageKind Language
Language, this format style is targeted at.
Definition: Format.h:3257
An unwrapped line is a sequence of Token, that we would like to put on a single line if there was no ...