1 //===- FileCheck.cpp - Check that File's Contents match what is expected --===//
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 // FileCheck does a line-by line check of a file that validates whether it
11 // contains the expected content. This is useful for regression tests etc.
13 // This program exits with an error status of 2 on error, exit status of 0 if
14 // the file matched the expected contents, and exit status of 1 if it did not
15 // contain the expected contents.
17 //===----------------------------------------------------------------------===//
19 #include "llvm/Support/CommandLine.h"
20 #include "llvm/Support/MemoryBuffer.h"
21 #include "llvm/Support/PrettyStackTrace.h"
22 #include "llvm/Support/Regex.h"
23 #include "llvm/Support/SourceMgr.h"
24 #include "llvm/Support/raw_ostream.h"
25 #include "llvm/System/Signals.h"
28 static cl::opt<std::string>
29 CheckFilename(cl::Positional, cl::desc("<check-file>"), cl::Required);
31 static cl::opt<std::string>
32 InputFilename("input-file", cl::desc("File to check (defaults to stdin)"),
33 cl::init("-"), cl::value_desc("filename"));
35 static cl::opt<std::string>
36 CheckPrefix("check-prefix", cl::init("CHECK"),
37 cl::desc("Prefix to use from check file (defaults to 'CHECK')"));
40 NoCanonicalizeWhiteSpace("strict-whitespace",
41 cl::desc("Do not treat all horizontal whitespace as equivalent"));
43 //===----------------------------------------------------------------------===//
44 // Pattern Handling Code.
45 //===----------------------------------------------------------------------===//
51 PatternChunk(StringRef S, bool isRE) : Str(S), isRegEx(isRE) {}
53 size_t Match(StringRef Buffer, size_t &MatchLen) const {
55 // Fixed string match.
56 MatchLen = Str.size();
57 return Buffer.find(Str);
61 SmallVector<StringRef, 4> MatchInfo;
62 if (!Regex(Str, Regex::Sub|Regex::Newline).match(Buffer, &MatchInfo))
63 return StringRef::npos;
65 // Successful regex match.
66 assert(!MatchInfo.empty() && "Didn't get any match");
67 StringRef FullMatch = MatchInfo[0];
69 MatchLen = FullMatch.size();
70 return FullMatch.data()-Buffer.data();
75 /// Chunks - The pattern chunks to match. If the bool is false, it is a fixed
76 /// string match, if it is true, it is a regex match.
77 SmallVector<PatternChunk, 4> Chunks;
84 bool ParsePattern(StringRef PatternStr, SourceMgr &SM);
86 /// Match - Match the pattern string against the input buffer Buffer. This
87 /// returns the position that is matched or npos if there is no match. If
88 /// there is a match, the size of the matched string is returned in MatchLen.
89 size_t Match(StringRef Buffer, size_t &MatchLen) const;
92 bool Pattern::ParsePattern(StringRef PatternStr, SourceMgr &SM) {
93 // Ignore trailing whitespace.
94 while (!PatternStr.empty() &&
95 (PatternStr.back() == ' ' || PatternStr.back() == '\t'))
96 PatternStr = PatternStr.substr(0, PatternStr.size()-1);
98 // Check that there is something on the line.
99 if (PatternStr.empty()) {
100 SM.PrintMessage(SMLoc::getFromPointer(PatternStr.data()),
101 "found empty check string with prefix '"+CheckPrefix+":'",
106 // Check to see if this is a fixed string, or if it has regex pieces.
107 if (PatternStr.size() < 2 || PatternStr.find("{{") == StringRef::npos) {
108 FixedStr = PatternStr;
112 // Otherwise, there is at least one regex piece.
114 // Scan the pattern to break it into regex and non-regex pieces.
115 while (!PatternStr.empty()) {
116 // Handle fixed string matches.
117 if (PatternStr.size() < 2 ||
118 PatternStr[0] != '{' || PatternStr[1] != '{') {
119 // Find the end, which is the start of the next regex.
120 size_t FixedMatchEnd = PatternStr.find("{{");
122 Chunks.push_back(PatternChunk(PatternStr.substr(0, FixedMatchEnd),false));
123 PatternStr = PatternStr.substr(FixedMatchEnd);
127 // Otherwise, this is the start of a regex match. Scan for the }}.
128 size_t End = PatternStr.find("}}");
129 if (End == StringRef::npos) {
130 SM.PrintMessage(SMLoc::getFromPointer(PatternStr.data()),
131 "found start of regex string with no end '}}'", "error");
135 Regex R(PatternStr.substr(2, End-2));
137 if (!R.isValid(Error)) {
138 SM.PrintMessage(SMLoc::getFromPointer(PatternStr.data()+2),
139 "invalid regex: " + Error, "error");
143 Chunks.push_back(PatternChunk(PatternStr.substr(2, End-2), true));
144 PatternStr = PatternStr.substr(End+2);
150 /// Match - Match the pattern string against the input buffer Buffer. This
151 /// returns the position that is matched or npos if there is no match. If
152 /// there is a match, the size of the matched string is returned in MatchLen.
153 size_t Pattern::Match(StringRef Buffer, size_t &MatchLen) const {
154 // If this is a fixed string pattern, just match it now.
155 if (!FixedStr.empty()) {
156 MatchLen = FixedStr.size();
157 return Buffer.find(FixedStr);
160 size_t FirstMatch = StringRef::npos;
163 while (!Buffer.empty()) {
164 StringRef MatchAttempt = Buffer;
166 unsigned ChunkNo = 0, e = Chunks.size();
167 for (; ChunkNo != e; ++ChunkNo) {
168 size_t ThisMatch, ThisLength = StringRef::npos;
169 ThisMatch = Chunks[ChunkNo].Match(MatchAttempt, ThisLength);
171 // Otherwise, what we do depends on if this is the first match or not. If
172 // this is the first match, it doesn't match to match at the start of
175 // If the first match fails then this pattern will never match in
177 if (ThisMatch == StringRef::npos)
180 FirstMatch = ThisMatch;
181 MatchAttempt = MatchAttempt.substr(FirstMatch);
185 // If this chunk didn't match, then the entire pattern didn't match from
186 // FirstMatch, try later in the buffer.
187 if (ThisMatch == StringRef::npos)
190 // Ok, if the match didn't match at the beginning of MatchAttempt, then we
191 // have something like "ABC{{DEF}} and something was in-between. Reject
196 // Otherwise, match the string and move to the next chunk.
197 MatchLen += ThisLength;
198 MatchAttempt = MatchAttempt.substr(ThisLength);
201 // If the whole thing matched, we win.
205 // Otherwise, try matching again after FirstMatch to see if this pattern
206 // matches later in the buffer.
207 Buffer = Buffer.substr(FirstMatch+1);
210 // If we ran out of stuff to scan, then we didn't match.
211 return StringRef::npos;
215 //===----------------------------------------------------------------------===//
217 //===----------------------------------------------------------------------===//
219 /// CheckString - This is a check that we found in the input file.
221 /// Pat - The pattern to match.
224 /// Loc - The location in the match file that the check string was specified.
227 /// IsCheckNext - This is true if this is a CHECK-NEXT: directive (as opposed
228 /// to a CHECK: directive.
231 /// NotStrings - These are all of the strings that are disallowed from
232 /// occurring between this match string and the previous one (or start of
234 std::vector<std::pair<SMLoc, Pattern> > NotStrings;
236 CheckString(const Pattern &P, SMLoc L, bool isCheckNext)
237 : Pat(P), Loc(L), IsCheckNext(isCheckNext) {}
240 /// CanonicalizeInputFile - Remove duplicate horizontal space from the specified
241 /// memory buffer, free it, and return a new one.
242 static MemoryBuffer *CanonicalizeInputFile(MemoryBuffer *MB) {
243 SmallVector<char, 16> NewFile;
244 NewFile.reserve(MB->getBufferSize());
246 for (const char *Ptr = MB->getBufferStart(), *End = MB->getBufferEnd();
248 // If C is not a horizontal whitespace, skip it.
249 if (*Ptr != ' ' && *Ptr != '\t') {
250 NewFile.push_back(*Ptr);
254 // Otherwise, add one space and advance over neighboring space.
255 NewFile.push_back(' ');
256 while (Ptr+1 != End &&
257 (Ptr[1] == ' ' || Ptr[1] == '\t'))
261 // Free the old buffer and return a new one.
263 MemoryBuffer::getMemBufferCopy(NewFile.data(),
264 NewFile.data() + NewFile.size(),
265 MB->getBufferIdentifier());
272 /// ReadCheckFile - Read the check file, which specifies the sequence of
273 /// expected strings. The strings are added to the CheckStrings vector.
274 static bool ReadCheckFile(SourceMgr &SM,
275 std::vector<CheckString> &CheckStrings) {
276 // Open the check file, and tell SourceMgr about it.
277 std::string ErrorStr;
279 MemoryBuffer::getFileOrSTDIN(CheckFilename.c_str(), &ErrorStr);
281 errs() << "Could not open check file '" << CheckFilename << "': "
286 // If we want to canonicalize whitespace, strip excess whitespace from the
287 // buffer containing the CHECK lines.
288 if (!NoCanonicalizeWhiteSpace)
289 F = CanonicalizeInputFile(F);
291 SM.AddNewSourceBuffer(F, SMLoc());
293 // Find all instances of CheckPrefix followed by : in the file.
294 StringRef Buffer = F->getBuffer();
296 std::vector<std::pair<SMLoc, Pattern> > NotMatches;
299 // See if Prefix occurs in the memory buffer.
300 Buffer = Buffer.substr(Buffer.find(CheckPrefix));
302 // If we didn't find a match, we're done.
306 const char *CheckPrefixStart = Buffer.data();
308 // When we find a check prefix, keep track of whether we find CHECK: or
310 bool IsCheckNext = false, IsCheckNot = false;
312 // Verify that the : is present after the prefix.
313 if (Buffer[CheckPrefix.size()] == ':') {
314 Buffer = Buffer.substr(CheckPrefix.size()+1);
315 } else if (Buffer.size() > CheckPrefix.size()+6 &&
316 memcmp(Buffer.data()+CheckPrefix.size(), "-NEXT:", 6) == 0) {
317 Buffer = Buffer.substr(CheckPrefix.size()+7);
319 } else if (Buffer.size() > CheckPrefix.size()+5 &&
320 memcmp(Buffer.data()+CheckPrefix.size(), "-NOT:", 5) == 0) {
321 Buffer = Buffer.substr(CheckPrefix.size()+6);
324 Buffer = Buffer.substr(1);
328 // Okay, we found the prefix, yay. Remember the rest of the line, but
329 // ignore leading and trailing whitespace.
330 Buffer = Buffer.substr(Buffer.find_first_not_of(" \t"));
332 // Scan ahead to the end of line.
333 size_t EOL = Buffer.find_first_of("\n\r");
335 // Parse the pattern.
337 if (P.ParsePattern(Buffer.substr(0, EOL), SM))
340 Buffer = Buffer.substr(EOL);
343 // Verify that CHECK-NEXT lines have at least one CHECK line before them.
344 if (IsCheckNext && CheckStrings.empty()) {
345 SM.PrintMessage(SMLoc::getFromPointer(CheckPrefixStart),
346 "found '"+CheckPrefix+"-NEXT:' without previous '"+
347 CheckPrefix+ ": line", "error");
353 NotMatches.push_back(std::make_pair(SMLoc::getFromPointer(Buffer.data()),
359 // Okay, add the string we captured to the output vector and move on.
360 CheckStrings.push_back(CheckString(P,
361 SMLoc::getFromPointer(Buffer.data()),
363 std::swap(NotMatches, CheckStrings.back().NotStrings);
366 if (CheckStrings.empty()) {
367 errs() << "error: no check strings found with prefix '" << CheckPrefix
372 if (!NotMatches.empty()) {
373 errs() << "error: '" << CheckPrefix
374 << "-NOT:' not supported after last check line.\n";
381 static void PrintCheckFailed(const SourceMgr &SM, const CheckString &CheckStr,
383 // Otherwise, we have an error, emit an error message.
384 SM.PrintMessage(CheckStr.Loc, "expected string not found in input",
387 // Print the "scanning from here" line. If the current position is at the
388 // end of a line, advance to the start of the next line.
389 Buffer = Buffer.substr(Buffer.find_first_not_of(" \t\n\r"));
391 SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()), "scanning from here",
395 /// CountNumNewlinesBetween - Count the number of newlines in the specified
397 static unsigned CountNumNewlinesBetween(StringRef Range) {
398 unsigned NumNewLines = 0;
401 Range = Range.substr(Range.find_first_of("\n\r"));
402 if (Range.empty()) return NumNewLines;
406 // Handle \n\r and \r\n as a single newline.
407 if (Range.size() > 1 &&
408 (Range[1] == '\n' || Range[1] == '\r') &&
409 (Range[0] != Range[1]))
410 Range = Range.substr(1);
411 Range = Range.substr(1);
415 int main(int argc, char **argv) {
416 sys::PrintStackTraceOnErrorSignal();
417 PrettyStackTraceProgram X(argc, argv);
418 cl::ParseCommandLineOptions(argc, argv);
422 // Read the expected strings from the check file.
423 std::vector<CheckString> CheckStrings;
424 if (ReadCheckFile(SM, CheckStrings))
427 // Open the file to check and add it to SourceMgr.
428 std::string ErrorStr;
430 MemoryBuffer::getFileOrSTDIN(InputFilename.c_str(), &ErrorStr);
432 errs() << "Could not open input file '" << InputFilename << "': "
437 // Remove duplicate spaces in the input file if requested.
438 if (!NoCanonicalizeWhiteSpace)
439 F = CanonicalizeInputFile(F);
441 SM.AddNewSourceBuffer(F, SMLoc());
443 // Check that we have all of the expected strings, in order, in the input
445 StringRef Buffer = F->getBuffer();
447 const char *LastMatch = Buffer.data();
449 for (unsigned StrNo = 0, e = CheckStrings.size(); StrNo != e; ++StrNo) {
450 const CheckString &CheckStr = CheckStrings[StrNo];
452 StringRef SearchFrom = Buffer;
454 // Find StrNo in the file.
456 Buffer = Buffer.substr(CheckStr.Pat.Match(Buffer, MatchLen));
458 // If we didn't find a match, reject the input.
459 if (Buffer.empty()) {
460 PrintCheckFailed(SM, CheckStr, SearchFrom);
464 StringRef SkippedRegion(LastMatch, Buffer.data()-LastMatch);
466 // If this check is a "CHECK-NEXT", verify that the previous match was on
467 // the previous line (i.e. that there is one newline between them).
468 if (CheckStr.IsCheckNext) {
469 // Count the number of newlines between the previous match and this one.
470 assert(LastMatch != F->getBufferStart() &&
471 "CHECK-NEXT can't be the first check in a file");
473 unsigned NumNewLines = CountNumNewlinesBetween(SkippedRegion);
474 if (NumNewLines == 0) {
475 SM.PrintMessage(CheckStr.Loc,
476 CheckPrefix+"-NEXT: is on the same line as previous match",
478 SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()),
479 "'next' match was here", "note");
480 SM.PrintMessage(SMLoc::getFromPointer(LastMatch),
481 "previous match was here", "note");
485 if (NumNewLines != 1) {
486 SM.PrintMessage(CheckStr.Loc,
488 "-NEXT: is not on the line after the previous match",
490 SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()),
491 "'next' match was here", "note");
492 SM.PrintMessage(SMLoc::getFromPointer(LastMatch),
493 "previous match was here", "note");
498 // If this match had "not strings", verify that they don't exist in the
500 for (unsigned ChunkNo = 0, e = CheckStr.NotStrings.size(); ChunkNo != e; ++ChunkNo) {
502 size_t Pos = CheckStr.NotStrings[ChunkNo].second.Match(SkippedRegion, MatchLen);
503 if (Pos == StringRef::npos) continue;
505 SM.PrintMessage(SMLoc::getFromPointer(LastMatch+Pos),
506 CheckPrefix+"-NOT: string occurred!", "error");
507 SM.PrintMessage(CheckStr.NotStrings[ChunkNo].first,
508 CheckPrefix+"-NOT: pattern specified here", "note");
513 // Otherwise, everything is good. Step over the matched text and remember
514 // the position after the match as the end of the last match.
515 Buffer = Buffer.substr(MatchLen);
516 LastMatch = Buffer.data();