1 //===- SourceMgr.cpp - Manager for Simple Source Buffers & Diagnostics ----===//
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 //===----------------------------------------------------------------------===//
10 // This file implements the SourceMgr class. This class is used as a simple
11 // substrate for diagnostics, #include handling, and other low level things for
14 //===----------------------------------------------------------------------===//
16 #include "llvm/Support/SourceMgr.h"
17 #include "llvm/ADT/ArrayRef.h"
18 #include "llvm/ADT/STLExtras.h"
19 #include "llvm/ADT/SmallVector.h"
20 #include "llvm/ADT/StringRef.h"
21 #include "llvm/ADT/Twine.h"
22 #include "llvm/Support/ErrorOr.h"
23 #include "llvm/Support/Locale.h"
24 #include "llvm/Support/MemoryBuffer.h"
25 #include "llvm/Support/Path.h"
26 #include "llvm/Support/SMLoc.h"
27 #include "llvm/Support/WithColor.h"
28 #include "llvm/Support/raw_ostream.h"
39 static const size_t TabStop = 8;
41 unsigned SourceMgr::AddIncludeFile(const std::string &Filename,
43 std::string &IncludedFile) {
44 IncludedFile = Filename;
45 ErrorOr<std::unique_ptr<MemoryBuffer>> NewBufOrErr =
46 MemoryBuffer::getFile(IncludedFile);
48 // If the file didn't exist directly, see if it's in an include path.
49 for (unsigned i = 0, e = IncludeDirectories.size(); i != e && !NewBufOrErr;
52 IncludeDirectories[i] + sys::path::get_separator().data() + Filename;
53 NewBufOrErr = MemoryBuffer::getFile(IncludedFile);
59 return AddNewSourceBuffer(std::move(*NewBufOrErr), IncludeLoc);
62 unsigned SourceMgr::FindBufferContainingLoc(SMLoc Loc) const {
63 for (unsigned i = 0, e = Buffers.size(); i != e; ++i)
64 if (Loc.getPointer() >= Buffers[i].Buffer->getBufferStart() &&
65 // Use <= here so that a pointer to the null at the end of the buffer
66 // is included as part of the buffer.
67 Loc.getPointer() <= Buffers[i].Buffer->getBufferEnd())
73 unsigned SourceMgr::SrcBuffer::getLineNumber(const char *Ptr) const {
75 // Ensure OffsetCache is allocated and populated with offsets of all the
77 std::vector<T> *Offsets = nullptr;
78 if (OffsetCache.isNull()) {
79 Offsets = new std::vector<T>();
80 OffsetCache = Offsets;
81 size_t Sz = Buffer->getBufferSize();
82 assert(Sz <= std::numeric_limits<T>::max());
83 StringRef S = Buffer->getBuffer();
84 for (size_t N = 0; N < Sz; ++N) {
86 Offsets->push_back(static_cast<T>(N));
90 Offsets = OffsetCache.get<std::vector<T> *>();
93 const char *BufStart = Buffer->getBufferStart();
94 assert(Ptr >= BufStart && Ptr <= Buffer->getBufferEnd());
95 ptrdiff_t PtrDiff = Ptr - BufStart;
96 assert(PtrDiff >= 0 && static_cast<size_t>(PtrDiff) <= std::numeric_limits<T>::max());
97 T PtrOffset = static_cast<T>(PtrDiff);
99 // std::lower_bound returns the first EOL offset that's not-less-than
100 // PtrOffset, meaning the EOL that _ends the line_ that PtrOffset is on
101 // (including if PtrOffset refers to the EOL itself). If there's no such
102 // EOL, returns end().
103 auto EOL = std::lower_bound(Offsets->begin(), Offsets->end(), PtrOffset);
105 // Lines count from 1, so add 1 to the distance from the 0th line.
106 return (1 + (EOL - Offsets->begin()));
109 SourceMgr::SrcBuffer::SrcBuffer(SourceMgr::SrcBuffer &&Other)
110 : Buffer(std::move(Other.Buffer)),
111 OffsetCache(Other.OffsetCache),
112 IncludeLoc(Other.IncludeLoc) {
113 Other.OffsetCache = nullptr;
116 SourceMgr::SrcBuffer::~SrcBuffer() {
117 if (!OffsetCache.isNull()) {
118 if (OffsetCache.is<std::vector<uint8_t>*>())
119 delete OffsetCache.get<std::vector<uint8_t>*>();
120 else if (OffsetCache.is<std::vector<uint16_t>*>())
121 delete OffsetCache.get<std::vector<uint16_t>*>();
122 else if (OffsetCache.is<std::vector<uint32_t>*>())
123 delete OffsetCache.get<std::vector<uint32_t>*>();
125 delete OffsetCache.get<std::vector<uint64_t>*>();
126 OffsetCache = nullptr;
130 std::pair<unsigned, unsigned>
131 SourceMgr::getLineAndColumn(SMLoc Loc, unsigned BufferID) const {
133 BufferID = FindBufferContainingLoc(Loc);
134 assert(BufferID && "Invalid Location!");
136 auto &SB = getBufferInfo(BufferID);
137 const char *Ptr = Loc.getPointer();
139 size_t Sz = SB.Buffer->getBufferSize();
141 if (Sz <= std::numeric_limits<uint8_t>::max())
142 LineNo = SB.getLineNumber<uint8_t>(Ptr);
143 else if (Sz <= std::numeric_limits<uint16_t>::max())
144 LineNo = SB.getLineNumber<uint16_t>(Ptr);
145 else if (Sz <= std::numeric_limits<uint32_t>::max())
146 LineNo = SB.getLineNumber<uint32_t>(Ptr);
148 LineNo = SB.getLineNumber<uint64_t>(Ptr);
150 const char *BufStart = SB.Buffer->getBufferStart();
151 size_t NewlineOffs = StringRef(BufStart, Ptr-BufStart).find_last_of("\n\r");
152 if (NewlineOffs == StringRef::npos) NewlineOffs = ~(size_t)0;
153 return std::make_pair(LineNo, Ptr-BufStart-NewlineOffs);
156 void SourceMgr::PrintIncludeStack(SMLoc IncludeLoc, raw_ostream &OS) const {
157 if (IncludeLoc == SMLoc()) return; // Top of stack.
159 unsigned CurBuf = FindBufferContainingLoc(IncludeLoc);
160 assert(CurBuf && "Invalid or unspecified location!");
162 PrintIncludeStack(getBufferInfo(CurBuf).IncludeLoc, OS);
164 OS << "Included from "
165 << getBufferInfo(CurBuf).Buffer->getBufferIdentifier()
166 << ":" << FindLineNumber(IncludeLoc, CurBuf) << ":\n";
169 SMDiagnostic SourceMgr::GetMessage(SMLoc Loc, SourceMgr::DiagKind Kind,
171 ArrayRef<SMRange> Ranges,
172 ArrayRef<SMFixIt> FixIts) const {
173 // First thing to do: find the current buffer containing the specified
174 // location to pull out the source line.
175 SmallVector<std::pair<unsigned, unsigned>, 4> ColRanges;
176 std::pair<unsigned, unsigned> LineAndCol;
177 StringRef BufferID = "<unknown>";
181 unsigned CurBuf = FindBufferContainingLoc(Loc);
182 assert(CurBuf && "Invalid or unspecified location!");
184 const MemoryBuffer *CurMB = getMemoryBuffer(CurBuf);
185 BufferID = CurMB->getBufferIdentifier();
187 // Scan backward to find the start of the line.
188 const char *LineStart = Loc.getPointer();
189 const char *BufStart = CurMB->getBufferStart();
190 while (LineStart != BufStart && LineStart[-1] != '\n' &&
191 LineStart[-1] != '\r')
194 // Get the end of the line.
195 const char *LineEnd = Loc.getPointer();
196 const char *BufEnd = CurMB->getBufferEnd();
197 while (LineEnd != BufEnd && LineEnd[0] != '\n' && LineEnd[0] != '\r')
199 LineStr = std::string(LineStart, LineEnd);
201 // Convert any ranges to column ranges that only intersect the line of the
203 for (unsigned i = 0, e = Ranges.size(); i != e; ++i) {
204 SMRange R = Ranges[i];
205 if (!R.isValid()) continue;
207 // If the line doesn't contain any part of the range, then ignore it.
208 if (R.Start.getPointer() > LineEnd || R.End.getPointer() < LineStart)
211 // Ignore pieces of the range that go onto other lines.
212 if (R.Start.getPointer() < LineStart)
213 R.Start = SMLoc::getFromPointer(LineStart);
214 if (R.End.getPointer() > LineEnd)
215 R.End = SMLoc::getFromPointer(LineEnd);
217 // Translate from SMLoc ranges to column ranges.
218 // FIXME: Handle multibyte characters.
219 ColRanges.push_back(std::make_pair(R.Start.getPointer()-LineStart,
220 R.End.getPointer()-LineStart));
223 LineAndCol = getLineAndColumn(Loc, CurBuf);
226 return SMDiagnostic(*this, Loc, BufferID, LineAndCol.first,
227 LineAndCol.second-1, Kind, Msg.str(),
228 LineStr, ColRanges, FixIts);
231 void SourceMgr::PrintMessage(raw_ostream &OS, const SMDiagnostic &Diagnostic,
232 bool ShowColors) const {
233 // Report the message with the diagnostic handler if present.
235 DiagHandler(Diagnostic, DiagContext);
239 if (Diagnostic.getLoc().isValid()) {
240 unsigned CurBuf = FindBufferContainingLoc(Diagnostic.getLoc());
241 assert(CurBuf && "Invalid or unspecified location!");
242 PrintIncludeStack(getBufferInfo(CurBuf).IncludeLoc, OS);
245 Diagnostic.print(nullptr, OS, ShowColors);
248 void SourceMgr::PrintMessage(raw_ostream &OS, SMLoc Loc,
249 SourceMgr::DiagKind Kind,
250 const Twine &Msg, ArrayRef<SMRange> Ranges,
251 ArrayRef<SMFixIt> FixIts, bool ShowColors) const {
252 PrintMessage(OS, GetMessage(Loc, Kind, Msg, Ranges, FixIts), ShowColors);
255 void SourceMgr::PrintMessage(SMLoc Loc, SourceMgr::DiagKind Kind,
256 const Twine &Msg, ArrayRef<SMRange> Ranges,
257 ArrayRef<SMFixIt> FixIts, bool ShowColors) const {
258 PrintMessage(errs(), Loc, Kind, Msg, Ranges, FixIts, ShowColors);
261 //===----------------------------------------------------------------------===//
262 // SMDiagnostic Implementation
263 //===----------------------------------------------------------------------===//
265 SMDiagnostic::SMDiagnostic(const SourceMgr &sm, SMLoc L, StringRef FN,
266 int Line, int Col, SourceMgr::DiagKind Kind,
267 StringRef Msg, StringRef LineStr,
268 ArrayRef<std::pair<unsigned,unsigned>> Ranges,
269 ArrayRef<SMFixIt> Hints)
270 : SM(&sm), Loc(L), Filename(FN), LineNo(Line), ColumnNo(Col), Kind(Kind),
271 Message(Msg), LineContents(LineStr), Ranges(Ranges.vec()),
272 FixIts(Hints.begin(), Hints.end()) {
276 static void buildFixItLine(std::string &CaretLine, std::string &FixItLine,
277 ArrayRef<SMFixIt> FixIts, ArrayRef<char> SourceLine){
281 const char *LineStart = SourceLine.begin();
282 const char *LineEnd = SourceLine.end();
284 size_t PrevHintEndCol = 0;
286 for (ArrayRef<SMFixIt>::iterator I = FixIts.begin(), E = FixIts.end();
288 // If the fixit contains a newline or tab, ignore it.
289 if (I->getText().find_first_of("\n\r\t") != StringRef::npos)
292 SMRange R = I->getRange();
294 // If the line doesn't contain any part of the range, then ignore it.
295 if (R.Start.getPointer() > LineEnd || R.End.getPointer() < LineStart)
298 // Translate from SMLoc to column.
299 // Ignore pieces of the range that go onto other lines.
300 // FIXME: Handle multibyte characters in the source line.
302 if (R.Start.getPointer() < LineStart)
305 FirstCol = R.Start.getPointer() - LineStart;
307 // If we inserted a long previous hint, push this one forwards, and add
308 // an extra space to show that this is not part of the previous
309 // completion. This is sort of the best we can do when two hints appear
312 // Note that if this hint is located immediately after the previous
313 // hint, no space will be added, since the location is more important.
314 unsigned HintCol = FirstCol;
315 if (HintCol < PrevHintEndCol)
316 HintCol = PrevHintEndCol + 1;
318 // FIXME: This assertion is intended to catch unintended use of multibyte
319 // characters in fixits. If we decide to do this, we'll have to track
320 // separate byte widths for the source and fixit lines.
321 assert((size_t)sys::locale::columnWidth(I->getText()) ==
322 I->getText().size());
324 // This relies on one byte per column in our fixit hints.
325 unsigned LastColumnModified = HintCol + I->getText().size();
326 if (LastColumnModified > FixItLine.size())
327 FixItLine.resize(LastColumnModified, ' ');
329 std::copy(I->getText().begin(), I->getText().end(),
330 FixItLine.begin() + HintCol);
332 PrevHintEndCol = LastColumnModified;
334 // For replacements, mark the removal range with '~'.
335 // FIXME: Handle multibyte characters in the source line.
337 if (R.End.getPointer() >= LineEnd)
338 LastCol = LineEnd - LineStart;
340 LastCol = R.End.getPointer() - LineStart;
342 std::fill(&CaretLine[FirstCol], &CaretLine[LastCol], '~');
346 static void printSourceLine(raw_ostream &S, StringRef LineContents) {
347 // Print out the source line one character at a time, so we can expand tabs.
348 for (unsigned i = 0, e = LineContents.size(), OutCol = 0; i != e; ++i) {
349 size_t NextTab = LineContents.find('\t', i);
350 // If there were no tabs left, print the rest, we are done.
351 if (NextTab == StringRef::npos) {
352 S << LineContents.drop_front(i);
356 // Otherwise, print from i to NextTab.
357 S << LineContents.slice(i, NextTab);
358 OutCol += NextTab - i;
361 // If we have a tab, emit at least one space, then round up to 8 columns.
365 } while ((OutCol % TabStop) != 0);
370 static bool isNonASCII(char c) {
374 void SMDiagnostic::print(const char *ProgName, raw_ostream &OS,
375 bool ShowColors, bool ShowKindLabel) const {
377 WithColor S(OS, raw_ostream::SAVEDCOLOR, true, false, !ShowColors);
379 if (ProgName && ProgName[0])
380 S << ProgName << ": ";
382 if (!Filename.empty()) {
391 S << ':' << (ColumnNo + 1);
399 case SourceMgr::DK_Error:
400 WithColor::error(OS, "", !ShowColors);
402 case SourceMgr::DK_Warning:
403 WithColor::warning(OS, "", !ShowColors);
405 case SourceMgr::DK_Note:
406 WithColor::note(OS, "", !ShowColors);
408 case SourceMgr::DK_Remark:
409 WithColor::remark(OS, "", !ShowColors);
414 WithColor(OS, raw_ostream::SAVEDCOLOR, true, false, !ShowColors)
417 if (LineNo == -1 || ColumnNo == -1)
420 // FIXME: If there are multibyte or multi-column characters in the source, all
421 // our ranges will be wrong. To do this properly, we'll need a byte-to-column
422 // map like Clang's TextDiagnostic. For now, we'll just handle tabs by
423 // expanding them later, and bail out rather than show incorrect ranges and
424 // misaligned fixits for any other odd characters.
425 if (find_if(LineContents, isNonASCII) != LineContents.end()) {
426 printSourceLine(OS, LineContents);
429 size_t NumColumns = LineContents.size();
431 // Build the line with the caret and ranges.
432 std::string CaretLine(NumColumns+1, ' ');
434 // Expand any ranges.
435 for (unsigned r = 0, e = Ranges.size(); r != e; ++r) {
436 std::pair<unsigned, unsigned> R = Ranges[r];
437 std::fill(&CaretLine[R.first],
438 &CaretLine[std::min((size_t)R.second, CaretLine.size())],
443 // FIXME: Find the beginning of the line properly for multibyte characters.
444 std::string FixItInsertionLine;
445 buildFixItLine(CaretLine, FixItInsertionLine, FixIts,
446 makeArrayRef(Loc.getPointer() - ColumnNo,
447 LineContents.size()));
449 // Finally, plop on the caret.
450 if (unsigned(ColumnNo) <= NumColumns)
451 CaretLine[ColumnNo] = '^';
453 CaretLine[NumColumns] = '^';
455 // ... and remove trailing whitespace so the output doesn't wrap for it. We
456 // know that the line isn't completely empty because it has the caret in it at
458 CaretLine.erase(CaretLine.find_last_not_of(' ')+1);
460 printSourceLine(OS, LineContents);
463 WithColor S(OS, raw_ostream::GREEN, true, false, !ShowColors);
465 // Print out the caret line, matching tabs in the source line.
466 for (unsigned i = 0, e = CaretLine.size(), OutCol = 0; i != e; ++i) {
467 if (i >= LineContents.size() || LineContents[i] != '\t') {
473 // Okay, we have a tab. Insert the appropriate number of characters.
477 } while ((OutCol % TabStop) != 0);
482 // Print out the replacement line, matching tabs in the source line.
483 if (FixItInsertionLine.empty())
486 for (size_t i = 0, e = FixItInsertionLine.size(), OutCol = 0; i < e; ++i) {
487 if (i >= LineContents.size() || LineContents[i] != '\t') {
488 OS << FixItInsertionLine[i];
493 // Okay, we have a tab. Insert the appropriate number of characters.
495 OS << FixItInsertionLine[i];
496 // FIXME: This is trying not to break up replacements, but then to re-sync
497 // with the tabs between replacements. This will fail, though, if two
498 // fix-it replacements are exactly adjacent, or if a fix-it contains a
499 // space. Really we should be precomputing column widths, which we'll
500 // need anyway for multibyte chars.
501 if (FixItInsertionLine[i] != ' ')
504 } while (((OutCol % TabStop) != 0) && i != e);