clang  10.0.0svn
Rewriter.cpp
Go to the documentation of this file.
1 //===- Rewriter.cpp - Code rewriting interface ----------------------------===//
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 // This file defines the Rewriter class, which is used for code
10 // transformations.
11 //
12 //===----------------------------------------------------------------------===//
13 
15 #include "clang/Basic/Diagnostic.h"
20 #include "clang/Lex/Lexer.h"
23 #include "llvm/ADT/SmallString.h"
24 #include "llvm/ADT/SmallVector.h"
25 #include "llvm/ADT/StringRef.h"
26 #include "llvm/Support/FileSystem.h"
27 #include "llvm/Support/raw_ostream.h"
28 #include <cassert>
29 #include <iterator>
30 #include <map>
31 #include <memory>
32 #include <system_error>
33 #include <utility>
34 
35 using namespace clang;
36 
37 raw_ostream &RewriteBuffer::write(raw_ostream &os) const {
38  // Walk RewriteRope chunks efficiently using MoveToNextPiece() instead of the
39  // character iterator.
40  for (RopePieceBTreeIterator I = begin(), E = end(); I != E;
41  I.MoveToNextPiece())
42  os << I.piece();
43  return os;
44 }
45 
46 /// Return true if this character is non-new-line whitespace:
47 /// ' ', '\\t', '\\f', '\\v', '\\r'.
48 static inline bool isWhitespaceExceptNL(unsigned char c) {
49  switch (c) {
50  case ' ':
51  case '\t':
52  case '\f':
53  case '\v':
54  case '\r':
55  return true;
56  default:
57  return false;
58  }
59 }
60 
61 void RewriteBuffer::RemoveText(unsigned OrigOffset, unsigned Size,
62  bool removeLineIfEmpty) {
63  // Nothing to remove, exit early.
64  if (Size == 0) return;
65 
66  unsigned RealOffset = getMappedOffset(OrigOffset, true);
67  assert(RealOffset+Size <= Buffer.size() && "Invalid location");
68 
69  // Remove the dead characters.
70  Buffer.erase(RealOffset, Size);
71 
72  // Add a delta so that future changes are offset correctly.
73  AddReplaceDelta(OrigOffset, -Size);
74 
75  if (removeLineIfEmpty) {
76  // Find the line that the remove occurred and if it is completely empty
77  // remove the line as well.
78 
79  iterator curLineStart = begin();
80  unsigned curLineStartOffs = 0;
81  iterator posI = begin();
82  for (unsigned i = 0; i != RealOffset; ++i) {
83  if (*posI == '\n') {
84  curLineStart = posI;
85  ++curLineStart;
86  curLineStartOffs = i + 1;
87  }
88  ++posI;
89  }
90 
91  unsigned lineSize = 0;
92  posI = curLineStart;
93  while (posI != end() && isWhitespaceExceptNL(*posI)) {
94  ++posI;
95  ++lineSize;
96  }
97  if (posI != end() && *posI == '\n') {
98  Buffer.erase(curLineStartOffs, lineSize + 1/* + '\n'*/);
99  // FIXME: Here, the offset of the start of the line is supposed to be
100  // expressed in terms of the original input not the "real" rewrite
101  // buffer. How do we compute that reliably? It might be tempting to use
102  // curLineStartOffs + OrigOffset - RealOffset, but that assumes the
103  // difference between the original and real offset is the same at the
104  // removed text and at the start of the line, but that's not true if
105  // edits were previously made earlier on the line. This bug is also
106  // documented by a FIXME on the definition of
107  // clang::Rewriter::RewriteOptions::RemoveLineIfEmpty. A reproducer for
108  // the implementation below is the test RemoveLineIfEmpty in
109  // clang/unittests/Rewrite/RewriteBufferTest.cpp.
110  AddReplaceDelta(curLineStartOffs, -(lineSize + 1/* + '\n'*/));
111  }
112  }
113 }
114 
115 void RewriteBuffer::InsertText(unsigned OrigOffset, StringRef Str,
116  bool InsertAfter) {
117  // Nothing to insert, exit early.
118  if (Str.empty()) return;
119 
120  unsigned RealOffset = getMappedOffset(OrigOffset, InsertAfter);
121  Buffer.insert(RealOffset, Str.begin(), Str.end());
122 
123  // Add a delta so that future changes are offset correctly.
124  AddInsertDelta(OrigOffset, Str.size());
125 }
126 
127 /// ReplaceText - This method replaces a range of characters in the input
128 /// buffer with a new string. This is effectively a combined "remove+insert"
129 /// operation.
130 void RewriteBuffer::ReplaceText(unsigned OrigOffset, unsigned OrigLength,
131  StringRef NewStr) {
132  unsigned RealOffset = getMappedOffset(OrigOffset, true);
133  Buffer.erase(RealOffset, OrigLength);
134  Buffer.insert(RealOffset, NewStr.begin(), NewStr.end());
135  if (OrigLength != NewStr.size())
136  AddReplaceDelta(OrigOffset, NewStr.size() - OrigLength);
137 }
138 
139 //===----------------------------------------------------------------------===//
140 // Rewriter class
141 //===----------------------------------------------------------------------===//
142 
143 /// getRangeSize - Return the size in bytes of the specified range if they
144 /// are in the same file. If not, this returns -1.
146  RewriteOptions opts) const {
147  if (!isRewritable(Range.getBegin()) ||
148  !isRewritable(Range.getEnd())) return -1;
149 
150  FileID StartFileID, EndFileID;
151  unsigned StartOff = getLocationOffsetAndFileID(Range.getBegin(), StartFileID);
152  unsigned EndOff = getLocationOffsetAndFileID(Range.getEnd(), EndFileID);
153 
154  if (StartFileID != EndFileID)
155  return -1;
156 
157  // If edits have been made to this buffer, the delta between the range may
158  // have changed.
159  std::map<FileID, RewriteBuffer>::const_iterator I =
160  RewriteBuffers.find(StartFileID);
161  if (I != RewriteBuffers.end()) {
162  const RewriteBuffer &RB = I->second;
163  EndOff = RB.getMappedOffset(EndOff, opts.IncludeInsertsAtEndOfRange);
164  StartOff = RB.getMappedOffset(StartOff, !opts.IncludeInsertsAtBeginOfRange);
165  }
166 
167  // Adjust the end offset to the end of the last token, instead of being the
168  // start of the last token if this is a token range.
169  if (Range.isTokenRange())
170  EndOff += Lexer::MeasureTokenLength(Range.getEnd(), *SourceMgr, *LangOpts);
171 
172  return EndOff-StartOff;
173 }
174 
176  return getRangeSize(CharSourceRange::getTokenRange(Range), opts);
177 }
178 
179 /// getRewrittenText - Return the rewritten form of the text in the specified
180 /// range. If the start or end of the range was unrewritable or if they are
181 /// in different buffers, this returns an empty string.
182 ///
183 /// Note that this method is not particularly efficient.
184 std::string Rewriter::getRewrittenText(CharSourceRange Range) const {
185  if (!isRewritable(Range.getBegin()) ||
186  !isRewritable(Range.getEnd()))
187  return {};
188 
189  FileID StartFileID, EndFileID;
190  unsigned StartOff, EndOff;
191  StartOff = getLocationOffsetAndFileID(Range.getBegin(), StartFileID);
192  EndOff = getLocationOffsetAndFileID(Range.getEnd(), EndFileID);
193 
194  if (StartFileID != EndFileID)
195  return {}; // Start and end in different buffers.
196 
197  // If edits have been made to this buffer, the delta between the range may
198  // have changed.
199  std::map<FileID, RewriteBuffer>::const_iterator I =
200  RewriteBuffers.find(StartFileID);
201  if (I == RewriteBuffers.end()) {
202  // If the buffer hasn't been rewritten, just return the text from the input.
203  const char *Ptr = SourceMgr->getCharacterData(Range.getBegin());
204 
205  // Adjust the end offset to the end of the last token, instead of being the
206  // start of the last token.
207  if (Range.isTokenRange())
208  EndOff +=
209  Lexer::MeasureTokenLength(Range.getEnd(), *SourceMgr, *LangOpts);
210  return std::string(Ptr, Ptr+EndOff-StartOff);
211  }
212 
213  const RewriteBuffer &RB = I->second;
214  EndOff = RB.getMappedOffset(EndOff, true);
215  StartOff = RB.getMappedOffset(StartOff);
216 
217  // Adjust the end offset to the end of the last token, instead of being the
218  // start of the last token.
219  if (Range.isTokenRange())
220  EndOff += Lexer::MeasureTokenLength(Range.getEnd(), *SourceMgr, *LangOpts);
221 
222  // Advance the iterators to the right spot, yay for linear time algorithms.
223  RewriteBuffer::iterator Start = RB.begin();
224  std::advance(Start, StartOff);
226  std::advance(End, EndOff-StartOff);
227 
228  return std::string(Start, End);
229 }
230 
231 unsigned Rewriter::getLocationOffsetAndFileID(SourceLocation Loc,
232  FileID &FID) const {
233  assert(Loc.isValid() && "Invalid location");
234  std::pair<FileID, unsigned> V = SourceMgr->getDecomposedLoc(Loc);
235  FID = V.first;
236  return V.second;
237 }
238 
239 /// getEditBuffer - Get or create a RewriteBuffer for the specified FileID.
241  std::map<FileID, RewriteBuffer>::iterator I =
242  RewriteBuffers.lower_bound(FID);
243  if (I != RewriteBuffers.end() && I->first == FID)
244  return I->second;
245  I = RewriteBuffers.insert(I, std::make_pair(FID, RewriteBuffer()));
246 
247  StringRef MB = SourceMgr->getBufferData(FID);
248  I->second.Initialize(MB.begin(), MB.end());
249 
250  return I->second;
251 }
252 
253 /// InsertText - Insert the specified string at the specified location in the
254 /// original buffer.
255 bool Rewriter::InsertText(SourceLocation Loc, StringRef Str,
256  bool InsertAfter, bool indentNewLines) {
257  if (!isRewritable(Loc)) return true;
258  FileID FID;
259  unsigned StartOffs = getLocationOffsetAndFileID(Loc, FID);
260 
261  SmallString<128> indentedStr;
262  if (indentNewLines && Str.find('\n') != StringRef::npos) {
263  StringRef MB = SourceMgr->getBufferData(FID);
264 
265  unsigned lineNo = SourceMgr->getLineNumber(FID, StartOffs) - 1;
266  const SrcMgr::ContentCache *
267  Content = SourceMgr->getSLocEntry(FID).getFile().getContentCache();
268  unsigned lineOffs = Content->SourceLineCache[lineNo];
269 
270  // Find the whitespace at the start of the line.
271  StringRef indentSpace;
272  {
273  unsigned i = lineOffs;
274  while (isWhitespaceExceptNL(MB[i]))
275  ++i;
276  indentSpace = MB.substr(lineOffs, i-lineOffs);
277  }
278 
280  Str.split(lines, "\n");
281 
282  for (unsigned i = 0, e = lines.size(); i != e; ++i) {
283  indentedStr += lines[i];
284  if (i < e-1) {
285  indentedStr += '\n';
286  indentedStr += indentSpace;
287  }
288  }
289  Str = indentedStr.str();
290  }
291 
292  getEditBuffer(FID).InsertText(StartOffs, Str, InsertAfter);
293  return false;
294 }
295 
297  if (!isRewritable(Loc)) return true;
298  FileID FID;
299  unsigned StartOffs = getLocationOffsetAndFileID(Loc, FID);
300  RewriteOptions rangeOpts;
301  rangeOpts.IncludeInsertsAtBeginOfRange = false;
302  StartOffs += getRangeSize(SourceRange(Loc, Loc), rangeOpts);
303  getEditBuffer(FID).InsertText(StartOffs, Str, /*InsertAfter*/true);
304  return false;
305 }
306 
307 /// RemoveText - Remove the specified text region.
308 bool Rewriter::RemoveText(SourceLocation Start, unsigned Length,
309  RewriteOptions opts) {
310  if (!isRewritable(Start)) return true;
311  FileID FID;
312  unsigned StartOffs = getLocationOffsetAndFileID(Start, FID);
313  getEditBuffer(FID).RemoveText(StartOffs, Length, opts.RemoveLineIfEmpty);
314  return false;
315 }
316 
317 /// ReplaceText - This method replaces a range of characters in the input
318 /// buffer with a new string. This is effectively a combined "remove/insert"
319 /// operation.
320 bool Rewriter::ReplaceText(SourceLocation Start, unsigned OrigLength,
321  StringRef NewStr) {
322  if (!isRewritable(Start)) return true;
323  FileID StartFileID;
324  unsigned StartOffs = getLocationOffsetAndFileID(Start, StartFileID);
325 
326  getEditBuffer(StartFileID).ReplaceText(StartOffs, OrigLength, NewStr);
327  return false;
328 }
329 
331  if (!isRewritable(range.getBegin())) return true;
332  if (!isRewritable(range.getEnd())) return true;
333  if (replacementRange.isInvalid()) return true;
334  SourceLocation start = range.getBegin();
335  unsigned origLength = getRangeSize(range);
336  unsigned newLength = getRangeSize(replacementRange);
337  FileID FID;
338  unsigned newOffs = getLocationOffsetAndFileID(replacementRange.getBegin(),
339  FID);
340  StringRef MB = SourceMgr->getBufferData(FID);
341  return ReplaceText(start, origLength, MB.substr(newOffs, newLength));
342 }
343 
345  SourceLocation parentIndent) {
346  if (range.isInvalid()) return true;
347  if (!isRewritable(range.getBegin())) return true;
348  if (!isRewritable(range.getEnd())) return true;
349  if (!isRewritable(parentIndent)) return true;
350 
351  FileID StartFileID, EndFileID, parentFileID;
352  unsigned StartOff, EndOff, parentOff;
353 
354  StartOff = getLocationOffsetAndFileID(range.getBegin(), StartFileID);
355  EndOff = getLocationOffsetAndFileID(range.getEnd(), EndFileID);
356  parentOff = getLocationOffsetAndFileID(parentIndent, parentFileID);
357 
358  if (StartFileID != EndFileID || StartFileID != parentFileID)
359  return true;
360  if (StartOff > EndOff)
361  return true;
362 
363  FileID FID = StartFileID;
364  StringRef MB = SourceMgr->getBufferData(FID);
365 
366  unsigned parentLineNo = SourceMgr->getLineNumber(FID, parentOff) - 1;
367  unsigned startLineNo = SourceMgr->getLineNumber(FID, StartOff) - 1;
368  unsigned endLineNo = SourceMgr->getLineNumber(FID, EndOff) - 1;
369 
370  const SrcMgr::ContentCache *
371  Content = SourceMgr->getSLocEntry(FID).getFile().getContentCache();
372 
373  // Find where the lines start.
374  unsigned parentLineOffs = Content->SourceLineCache[parentLineNo];
375  unsigned startLineOffs = Content->SourceLineCache[startLineNo];
376 
377  // Find the whitespace at the start of each line.
378  StringRef parentSpace, startSpace;
379  {
380  unsigned i = parentLineOffs;
381  while (isWhitespaceExceptNL(MB[i]))
382  ++i;
383  parentSpace = MB.substr(parentLineOffs, i-parentLineOffs);
384 
385  i = startLineOffs;
386  while (isWhitespaceExceptNL(MB[i]))
387  ++i;
388  startSpace = MB.substr(startLineOffs, i-startLineOffs);
389  }
390  if (parentSpace.size() >= startSpace.size())
391  return true;
392  if (!startSpace.startswith(parentSpace))
393  return true;
394 
395  StringRef indent = startSpace.substr(parentSpace.size());
396 
397  // Indent the lines between start/end offsets.
398  RewriteBuffer &RB = getEditBuffer(FID);
399  for (unsigned lineNo = startLineNo; lineNo <= endLineNo; ++lineNo) {
400  unsigned offs = Content->SourceLineCache[lineNo];
401  unsigned i = offs;
402  while (isWhitespaceExceptNL(MB[i]))
403  ++i;
404  StringRef origIndent = MB.substr(offs, i-offs);
405  if (origIndent.startswith(startSpace))
406  RB.InsertText(offs, indent, /*InsertAfter=*/false);
407  }
408 
409  return false;
410 }
411 
412 namespace {
413 
414 // A wrapper for a file stream that atomically overwrites the target.
415 //
416 // Creates a file output stream for a temporary file in the constructor,
417 // which is later accessible via getStream() if ok() return true.
418 // Flushes the stream and moves the temporary file to the target location
419 // in the destructor.
420 class AtomicallyMovedFile {
421 public:
422  AtomicallyMovedFile(DiagnosticsEngine &Diagnostics, StringRef Filename,
423  bool &AllWritten)
424  : Diagnostics(Diagnostics), Filename(Filename), AllWritten(AllWritten) {
425  TempFilename = Filename;
426  TempFilename += "-%%%%%%%%";
427  int FD;
428  if (llvm::sys::fs::createUniqueFile(TempFilename, FD, TempFilename)) {
429  AllWritten = false;
430  Diagnostics.Report(clang::diag::err_unable_to_make_temp)
431  << TempFilename;
432  } else {
433  FileStream.reset(new llvm::raw_fd_ostream(FD, /*shouldClose=*/true));
434  }
435  }
436 
437  ~AtomicallyMovedFile() {
438  if (!ok()) return;
439 
440  // Close (will also flush) theFileStream.
441  FileStream->close();
442  if (std::error_code ec = llvm::sys::fs::rename(TempFilename, Filename)) {
443  AllWritten = false;
444  Diagnostics.Report(clang::diag::err_unable_to_rename_temp)
445  << TempFilename << Filename << ec.message();
446  // If the remove fails, there's not a lot we can do - this is already an
447  // error.
448  llvm::sys::fs::remove(TempFilename);
449  }
450  }
451 
452  bool ok() { return (bool)FileStream; }
453  raw_ostream &getStream() { return *FileStream; }
454 
455 private:
456  DiagnosticsEngine &Diagnostics;
457  StringRef Filename;
458  SmallString<128> TempFilename;
459  std::unique_ptr<llvm::raw_fd_ostream> FileStream;
460  bool &AllWritten;
461 };
462 
463 } // namespace
464 
466  bool AllWritten = true;
467  for (buffer_iterator I = buffer_begin(), E = buffer_end(); I != E; ++I) {
468  const FileEntry *Entry =
469  getSourceMgr().getFileEntryForID(I->first);
470  AtomicallyMovedFile File(getSourceMgr().getDiagnostics(), Entry->getName(),
471  AllWritten);
472  if (File.ok()) {
473  I->second.write(File.getStream());
474  }
475  }
476  return !AllWritten;
477 }
void insert(unsigned Offset, const char *Start, const char *End)
Definition: RewriteRope.h:196
bool IncreaseIndentation(CharSourceRange range, SourceLocation parentIndent)
Increase indentation for the lines between the given source range.
Definition: Rewriter.cpp:344
Defines the clang::FileManager interface and associated types.
Defines the SourceManager interface.
static CharSourceRange getTokenRange(SourceRange R)
void Initialize(const char *BufStart, const char *BufEnd)
Initialize - Start this rewrite buffer out with a copy of the unmodified input buffer.
Definition: RewriteBuffer.h:43
DiagnosticBuilder Report(SourceLocation Loc, unsigned DiagID)
Issue the message to the client.
Definition: Diagnostic.h:1290
bool RemoveText(SourceLocation Start, unsigned Length, RewriteOptions opts=RewriteOptions())
RemoveText - Remove the specified text region.
Definition: Rewriter.cpp:308
void erase(unsigned Offset, unsigned NumBytes)
Definition: RewriteRope.h:202
RewriteBuffer & getEditBuffer(FileID FID)
getEditBuffer - This is like getRewriteBufferFor, but always returns a buffer, and allows you to writ...
Definition: Rewriter.cpp:240
RewriteBuffer - As code is rewritten, SourceBuffer&#39;s from the original input with modifications get a...
Definition: RewriteBuffer.h:25
RangeSelector range(RangeSelector Begin, RangeSelector End)
Selects from the start of Begin and to the end of End.
std::string getRewrittenText(CharSourceRange Range) const
getRewrittenText - Return the rewritten form of the text in the specified range.
Definition: Rewriter.cpp:184
iterator end() const
Definition: RewriteBuffer.h:38
SourceLocation getBegin() const
One instance of this struct is kept for every file loaded or used.
Definition: SourceManager.h:94
bool ReplaceText(SourceLocation Start, unsigned OrigLength, StringRef NewStr)
ReplaceText - This method replaces a range of characters in the input buffer with a new string...
Definition: Rewriter.cpp:320
Concrete class used by the front-end to report problems and issues.
Definition: Diagnostic.h:149
Defines the Diagnostic-related interfaces.
StringRef Filename
Definition: Format.cpp:1756
SourceLocation End
Represents a character-granular source range.
static unsigned MeasureTokenLength(SourceLocation Loc, const SourceManager &SM, const LangOptions &LangOpts)
MeasureTokenLength - Relex the token at the specified location and return its length in bytes in the ...
Definition: Lexer.cpp:444
#define V(N, I)
Definition: ASTContext.h:2913
ASTEdit remove(RangeSelector S)
Removes the source selected by S.
Definition: Transformer.h:230
bool isInvalid() const
SourceLocation getEnd() const
int getRangeSize(SourceRange Range, RewriteOptions opts=RewriteOptions()) const
getRangeSize - Return the size in bytes of the specified range if they are in the same file...
Definition: Rewriter.cpp:175
void InsertText(unsigned OrigOffset, StringRef Str, bool InsertAfter=true)
InsertText - Insert some text at the specified point, where the offset in the buffer is specified rel...
Definition: Rewriter.cpp:115
bool overwriteChangedFiles()
overwriteChangedFiles - Save all changed files to disk.
Definition: Rewriter.cpp:465
iterator begin() const
Definition: RewriteBuffer.h:37
void ReplaceText(unsigned OrigOffset, unsigned OrigLength, StringRef NewStr)
ReplaceText - This method replaces a range of characters in the input buffer with a new string...
Definition: Rewriter.cpp:130
Encodes a location in the source.
static bool isWhitespaceExceptNL(unsigned char c)
Return true if this character is non-new-line whitespace: &#39; &#39;, &#39;\t&#39;, &#39;\f&#39;, &#39;\v&#39;, &#39;\r&#39;.
Definition: Rewriter.cpp:48
raw_ostream & write(raw_ostream &Stream) const
Write to Stream the result of applying all changes to the original buffer.
Definition: Rewriter.cpp:37
StringRef getName() const
Definition: FileManager.h:102
unsigned size() const
Definition: RewriteRope.h:184
Cached information about one file (either on disk or in the virtual file system). ...
Definition: FileManager.h:78
bool InsertText(SourceLocation Loc, StringRef Str, bool InsertAfter=true, bool indentNewLines=false)
InsertText - Insert the specified string at the specified location in the original buffer...
Definition: Rewriter.cpp:255
bool RemoveLineIfEmpty
If true and removing some text leaves a blank line also remove the empty line (false by default)...
Definition: Rewriter.h:60
void RemoveText(unsigned OrigOffset, unsigned Size, bool removeLineIfEmpty=false)
RemoveText - Remove the specified text.
Definition: Rewriter.cpp:61
unsigned * SourceLineCache
A bump pointer allocated array of offsets for each source line.
std::map< FileID, RewriteBuffer >::iterator buffer_iterator
Definition: Rewriter.h:65
bool isTokenRange() const
Return true if the end of this range specifies the start of the last token.
An opaque identifier used by SourceManager which refers to a source file (MemoryBuffer) along with it...
Dataflow Directional Tag Classes.
bool isValid() const
Return true if this is a valid SourceLocation object.
RopePieceBTreeIterator - This class provides read-only forward iteration over bytes that are in a Rop...
Definition: RewriteRope.h:86
static int getRangeSize(const SourceManager &Sources, const CharSourceRange &Range, const LangOptions &LangOpts)
bool IncludeInsertsAtBeginOfRange
Given a source range, true to include previous inserts at the beginning of the range as part of the r...
Definition: Rewriter.h:41
SourceLocation getEnd() const
Defines the clang::SourceLocation class and associated facilities.
Defines the Diagnostic IDs-related interfaces.
A trivial tuple used to represent a source range.
bool InsertTextAfterToken(SourceLocation Loc, StringRef Str)
Insert the specified string after the token in the specified location.
Definition: Rewriter.cpp:296
SourceLocation getBegin() const
bool IncludeInsertsAtEndOfRange
Given a source range, true to include previous inserts at the end of the range as part of the range i...
Definition: Rewriter.h:45