blob: 078028a115cf27a0ea31e4f288c05c98a131519c [file] [log] [blame]
Chris Lattner81cb8ca2009-07-08 18:44:05 +00001//===- FileCheck.cpp - Check that File's Contents match what is expected --===//
2//
3// The LLVM Compiler Infrastructure
4//
5// This file is distributed under the University of Illinois Open Source
6// License. See LICENSE.TXT for details.
7//
8//===----------------------------------------------------------------------===//
9//
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.
12//
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.
16//
17//===----------------------------------------------------------------------===//
18
19#include "llvm/Support/CommandLine.h"
20#include "llvm/Support/MemoryBuffer.h"
21#include "llvm/Support/PrettyStackTrace.h"
Chris Lattner52870082009-09-24 21:47:32 +000022#include "llvm/Support/Regex.h"
Chris Lattner81cb8ca2009-07-08 18:44:05 +000023#include "llvm/Support/SourceMgr.h"
24#include "llvm/Support/raw_ostream.h"
25#include "llvm/System/Signals.h"
Daniel Dunbarfafe93c2009-11-22 22:08:06 +000026#include "llvm/ADT/SmallString.h"
Chris Lattnereec96952009-09-27 07:56:52 +000027#include "llvm/ADT/StringMap.h"
28#include <algorithm>
Chris Lattner81cb8ca2009-07-08 18:44:05 +000029using namespace llvm;
30
31static cl::opt<std::string>
32CheckFilename(cl::Positional, cl::desc("<check-file>"), cl::Required);
33
34static cl::opt<std::string>
35InputFilename("input-file", cl::desc("File to check (defaults to stdin)"),
36 cl::init("-"), cl::value_desc("filename"));
37
38static cl::opt<std::string>
39CheckPrefix("check-prefix", cl::init("CHECK"),
40 cl::desc("Prefix to use from check file (defaults to 'CHECK')"));
41
Chris Lattner88a7e9e2009-07-11 18:58:15 +000042static cl::opt<bool>
43NoCanonicalizeWhiteSpace("strict-whitespace",
44 cl::desc("Do not treat all horizontal whitespace as equivalent"));
45
Chris Lattnera29703e2009-09-24 20:39:13 +000046//===----------------------------------------------------------------------===//
47// Pattern Handling Code.
48//===----------------------------------------------------------------------===//
49
Chris Lattner9fc66782009-09-24 20:25:55 +000050class Pattern {
Chris Lattner94638f02009-09-25 17:29:36 +000051 SMLoc PatternLoc;
52
Chris Lattner5d6a05f2009-09-25 17:23:43 +000053 /// FixedStr - If non-empty, this pattern is a fixed string match with the
54 /// specified fixed string.
Chris Lattner2702e6a2009-09-25 17:09:12 +000055 StringRef FixedStr;
Chris Lattner5d6a05f2009-09-25 17:23:43 +000056
57 /// RegEx - If non-empty, this is a regex pattern.
58 std::string RegExStr;
Chris Lattnereec96952009-09-27 07:56:52 +000059
60 /// VariableUses - Entries in this vector map to uses of a variable in the
61 /// pattern, e.g. "foo[[bar]]baz". In this case, the RegExStr will contain
62 /// "foobaz" and we'll get an entry in this vector that tells us to insert the
63 /// value of bar at offset 3.
64 std::vector<std::pair<StringRef, unsigned> > VariableUses;
65
66 /// VariableDefs - Entries in this vector map to definitions of a variable in
67 /// the pattern, e.g. "foo[[bar:.*]]baz". In this case, the RegExStr will
68 /// contain "foo(.*)baz" and VariableDefs will contain the pair "bar",1. The
69 /// index indicates what parenthesized value captures the variable value.
70 std::vector<std::pair<StringRef, unsigned> > VariableDefs;
71
Chris Lattner9fc66782009-09-24 20:25:55 +000072public:
73
Chris Lattnera29703e2009-09-24 20:39:13 +000074 Pattern() { }
75
76 bool ParsePattern(StringRef PatternStr, SourceMgr &SM);
Chris Lattner9fc66782009-09-24 20:25:55 +000077
78 /// Match - Match the pattern string against the input buffer Buffer. This
79 /// returns the position that is matched or npos if there is no match. If
80 /// there is a match, the size of the matched string is returned in MatchLen.
Chris Lattnereec96952009-09-27 07:56:52 +000081 ///
82 /// The VariableTable StringMap provides the current values of filecheck
83 /// variables and is updated if this match defines new values.
84 size_t Match(StringRef Buffer, size_t &MatchLen,
85 StringMap<StringRef> &VariableTable) const;
Daniel Dunbarfafe93c2009-11-22 22:08:06 +000086
87 /// PrintFailureInfo - Print additional information about a failure to match
88 /// involving this pattern.
89 void PrintFailureInfo(const SourceMgr &SM, StringRef Buffer,
90 const StringMap<StringRef> &VariableTable) const;
91
Chris Lattner5d6a05f2009-09-25 17:23:43 +000092private:
Chris Lattnereec96952009-09-27 07:56:52 +000093 static void AddFixedStringToRegEx(StringRef FixedStr, std::string &TheStr);
94 bool AddRegExToRegEx(StringRef RegExStr, unsigned &CurParen, SourceMgr &SM);
Daniel Dunbaread2dac2009-11-22 22:59:26 +000095
96 /// ComputeMatchDistance - Compute an arbitrary estimate for the quality of
97 /// matching this pattern at the start of \arg Buffer; a distance of zero
98 /// should correspond to a perfect match.
99 unsigned ComputeMatchDistance(StringRef Buffer,
100 const StringMap<StringRef> &VariableTable) const;
Chris Lattner9fc66782009-09-24 20:25:55 +0000101};
102
Chris Lattnereec96952009-09-27 07:56:52 +0000103
Chris Lattnera29703e2009-09-24 20:39:13 +0000104bool Pattern::ParsePattern(StringRef PatternStr, SourceMgr &SM) {
Chris Lattner94638f02009-09-25 17:29:36 +0000105 PatternLoc = SMLoc::getFromPointer(PatternStr.data());
106
Chris Lattnera29703e2009-09-24 20:39:13 +0000107 // Ignore trailing whitespace.
108 while (!PatternStr.empty() &&
109 (PatternStr.back() == ' ' || PatternStr.back() == '\t'))
110 PatternStr = PatternStr.substr(0, PatternStr.size()-1);
111
112 // Check that there is something on the line.
113 if (PatternStr.empty()) {
Chris Lattner94638f02009-09-25 17:29:36 +0000114 SM.PrintMessage(PatternLoc, "found empty check string with prefix '" +
115 CheckPrefix+":'", "error");
Chris Lattnera29703e2009-09-24 20:39:13 +0000116 return true;
117 }
Chris Lattner52870082009-09-24 21:47:32 +0000118
Chris Lattner2702e6a2009-09-25 17:09:12 +0000119 // Check to see if this is a fixed string, or if it has regex pieces.
Chris Lattnereec96952009-09-27 07:56:52 +0000120 if (PatternStr.size() < 2 ||
121 (PatternStr.find("{{") == StringRef::npos &&
122 PatternStr.find("[[") == StringRef::npos)) {
Chris Lattner2702e6a2009-09-25 17:09:12 +0000123 FixedStr = PatternStr;
124 return false;
125 }
126
Chris Lattnereec96952009-09-27 07:56:52 +0000127 // Paren value #0 is for the fully matched string. Any new parenthesized
128 // values add from their.
129 unsigned CurParen = 1;
130
Chris Lattner5d6a05f2009-09-25 17:23:43 +0000131 // Otherwise, there is at least one regex piece. Build up the regex pattern
132 // by escaping scary characters in fixed strings, building up one big regex.
Chris Lattner52870082009-09-24 21:47:32 +0000133 while (!PatternStr.empty()) {
Chris Lattnereec96952009-09-27 07:56:52 +0000134 // RegEx matches.
135 if (PatternStr.size() >= 2 &&
136 PatternStr[0] == '{' && PatternStr[1] == '{') {
137
138 // Otherwise, this is the start of a regex match. Scan for the }}.
139 size_t End = PatternStr.find("}}");
140 if (End == StringRef::npos) {
141 SM.PrintMessage(SMLoc::getFromPointer(PatternStr.data()),
142 "found start of regex string with no end '}}'", "error");
143 return true;
144 }
145
146 if (AddRegExToRegEx(PatternStr.substr(2, End-2), CurParen, SM))
147 return true;
148 PatternStr = PatternStr.substr(End+2);
Chris Lattner52870082009-09-24 21:47:32 +0000149 continue;
150 }
151
Chris Lattnereec96952009-09-27 07:56:52 +0000152 // Named RegEx matches. These are of two forms: [[foo:.*]] which matches .*
153 // (or some other regex) and assigns it to the FileCheck variable 'foo'. The
154 // second form is [[foo]] which is a reference to foo. The variable name
Daniel Dunbar964ac012009-11-22 22:07:50 +0000155 // itself must be of the form "[a-zA-Z_][0-9a-zA-Z_]*", otherwise we reject
Chris Lattnereec96952009-09-27 07:56:52 +0000156 // it. This is to catch some common errors.
157 if (PatternStr.size() >= 2 &&
158 PatternStr[0] == '[' && PatternStr[1] == '[') {
159 // Verify that it is terminated properly.
160 size_t End = PatternStr.find("]]");
161 if (End == StringRef::npos) {
162 SM.PrintMessage(SMLoc::getFromPointer(PatternStr.data()),
163 "invalid named regex reference, no ]] found", "error");
164 return true;
165 }
166
167 StringRef MatchStr = PatternStr.substr(2, End-2);
168 PatternStr = PatternStr.substr(End+2);
169
170 // Get the regex name (e.g. "foo").
171 size_t NameEnd = MatchStr.find(':');
172 StringRef Name = MatchStr.substr(0, NameEnd);
173
174 if (Name.empty()) {
175 SM.PrintMessage(SMLoc::getFromPointer(Name.data()),
176 "invalid name in named regex: empty name", "error");
177 return true;
178 }
179
180 // Verify that the name is well formed.
181 for (unsigned i = 0, e = Name.size(); i != e; ++i)
Daniel Dunbar964ac012009-11-22 22:07:50 +0000182 if (Name[i] != '_' &&
183 (Name[i] < 'a' || Name[i] > 'z') &&
Chris Lattnereec96952009-09-27 07:56:52 +0000184 (Name[i] < 'A' || Name[i] > 'Z') &&
185 (Name[i] < '0' || Name[i] > '9')) {
186 SM.PrintMessage(SMLoc::getFromPointer(Name.data()+i),
187 "invalid name in named regex", "error");
188 return true;
189 }
190
191 // Name can't start with a digit.
192 if (isdigit(Name[0])) {
193 SM.PrintMessage(SMLoc::getFromPointer(Name.data()),
194 "invalid name in named regex", "error");
195 return true;
196 }
197
198 // Handle [[foo]].
199 if (NameEnd == StringRef::npos) {
200 VariableUses.push_back(std::make_pair(Name, RegExStr.size()));
201 continue;
202 }
203
204 // Handle [[foo:.*]].
205 VariableDefs.push_back(std::make_pair(Name, CurParen));
206 RegExStr += '(';
207 ++CurParen;
208
209 if (AddRegExToRegEx(MatchStr.substr(NameEnd+1), CurParen, SM))
210 return true;
211
212 RegExStr += ')';
Chris Lattner52870082009-09-24 21:47:32 +0000213 }
214
Chris Lattnereec96952009-09-27 07:56:52 +0000215 // Handle fixed string matches.
216 // Find the end, which is the start of the next regex.
217 size_t FixedMatchEnd = PatternStr.find("{{");
218 FixedMatchEnd = std::min(FixedMatchEnd, PatternStr.find("[["));
219 AddFixedStringToRegEx(PatternStr.substr(0, FixedMatchEnd), RegExStr);
220 PatternStr = PatternStr.substr(FixedMatchEnd);
221 continue;
Chris Lattner52870082009-09-24 21:47:32 +0000222 }
Chris Lattneradea46e2009-09-24 20:45:07 +0000223
Chris Lattnera29703e2009-09-24 20:39:13 +0000224 return false;
225}
226
Chris Lattnereec96952009-09-27 07:56:52 +0000227void Pattern::AddFixedStringToRegEx(StringRef FixedStr, std::string &TheStr) {
Chris Lattner5d6a05f2009-09-25 17:23:43 +0000228 // Add the characters from FixedStr to the regex, escaping as needed. This
229 // avoids "leaning toothpicks" in common patterns.
230 for (unsigned i = 0, e = FixedStr.size(); i != e; ++i) {
231 switch (FixedStr[i]) {
232 // These are the special characters matched in "p_ere_exp".
233 case '(':
234 case ')':
235 case '^':
236 case '$':
237 case '|':
238 case '*':
239 case '+':
240 case '?':
241 case '.':
242 case '[':
243 case '\\':
244 case '{':
Chris Lattnereec96952009-09-27 07:56:52 +0000245 TheStr += '\\';
Chris Lattner5d6a05f2009-09-25 17:23:43 +0000246 // FALL THROUGH.
247 default:
Chris Lattnereec96952009-09-27 07:56:52 +0000248 TheStr += FixedStr[i];
Chris Lattner5d6a05f2009-09-25 17:23:43 +0000249 break;
250 }
251 }
252}
253
Chris Lattnereec96952009-09-27 07:56:52 +0000254bool Pattern::AddRegExToRegEx(StringRef RegexStr, unsigned &CurParen,
255 SourceMgr &SM) {
256 Regex R(RegexStr);
257 std::string Error;
258 if (!R.isValid(Error)) {
259 SM.PrintMessage(SMLoc::getFromPointer(RegexStr.data()),
260 "invalid regex: " + Error, "error");
261 return true;
262 }
263
264 RegExStr += RegexStr.str();
265 CurParen += R.getNumMatches();
266 return false;
267}
Chris Lattner5d6a05f2009-09-25 17:23:43 +0000268
Chris Lattner52870082009-09-24 21:47:32 +0000269/// Match - Match the pattern string against the input buffer Buffer. This
270/// returns the position that is matched or npos if there is no match. If
271/// there is a match, the size of the matched string is returned in MatchLen.
Chris Lattnereec96952009-09-27 07:56:52 +0000272size_t Pattern::Match(StringRef Buffer, size_t &MatchLen,
273 StringMap<StringRef> &VariableTable) const {
Chris Lattner2702e6a2009-09-25 17:09:12 +0000274 // If this is a fixed string pattern, just match it now.
275 if (!FixedStr.empty()) {
276 MatchLen = FixedStr.size();
277 return Buffer.find(FixedStr);
278 }
Chris Lattnereec96952009-09-27 07:56:52 +0000279
Chris Lattner5d6a05f2009-09-25 17:23:43 +0000280 // Regex match.
Chris Lattnereec96952009-09-27 07:56:52 +0000281
282 // If there are variable uses, we need to create a temporary string with the
283 // actual value.
284 StringRef RegExToMatch = RegExStr;
285 std::string TmpStr;
286 if (!VariableUses.empty()) {
287 TmpStr = RegExStr;
288
289 unsigned InsertOffset = 0;
290 for (unsigned i = 0, e = VariableUses.size(); i != e; ++i) {
Daniel Dunbarfafe93c2009-11-22 22:08:06 +0000291 StringMap<StringRef>::iterator it =
292 VariableTable.find(VariableUses[i].first);
293 // If the variable is undefined, return an error.
294 if (it == VariableTable.end())
295 return StringRef::npos;
296
Chris Lattnereec96952009-09-27 07:56:52 +0000297 // Look up the value and escape it so that we can plop it into the regex.
298 std::string Value;
Daniel Dunbarfafe93c2009-11-22 22:08:06 +0000299 AddFixedStringToRegEx(it->second, Value);
Chris Lattnereec96952009-09-27 07:56:52 +0000300
301 // Plop it into the regex at the adjusted offset.
302 TmpStr.insert(TmpStr.begin()+VariableUses[i].second+InsertOffset,
303 Value.begin(), Value.end());
304 InsertOffset += Value.size();
305 }
306
307 // Match the newly constructed regex.
308 RegExToMatch = TmpStr;
309 }
310
311
Chris Lattner5d6a05f2009-09-25 17:23:43 +0000312 SmallVector<StringRef, 4> MatchInfo;
Chris Lattnereec96952009-09-27 07:56:52 +0000313 if (!Regex(RegExToMatch, Regex::Newline).match(Buffer, &MatchInfo))
Chris Lattner5d6a05f2009-09-25 17:23:43 +0000314 return StringRef::npos;
Chris Lattner52870082009-09-24 21:47:32 +0000315
Chris Lattner5d6a05f2009-09-25 17:23:43 +0000316 // Successful regex match.
317 assert(!MatchInfo.empty() && "Didn't get any match");
318 StringRef FullMatch = MatchInfo[0];
Chris Lattner52870082009-09-24 21:47:32 +0000319
Chris Lattnereec96952009-09-27 07:56:52 +0000320 // If this defines any variables, remember their values.
321 for (unsigned i = 0, e = VariableDefs.size(); i != e; ++i) {
322 assert(VariableDefs[i].second < MatchInfo.size() &&
323 "Internal paren error");
324 VariableTable[VariableDefs[i].first] = MatchInfo[VariableDefs[i].second];
Chris Lattner94638f02009-09-25 17:29:36 +0000325 }
Chris Lattner94638f02009-09-25 17:29:36 +0000326
Chris Lattner5d6a05f2009-09-25 17:23:43 +0000327 MatchLen = FullMatch.size();
328 return FullMatch.data()-Buffer.data();
Chris Lattner52870082009-09-24 21:47:32 +0000329}
330
Daniel Dunbaread2dac2009-11-22 22:59:26 +0000331unsigned Pattern::ComputeMatchDistance(StringRef Buffer,
332 const StringMap<StringRef> &VariableTable) const {
333 // Just compute the number of matching characters. For regular expressions, we
334 // just compare against the regex itself and hope for the best.
335 //
336 // FIXME: One easy improvement here is have the regex lib generate a single
337 // example regular expression which matches, and use that as the example
338 // string.
339 StringRef ExampleString(FixedStr);
340 if (ExampleString.empty())
341 ExampleString = RegExStr;
342
343 unsigned Distance = 0;
344 for (unsigned i = 0, e = ExampleString.size(); i != e; ++i)
345 if (Buffer.substr(i, 1) != ExampleString.substr(i, 1))
346 ++Distance;
347
348 return Distance;
349}
350
Daniel Dunbarfafe93c2009-11-22 22:08:06 +0000351void Pattern::PrintFailureInfo(const SourceMgr &SM, StringRef Buffer,
352 const StringMap<StringRef> &VariableTable) const{
Daniel Dunbarfafe93c2009-11-22 22:08:06 +0000353 // If this was a regular expression using variables, print the current
354 // variable values.
355 if (!VariableUses.empty()) {
356 for (unsigned i = 0, e = VariableUses.size(); i != e; ++i) {
357 StringRef Var = VariableUses[i].first;
358 StringMap<StringRef>::const_iterator it = VariableTable.find(Var);
359 SmallString<256> Msg;
360 raw_svector_ostream OS(Msg);
361
362 // Check for undefined variable references.
363 if (it == VariableTable.end()) {
364 OS << "uses undefined variable \"";
365 OS.write_escaped(Var) << "\"";;
366 } else {
367 OS << "with variable \"";
368 OS.write_escaped(Var) << "\" equal to \"";
369 OS.write_escaped(it->second) << "\"";
370 }
371
372 SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()), OS.str(), "note",
373 /*ShowLine=*/false);
374 }
375 }
Daniel Dunbaread2dac2009-11-22 22:59:26 +0000376
377 // Attempt to find the closest/best fuzzy match. Usually an error happens
378 // because some string in the output didn't exactly match. In these cases, we
379 // would like to show the user a best guess at what "should have" matched, to
380 // save them having to actually check the input manually.
381 size_t NumLinesForward = 0;
382 size_t Best = StringRef::npos;
383 double BestQuality = 0;
384
385 // Use an arbitrary 4k limit on how far we will search.
386 for (size_t i = 0, e = std::min(4096, int(Buffer.size())); i != e; ++i) {
387 if (Buffer[i] == '\n')
388 ++NumLinesForward;
389
390 // Compute the "quality" of this match as an arbitrary combination of the
391 // match distance and the number of lines skipped to get to this match.
392 unsigned Distance = ComputeMatchDistance(Buffer.substr(i), VariableTable);
393 double Quality = Distance + (NumLinesForward / 100.);
394
395 if (Quality < BestQuality || Best == StringRef::npos) {
396 Best = i;
397 BestQuality = Quality;
398 }
399 }
400
Daniel Dunbara76a7882009-11-29 08:30:24 +0000401 if (Best != StringRef::npos && BestQuality < 50) {
Daniel Dunbaread2dac2009-11-22 22:59:26 +0000402 // Print the "possible intended match here" line if we found something
403 // reasonable.
404 SM.PrintMessage(SMLoc::getFromPointer(Buffer.data() + Best),
405 "possible intended match here", "note");
406
407 // FIXME: If we wanted to be really friendly we would show why the match
408 // failed, as it can be hard to spot simple one character differences.
409 }
Daniel Dunbarfafe93c2009-11-22 22:08:06 +0000410}
Chris Lattnera29703e2009-09-24 20:39:13 +0000411
412//===----------------------------------------------------------------------===//
413// Check Strings.
414//===----------------------------------------------------------------------===//
Chris Lattner9fc66782009-09-24 20:25:55 +0000415
416/// CheckString - This is a check that we found in the input file.
417struct CheckString {
418 /// Pat - The pattern to match.
419 Pattern Pat;
Chris Lattner207e1bc2009-08-15 17:41:04 +0000420
421 /// Loc - The location in the match file that the check string was specified.
422 SMLoc Loc;
423
Chris Lattner5dafafd2009-08-15 18:32:21 +0000424 /// IsCheckNext - This is true if this is a CHECK-NEXT: directive (as opposed
425 /// to a CHECK: directive.
426 bool IsCheckNext;
427
Chris Lattnerf15380b2009-09-20 22:35:26 +0000428 /// NotStrings - These are all of the strings that are disallowed from
429 /// occurring between this match string and the previous one (or start of
430 /// file).
Chris Lattnera29703e2009-09-24 20:39:13 +0000431 std::vector<std::pair<SMLoc, Pattern> > NotStrings;
Chris Lattnerf15380b2009-09-20 22:35:26 +0000432
Chris Lattner9fc66782009-09-24 20:25:55 +0000433 CheckString(const Pattern &P, SMLoc L, bool isCheckNext)
434 : Pat(P), Loc(L), IsCheckNext(isCheckNext) {}
Chris Lattner207e1bc2009-08-15 17:41:04 +0000435};
436
Chris Lattneradea46e2009-09-24 20:45:07 +0000437/// CanonicalizeInputFile - Remove duplicate horizontal space from the specified
438/// memory buffer, free it, and return a new one.
439static MemoryBuffer *CanonicalizeInputFile(MemoryBuffer *MB) {
440 SmallVector<char, 16> NewFile;
441 NewFile.reserve(MB->getBufferSize());
442
443 for (const char *Ptr = MB->getBufferStart(), *End = MB->getBufferEnd();
444 Ptr != End; ++Ptr) {
445 // If C is not a horizontal whitespace, skip it.
446 if (*Ptr != ' ' && *Ptr != '\t') {
447 NewFile.push_back(*Ptr);
448 continue;
449 }
450
451 // Otherwise, add one space and advance over neighboring space.
452 NewFile.push_back(' ');
453 while (Ptr+1 != End &&
454 (Ptr[1] == ' ' || Ptr[1] == '\t'))
455 ++Ptr;
456 }
457
458 // Free the old buffer and return a new one.
459 MemoryBuffer *MB2 =
460 MemoryBuffer::getMemBufferCopy(NewFile.data(),
461 NewFile.data() + NewFile.size(),
462 MB->getBufferIdentifier());
463
464 delete MB;
465 return MB2;
466}
467
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000468
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000469/// ReadCheckFile - Read the check file, which specifies the sequence of
470/// expected strings. The strings are added to the CheckStrings vector.
471static bool ReadCheckFile(SourceMgr &SM,
Chris Lattner207e1bc2009-08-15 17:41:04 +0000472 std::vector<CheckString> &CheckStrings) {
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000473 // Open the check file, and tell SourceMgr about it.
474 std::string ErrorStr;
475 MemoryBuffer *F =
476 MemoryBuffer::getFileOrSTDIN(CheckFilename.c_str(), &ErrorStr);
477 if (F == 0) {
478 errs() << "Could not open check file '" << CheckFilename << "': "
479 << ErrorStr << '\n';
480 return true;
481 }
Chris Lattneradea46e2009-09-24 20:45:07 +0000482
483 // If we want to canonicalize whitespace, strip excess whitespace from the
484 // buffer containing the CHECK lines.
485 if (!NoCanonicalizeWhiteSpace)
486 F = CanonicalizeInputFile(F);
487
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000488 SM.AddNewSourceBuffer(F, SMLoc());
489
Chris Lattnerd7e25052009-08-15 18:00:42 +0000490 // Find all instances of CheckPrefix followed by : in the file.
Chris Lattner96077032009-09-20 22:11:44 +0000491 StringRef Buffer = F->getBuffer();
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000492
Chris Lattnera29703e2009-09-24 20:39:13 +0000493 std::vector<std::pair<SMLoc, Pattern> > NotMatches;
Chris Lattnerf15380b2009-09-20 22:35:26 +0000494
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000495 while (1) {
496 // See if Prefix occurs in the memory buffer.
Chris Lattner96077032009-09-20 22:11:44 +0000497 Buffer = Buffer.substr(Buffer.find(CheckPrefix));
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000498
499 // If we didn't find a match, we're done.
Chris Lattner96077032009-09-20 22:11:44 +0000500 if (Buffer.empty())
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000501 break;
502
Chris Lattner96077032009-09-20 22:11:44 +0000503 const char *CheckPrefixStart = Buffer.data();
Chris Lattner5dafafd2009-08-15 18:32:21 +0000504
505 // When we find a check prefix, keep track of whether we find CHECK: or
506 // CHECK-NEXT:
Chris Lattnerf15380b2009-09-20 22:35:26 +0000507 bool IsCheckNext = false, IsCheckNot = false;
Chris Lattner5dafafd2009-08-15 18:32:21 +0000508
Chris Lattnerd7e25052009-08-15 18:00:42 +0000509 // Verify that the : is present after the prefix.
Chris Lattner96077032009-09-20 22:11:44 +0000510 if (Buffer[CheckPrefix.size()] == ':') {
511 Buffer = Buffer.substr(CheckPrefix.size()+1);
Chris Lattner96077032009-09-20 22:11:44 +0000512 } else if (Buffer.size() > CheckPrefix.size()+6 &&
513 memcmp(Buffer.data()+CheckPrefix.size(), "-NEXT:", 6) == 0) {
514 Buffer = Buffer.substr(CheckPrefix.size()+7);
Chris Lattner5dafafd2009-08-15 18:32:21 +0000515 IsCheckNext = true;
Chris Lattnerf15380b2009-09-20 22:35:26 +0000516 } else if (Buffer.size() > CheckPrefix.size()+5 &&
517 memcmp(Buffer.data()+CheckPrefix.size(), "-NOT:", 5) == 0) {
518 Buffer = Buffer.substr(CheckPrefix.size()+6);
519 IsCheckNot = true;
Chris Lattner5dafafd2009-08-15 18:32:21 +0000520 } else {
Chris Lattner96077032009-09-20 22:11:44 +0000521 Buffer = Buffer.substr(1);
Chris Lattnerd7e25052009-08-15 18:00:42 +0000522 continue;
523 }
524
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000525 // Okay, we found the prefix, yay. Remember the rest of the line, but
526 // ignore leading and trailing whitespace.
Chris Lattnerf15380b2009-09-20 22:35:26 +0000527 Buffer = Buffer.substr(Buffer.find_first_not_of(" \t"));
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000528
529 // Scan ahead to the end of line.
Chris Lattner96077032009-09-20 22:11:44 +0000530 size_t EOL = Buffer.find_first_of("\n\r");
Chris Lattnera29703e2009-09-24 20:39:13 +0000531
532 // Parse the pattern.
533 Pattern P;
534 if (P.ParsePattern(Buffer.substr(0, EOL), SM))
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000535 return true;
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000536
Chris Lattnera29703e2009-09-24 20:39:13 +0000537 Buffer = Buffer.substr(EOL);
538
Chris Lattnerf15380b2009-09-20 22:35:26 +0000539
Chris Lattner5dafafd2009-08-15 18:32:21 +0000540 // Verify that CHECK-NEXT lines have at least one CHECK line before them.
541 if (IsCheckNext && CheckStrings.empty()) {
542 SM.PrintMessage(SMLoc::getFromPointer(CheckPrefixStart),
543 "found '"+CheckPrefix+"-NEXT:' without previous '"+
544 CheckPrefix+ ": line", "error");
545 return true;
546 }
547
Chris Lattnera29703e2009-09-24 20:39:13 +0000548 // Handle CHECK-NOT.
549 if (IsCheckNot) {
550 NotMatches.push_back(std::make_pair(SMLoc::getFromPointer(Buffer.data()),
551 P));
552 continue;
553 }
554
Chris Lattner9fc66782009-09-24 20:25:55 +0000555
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000556 // Okay, add the string we captured to the output vector and move on.
Chris Lattner9fc66782009-09-24 20:25:55 +0000557 CheckStrings.push_back(CheckString(P,
Chris Lattner96077032009-09-20 22:11:44 +0000558 SMLoc::getFromPointer(Buffer.data()),
Chris Lattner5dafafd2009-08-15 18:32:21 +0000559 IsCheckNext));
Chris Lattnerf15380b2009-09-20 22:35:26 +0000560 std::swap(NotMatches, CheckStrings.back().NotStrings);
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000561 }
562
563 if (CheckStrings.empty()) {
Chris Lattnerd7e25052009-08-15 18:00:42 +0000564 errs() << "error: no check strings found with prefix '" << CheckPrefix
565 << ":'\n";
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000566 return true;
567 }
568
Chris Lattnerf15380b2009-09-20 22:35:26 +0000569 if (!NotMatches.empty()) {
570 errs() << "error: '" << CheckPrefix
571 << "-NOT:' not supported after last check line.\n";
572 return true;
573 }
574
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000575 return false;
576}
577
Chris Lattner5dafafd2009-08-15 18:32:21 +0000578static void PrintCheckFailed(const SourceMgr &SM, const CheckString &CheckStr,
Daniel Dunbarfafe93c2009-11-22 22:08:06 +0000579 StringRef Buffer,
580 StringMap<StringRef> &VariableTable) {
Chris Lattner5dafafd2009-08-15 18:32:21 +0000581 // Otherwise, we have an error, emit an error message.
582 SM.PrintMessage(CheckStr.Loc, "expected string not found in input",
583 "error");
584
585 // Print the "scanning from here" line. If the current position is at the
586 // end of a line, advance to the start of the next line.
Chris Lattner96077032009-09-20 22:11:44 +0000587 Buffer = Buffer.substr(Buffer.find_first_not_of(" \t\n\r"));
Chris Lattner5dafafd2009-08-15 18:32:21 +0000588
Chris Lattner96077032009-09-20 22:11:44 +0000589 SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()), "scanning from here",
Chris Lattner5dafafd2009-08-15 18:32:21 +0000590 "note");
Daniel Dunbarfafe93c2009-11-22 22:08:06 +0000591
592 // Allow the pattern to print additional information if desired.
593 CheckStr.Pat.PrintFailureInfo(SM, Buffer, VariableTable);
Chris Lattner5dafafd2009-08-15 18:32:21 +0000594}
595
Chris Lattner3711b7a2009-09-20 22:42:44 +0000596/// CountNumNewlinesBetween - Count the number of newlines in the specified
597/// range.
598static unsigned CountNumNewlinesBetween(StringRef Range) {
Chris Lattner5dafafd2009-08-15 18:32:21 +0000599 unsigned NumNewLines = 0;
Chris Lattner3711b7a2009-09-20 22:42:44 +0000600 while (1) {
Chris Lattner5dafafd2009-08-15 18:32:21 +0000601 // Scan for newline.
Chris Lattner3711b7a2009-09-20 22:42:44 +0000602 Range = Range.substr(Range.find_first_of("\n\r"));
603 if (Range.empty()) return NumNewLines;
Chris Lattner5dafafd2009-08-15 18:32:21 +0000604
605 ++NumNewLines;
606
607 // Handle \n\r and \r\n as a single newline.
Chris Lattner3711b7a2009-09-20 22:42:44 +0000608 if (Range.size() > 1 &&
609 (Range[1] == '\n' || Range[1] == '\r') &&
610 (Range[0] != Range[1]))
611 Range = Range.substr(1);
612 Range = Range.substr(1);
Chris Lattner5dafafd2009-08-15 18:32:21 +0000613 }
Chris Lattner5dafafd2009-08-15 18:32:21 +0000614}
615
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000616int main(int argc, char **argv) {
617 sys::PrintStackTraceOnErrorSignal();
618 PrettyStackTraceProgram X(argc, argv);
619 cl::ParseCommandLineOptions(argc, argv);
620
621 SourceMgr SM;
622
623 // Read the expected strings from the check file.
Chris Lattner207e1bc2009-08-15 17:41:04 +0000624 std::vector<CheckString> CheckStrings;
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000625 if (ReadCheckFile(SM, CheckStrings))
626 return 2;
627
628 // Open the file to check and add it to SourceMgr.
629 std::string ErrorStr;
630 MemoryBuffer *F =
631 MemoryBuffer::getFileOrSTDIN(InputFilename.c_str(), &ErrorStr);
632 if (F == 0) {
633 errs() << "Could not open input file '" << InputFilename << "': "
634 << ErrorStr << '\n';
635 return true;
636 }
Chris Lattner88a7e9e2009-07-11 18:58:15 +0000637
638 // Remove duplicate spaces in the input file if requested.
639 if (!NoCanonicalizeWhiteSpace)
640 F = CanonicalizeInputFile(F);
641
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000642 SM.AddNewSourceBuffer(F, SMLoc());
643
Chris Lattnereec96952009-09-27 07:56:52 +0000644 /// VariableTable - This holds all the current filecheck variables.
645 StringMap<StringRef> VariableTable;
646
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000647 // Check that we have all of the expected strings, in order, in the input
648 // file.
Chris Lattner96077032009-09-20 22:11:44 +0000649 StringRef Buffer = F->getBuffer();
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000650
Chris Lattnerf15380b2009-09-20 22:35:26 +0000651 const char *LastMatch = Buffer.data();
652
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000653 for (unsigned StrNo = 0, e = CheckStrings.size(); StrNo != e; ++StrNo) {
Chris Lattner207e1bc2009-08-15 17:41:04 +0000654 const CheckString &CheckStr = CheckStrings[StrNo];
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000655
Chris Lattner96077032009-09-20 22:11:44 +0000656 StringRef SearchFrom = Buffer;
657
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000658 // Find StrNo in the file.
Chris Lattner9fc66782009-09-24 20:25:55 +0000659 size_t MatchLen = 0;
Chris Lattnereec96952009-09-27 07:56:52 +0000660 Buffer = Buffer.substr(CheckStr.Pat.Match(Buffer, MatchLen, VariableTable));
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000661
Chris Lattner5dafafd2009-08-15 18:32:21 +0000662 // If we didn't find a match, reject the input.
Chris Lattner96077032009-09-20 22:11:44 +0000663 if (Buffer.empty()) {
Daniel Dunbarfafe93c2009-11-22 22:08:06 +0000664 PrintCheckFailed(SM, CheckStr, SearchFrom, VariableTable);
Chris Lattner5dafafd2009-08-15 18:32:21 +0000665 return 1;
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000666 }
Chris Lattner3711b7a2009-09-20 22:42:44 +0000667
668 StringRef SkippedRegion(LastMatch, Buffer.data()-LastMatch);
669
Chris Lattner5dafafd2009-08-15 18:32:21 +0000670 // If this check is a "CHECK-NEXT", verify that the previous match was on
671 // the previous line (i.e. that there is one newline between them).
672 if (CheckStr.IsCheckNext) {
673 // Count the number of newlines between the previous match and this one.
Chris Lattnerf15380b2009-09-20 22:35:26 +0000674 assert(LastMatch != F->getBufferStart() &&
675 "CHECK-NEXT can't be the first check in a file");
Chris Lattner5dafafd2009-08-15 18:32:21 +0000676
Chris Lattner3711b7a2009-09-20 22:42:44 +0000677 unsigned NumNewLines = CountNumNewlinesBetween(SkippedRegion);
Chris Lattner5dafafd2009-08-15 18:32:21 +0000678 if (NumNewLines == 0) {
Chris Lattner0b2353f2009-08-16 02:22:31 +0000679 SM.PrintMessage(CheckStr.Loc,
Chris Lattner5dafafd2009-08-15 18:32:21 +0000680 CheckPrefix+"-NEXT: is on the same line as previous match",
681 "error");
Chris Lattner96077032009-09-20 22:11:44 +0000682 SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()),
Chris Lattner0b2353f2009-08-16 02:22:31 +0000683 "'next' match was here", "note");
Chris Lattner5dafafd2009-08-15 18:32:21 +0000684 SM.PrintMessage(SMLoc::getFromPointer(LastMatch),
685 "previous match was here", "note");
686 return 1;
687 }
688
689 if (NumNewLines != 1) {
Chris Lattner0b2353f2009-08-16 02:22:31 +0000690 SM.PrintMessage(CheckStr.Loc,
Chris Lattner5dafafd2009-08-15 18:32:21 +0000691 CheckPrefix+
692 "-NEXT: is not on the line after the previous match",
693 "error");
Chris Lattner96077032009-09-20 22:11:44 +0000694 SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()),
Chris Lattner0b2353f2009-08-16 02:22:31 +0000695 "'next' match was here", "note");
Chris Lattner5dafafd2009-08-15 18:32:21 +0000696 SM.PrintMessage(SMLoc::getFromPointer(LastMatch),
697 "previous match was here", "note");
698 return 1;
699 }
700 }
Chris Lattnerf15380b2009-09-20 22:35:26 +0000701
702 // If this match had "not strings", verify that they don't exist in the
703 // skipped region.
Chris Lattnereec96952009-09-27 07:56:52 +0000704 for (unsigned ChunkNo = 0, e = CheckStr.NotStrings.size();
705 ChunkNo != e; ++ChunkNo) {
Chris Lattnera29703e2009-09-24 20:39:13 +0000706 size_t MatchLen = 0;
Chris Lattnereec96952009-09-27 07:56:52 +0000707 size_t Pos = CheckStr.NotStrings[ChunkNo].second.Match(SkippedRegion,
708 MatchLen,
709 VariableTable);
Chris Lattnerf15380b2009-09-20 22:35:26 +0000710 if (Pos == StringRef::npos) continue;
711
712 SM.PrintMessage(SMLoc::getFromPointer(LastMatch+Pos),
713 CheckPrefix+"-NOT: string occurred!", "error");
Chris Lattner52870082009-09-24 21:47:32 +0000714 SM.PrintMessage(CheckStr.NotStrings[ChunkNo].first,
Chris Lattnerf15380b2009-09-20 22:35:26 +0000715 CheckPrefix+"-NOT: pattern specified here", "note");
716 return 1;
717 }
718
Chris Lattner5dafafd2009-08-15 18:32:21 +0000719
Chris Lattner81115762009-09-21 02:30:42 +0000720 // Otherwise, everything is good. Step over the matched text and remember
721 // the position after the match as the end of the last match.
Chris Lattner9fc66782009-09-24 20:25:55 +0000722 Buffer = Buffer.substr(MatchLen);
Chris Lattner81115762009-09-21 02:30:42 +0000723 LastMatch = Buffer.data();
Chris Lattner81cb8ca2009-07-08 18:44:05 +0000724 }
725
726 return 0;
727}