Coverage Report

Created: 2022-05-17 06:19

/Users/buildslave/jenkins/workspace/coverage/llvm-project/clang/lib/Format/WhitespaceManager.h
Line
Count
Source (jump to first uncovered line)
1
//===--- WhitespaceManager.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
/// WhitespaceManager class manages whitespace around tokens and their
11
/// replacements.
12
///
13
//===----------------------------------------------------------------------===//
14
15
#ifndef LLVM_CLANG_LIB_FORMAT_WHITESPACEMANAGER_H
16
#define LLVM_CLANG_LIB_FORMAT_WHITESPACEMANAGER_H
17
18
#include "TokenAnnotator.h"
19
#include "clang/Basic/SourceManager.h"
20
#include "clang/Format/Format.h"
21
#include "llvm/ADT/SmallVector.h"
22
#include <algorithm>
23
#include <string>
24
#include <tuple>
25
26
namespace clang {
27
namespace format {
28
29
/// Manages the whitespaces around tokens and their replacements.
30
///
31
/// This includes special handling for certain constructs, e.g. the alignment of
32
/// trailing line comments.
33
///
34
/// To guarantee correctness of alignment operations, the \c WhitespaceManager
35
/// must be informed about every token in the source file; for each token, there
36
/// must be exactly one call to either \c replaceWhitespace or
37
/// \c addUntouchableToken.
38
///
39
/// There may be multiple calls to \c breakToken for a given token.
40
class WhitespaceManager {
41
public:
42
  WhitespaceManager(const SourceManager &SourceMgr, const FormatStyle &Style,
43
                    bool UseCRLF)
44
22.2k
      : SourceMgr(SourceMgr), Style(Style), UseCRLF(UseCRLF) {}
45
46
2.69k
  bool useCRLF() const { return UseCRLF; }
47
48
  /// Infers whether the input is using CRLF.
49
  static bool inputUsesCRLF(StringRef Text, bool DefaultToCRLF);
50
51
  /// Replaces the whitespace in front of \p Tok. Only call once for
52
  /// each \c AnnotatedToken.
53
  ///
54
  /// \p StartOfTokenColumn is the column at which the token will start after
55
  /// this replacement. It is needed for determining how \p Spaces is turned
56
  /// into tabs and spaces for some format styles.
57
  void replaceWhitespace(FormatToken &Tok, unsigned Newlines, unsigned Spaces,
58
                         unsigned StartOfTokenColumn, bool isAligned = false,
59
                         bool InPPDirective = false);
60
61
  /// Adds information about an unchangeable token's whitespace.
62
  ///
63
  /// Needs to be called for every token for which \c replaceWhitespace
64
  /// was not called.
65
  void addUntouchableToken(const FormatToken &Tok, bool InPPDirective);
66
67
  llvm::Error addReplacement(const tooling::Replacement &Replacement);
68
69
  /// Inserts or replaces whitespace in the middle of a token.
70
  ///
71
  /// Inserts \p PreviousPostfix, \p Newlines, \p Spaces and \p CurrentPrefix
72
  /// (in this order) at \p Offset inside \p Tok, replacing \p ReplaceChars
73
  /// characters.
74
  ///
75
  /// Note: \p Spaces can be negative to retain information about initial
76
  /// relative column offset between a line of a block comment and the start of
77
  /// the comment. This negative offset may be compensated by trailing comment
78
  /// alignment here. In all other cases negative \p Spaces will be truncated to
79
  /// 0.
80
  ///
81
  /// When \p InPPDirective is true, escaped newlines are inserted. \p Spaces is
82
  /// used to align backslashes correctly.
83
  void replaceWhitespaceInToken(const FormatToken &Tok, unsigned Offset,
84
                                unsigned ReplaceChars,
85
                                StringRef PreviousPostfix,
86
                                StringRef CurrentPrefix, bool InPPDirective,
87
                                unsigned Newlines, int Spaces);
88
89
  /// Returns all the \c Replacements created during formatting.
90
  const tooling::Replacements &generateReplacements();
91
92
  /// Represents a change before a token, a break inside a token,
93
  /// or the layout of an unchanged token (or whitespace within).
94
  struct Change {
95
    /// Functor to sort changes in original source order.
96
    class IsBeforeInFile {
97
    public:
98
22.2k
      IsBeforeInFile(const SourceManager &SourceMgr) : SourceMgr(SourceMgr) {}
99
      bool operator()(const Change &C1, const Change &C2) const;
100
101
    private:
102
      const SourceManager &SourceMgr;
103
    };
104
105
    /// Creates a \c Change.
106
    ///
107
    /// The generated \c Change will replace the characters at
108
    /// \p OriginalWhitespaceRange with a concatenation of
109
    /// \p PreviousLinePostfix, \p NewlinesBefore line breaks, \p Spaces spaces
110
    /// and \p CurrentLinePrefix.
111
    ///
112
    /// \p StartOfTokenColumn and \p InPPDirective will be used to lay out
113
    /// trailing comments and escaped newlines.
114
    Change(const FormatToken &Tok, bool CreateReplacement,
115
           SourceRange OriginalWhitespaceRange, int Spaces,
116
           unsigned StartOfTokenColumn, unsigned NewlinesBefore,
117
           StringRef PreviousLinePostfix, StringRef CurrentLinePrefix,
118
           bool IsAligned, bool ContinuesPPDirective, bool IsInsideToken);
119
120
    // The kind of the token whose whitespace this change replaces, or in which
121
    // this change inserts whitespace.
122
    // FIXME: Currently this is not set correctly for breaks inside comments, as
123
    // the \c BreakableToken is still doing its own alignment.
124
    const FormatToken *Tok;
125
126
    bool CreateReplacement;
127
    // Changes might be in the middle of a token, so we cannot just keep the
128
    // FormatToken around to query its information.
129
    SourceRange OriginalWhitespaceRange;
130
    unsigned StartOfTokenColumn;
131
    unsigned NewlinesBefore;
132
    std::string PreviousLinePostfix;
133
    std::string CurrentLinePrefix;
134
    bool IsAligned;
135
    bool ContinuesPPDirective;
136
137
    // The number of spaces in front of the token or broken part of the token.
138
    // This will be adapted when aligning tokens.
139
    // Can be negative to retain information about the initial relative offset
140
    // of the lines in a block comment. This is used when aligning trailing
141
    // comments. Uncompensated negative offset is truncated to 0.
142
    int Spaces;
143
144
    // If this change is inside of a token but not at the start of the token or
145
    // directly after a newline.
146
    bool IsInsideToken;
147
148
    // \c IsTrailingComment, \c TokenLength, \c PreviousEndOfTokenColumn and
149
    // \c EscapedNewlineColumn will be calculated in
150
    // \c calculateLineBreakInformation.
151
    bool IsTrailingComment;
152
    unsigned TokenLength;
153
    unsigned PreviousEndOfTokenColumn;
154
    unsigned EscapedNewlineColumn;
155
156
    // These fields are used to retain correct relative line indentation in a
157
    // block comment when aligning trailing comments.
158
    //
159
    // If this Change represents a continuation of a block comment,
160
    // \c StartOfBlockComment is pointer to the first Change in the block
161
    // comment. \c IndentationOffset is a relative column offset to this
162
    // change, so that the correct column can be reconstructed at the end of
163
    // the alignment process.
164
    const Change *StartOfBlockComment;
165
    int IndentationOffset;
166
167
    // Depth of conditionals. Computed from tracking fake parenthesis, except
168
    // it does not increase the indent for "chained" conditionals.
169
    int ConditionalsLevel;
170
171
    // A combination of indent, nesting and conditionals levels, which are used
172
    // in tandem to compute lexical scope, for the purposes of deciding
173
    // when to stop consecutive alignment runs.
174
1.12M
    std::tuple<unsigned, unsigned, unsigned> indentAndNestingLevel() const {
175
1.12M
      return std::make_tuple(Tok->IndentLevel, Tok->NestingLevel,
176
1.12M
                             ConditionalsLevel);
177
1.12M
    }
178
  };
179
180
private:
181
  struct CellDescription {
182
    unsigned Index = 0;
183
    unsigned Cell = 0;
184
    unsigned EndIndex = 0;
185
    bool HasSplit = false;
186
    CellDescription *NextColumnElement = nullptr;
187
188
0
    constexpr bool operator==(const CellDescription &Other) const {
189
0
      return Index == Other.Index && Cell == Other.Cell &&
190
0
             EndIndex == Other.EndIndex;
191
0
    }
192
0
    constexpr bool operator!=(const CellDescription &Other) const {
193
0
      return !(*this == Other);
194
0
    }
195
  };
196
197
  struct CellDescriptions {
198
    SmallVector<CellDescription> Cells;
199
    SmallVector<unsigned> CellCounts;
200
    unsigned InitialSpaces = 0;
201
202
    // Determine if every row in the the array
203
    // has the same number of columns.
204
157
    bool isRectangular() const {
205
157
      if (CellCounts.empty())
206
0
        return false;
207
208
157
      for (auto NumberOfColumns : CellCounts)
209
414
        if (NumberOfColumns != CellCounts[0])
210
36
          return false;
211
121
      return true;
212
157
    }
213
  };
214
215
  /// Calculate \c IsTrailingComment, \c TokenLength for the last tokens
216
  /// or token parts in a line and \c PreviousEndOfTokenColumn and
217
  /// \c EscapedNewlineColumn for the first tokens or token parts in a line.
218
  void calculateLineBreakInformation();
219
220
  /// \brief Align consecutive C/C++ preprocessor macros over all \c Changes.
221
  void alignConsecutiveMacros();
222
223
  /// Align consecutive assignments over all \c Changes.
224
  void alignConsecutiveAssignments();
225
226
  /// Align consecutive bitfields over all \c Changes.
227
  void alignConsecutiveBitFields();
228
229
  /// Align consecutive declarations over all \c Changes.
230
  void alignConsecutiveDeclarations();
231
232
  /// Align consecutive declarations over all \c Changes.
233
  void alignChainedConditionals();
234
235
  /// Align trailing comments over all \c Changes.
236
  void alignTrailingComments();
237
238
  /// Align trailing comments from change \p Start to change \p End at
239
  /// the specified \p Column.
240
  void alignTrailingComments(unsigned Start, unsigned End, unsigned Column);
241
242
  /// Align escaped newlines over all \c Changes.
243
  void alignEscapedNewlines();
244
245
  /// Align escaped newlines from change \p Start to change \p End at
246
  /// the specified \p Column.
247
  void alignEscapedNewlines(unsigned Start, unsigned End, unsigned Column);
248
249
  /// Align Array Initializers over all \c Changes.
250
  void alignArrayInitializers();
251
252
  /// Align Array Initializers from change \p Start to change \p End at
253
  /// the specified \p Column.
254
  void alignArrayInitializers(unsigned Start, unsigned End);
255
256
  /// Align Array Initializers being careful to right justify the columns
257
  /// as described by \p CellDescs.
258
  void alignArrayInitializersRightJustified(CellDescriptions &&CellDescs);
259
260
  /// Align Array Initializers being careful to left justify the columns
261
  /// as described by \p CellDescs.
262
  void alignArrayInitializersLeftJustified(CellDescriptions &&CellDescs);
263
264
  /// Calculate the cell width between two indexes.
265
  unsigned calculateCellWidth(unsigned Start, unsigned End,
266
                              bool WithSpaces = false) const;
267
268
  /// Get a set of fully specified CellDescriptions between \p Start and
269
  /// \p End of the change list.
270
  CellDescriptions getCells(unsigned Start, unsigned End);
271
272
  /// Does this \p Cell contain a split element?
273
  static bool isSplitCell(const CellDescription &Cell);
274
275
  /// Get the width of the preceding cells from \p Start to \p End.
276
  template <typename I>
277
1.24k
  auto getNetWidth(const I &Start, const I &End, unsigned InitialSpaces) const {
278
1.24k
    auto NetWidth = InitialSpaces;
279
4.11k
    for (auto PrevIter = Start; PrevIter != End; 
++PrevIter2.87k
) {
280
      // If we broke the line the initial spaces are already
281
      // accounted for.
282
2.87k
      if (Changes[PrevIter->Index].NewlinesBefore > 0)
283
87
        NetWidth = 0;
284
2.87k
      NetWidth +=
285
2.87k
          calculateCellWidth(PrevIter->Index, PrevIter->EndIndex, true) + 1;
286
2.87k
    }
287
1.24k
    return NetWidth;
288
1.24k
  }
289
290
  /// Get the maximum width of a cell in a sequence of columns.
291
  template <typename I>
292
262
  unsigned getMaximumCellWidth(I CellIter, unsigned NetWidth) const {
293
262
    unsigned CellWidth =
294
262
        calculateCellWidth(CellIter->Index, CellIter->EndIndex, true);
295
262
    if (Changes[CellIter->Index].NewlinesBefore == 0)
296
240
      CellWidth += NetWidth;
297
735
    for (const auto *Next = CellIter->NextColumnElement; Next != nullptr;
298
473
         Next = Next->NextColumnElement) {
299
473
      auto ThisWidth = calculateCellWidth(Next->Index, Next->EndIndex, true);
300
473
      if (Changes[Next->Index].NewlinesBefore == 0)
301
471
        ThisWidth += NetWidth;
302
473
      CellWidth = std::max(CellWidth, ThisWidth);
303
473
    }
304
262
    return CellWidth;
305
262
  }
306
307
  /// Get The maximum width of all columns to a given cell.
308
  template <typename I>
309
  unsigned getMaximumNetWidth(const I &CellStart, const I &CellStop,
310
                              unsigned InitialSpaces, unsigned CellCount,
311
217
                              unsigned MaxRowCount) const {
312
217
    auto MaxNetWidth = getNetWidth(CellStart, CellStop, InitialSpaces);
313
217
    auto RowCount = 1U;
314
217
    auto Offset = std::distance(CellStart, CellStop);
315
612
    for (const auto *Next = CellStop->NextColumnElement; Next != nullptr;
316
395
         Next = Next->NextColumnElement) {
317
395
      if (RowCount > MaxRowCount)
318
0
        break;
319
395
      auto Start = (CellStart + RowCount * CellCount);
320
395
      auto End = Start + Offset;
321
395
      MaxNetWidth =
322
395
          std::max(MaxNetWidth, getNetWidth(Start, End, InitialSpaces));
323
395
      ++RowCount;
324
395
    }
325
217
    return MaxNetWidth;
326
217
  }
327
328
  /// Align a split cell with a newline to the first element in the cell.
329
  void alignToStartOfCell(unsigned Start, unsigned End);
330
331
  /// Link the Cell pointers in the list of Cells.
332
  static CellDescriptions linkCells(CellDescriptions &&CellDesc);
333
334
  /// Fill \c Replaces with the replacements for all effective changes.
335
  void generateChanges();
336
337
  /// Stores \p Text as the replacement for the whitespace in \p Range.
338
  void storeReplacement(SourceRange Range, StringRef Text);
339
  void appendNewlineText(std::string &Text, unsigned Newlines);
340
  void appendEscapedNewlineText(std::string &Text, unsigned Newlines,
341
                                unsigned PreviousEndOfTokenColumn,
342
                                unsigned EscapedNewlineColumn);
343
  void appendIndentText(std::string &Text, unsigned IndentLevel,
344
                        unsigned Spaces, unsigned WhitespaceStartColumn,
345
                        bool IsAligned);
346
  unsigned appendTabIndent(std::string &Text, unsigned Spaces,
347
                           unsigned Indentation);
348
349
  SmallVector<Change, 16> Changes;
350
  const SourceManager &SourceMgr;
351
  tooling::Replacements Replaces;
352
  const FormatStyle &Style;
353
  bool UseCRLF;
354
};
355
356
} // namespace format
357
} // namespace clang
358
359
#endif