FileCheck.cpp revision 7f8e76f5140be7cc9ed1cb66cdcdedfa28147641 
1//===- 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/ADT/OwningPtr.h"
20#include "llvm/Support/CommandLine.h"
21#include "llvm/Support/MemoryBuffer.h"
22#include "llvm/Support/PrettyStackTrace.h"
23#include "llvm/Support/Regex.h"
24#include "llvm/Support/SourceMgr.h"
25#include "llvm/Support/raw_ostream.h"
26#include "llvm/Support/Signals.h"
27#include "llvm/Support/system_error.h"
28#include "llvm/ADT/SmallString.h"
29#include "llvm/ADT/StringExtras.h"
30#include "llvm/ADT/StringMap.h"
31#include <algorithm>
32using namespace llvm;
33
34static cl::opt<std::string>
35CheckFilename(cl::Positional, cl::desc("<check-file>"), cl::Required);
36
37static cl::opt<std::string>
38InputFilename("input-file", cl::desc("File to check (defaults to stdin)"),
39              cl::init("-"), cl::value_desc("filename"));
40
41static cl::opt<std::string>
42CheckPrefix("check-prefix", cl::init("CHECK"),
43            cl::desc("Prefix to use from check file (defaults to 'CHECK')"));
44
45static cl::opt<bool>
46NoCanonicalizeWhiteSpace("strict-whitespace",
47              cl::desc("Do not treat all horizontal whitespace as equivalent"));
48
49//===----------------------------------------------------------------------===//
50// Pattern Handling Code.
51//===----------------------------------------------------------------------===//
52
53class Pattern {
54  SMLoc PatternLoc;
55
56  /// MatchEOF - When set, this pattern only matches the end of file. This is
57  /// used for trailing CHECK-NOTs.
58  bool MatchEOF;
59
60  /// FixedStr - If non-empty, this pattern is a fixed string match with the
61  /// specified fixed string.
62  StringRef FixedStr;
63
64  /// RegEx - If non-empty, this is a regex pattern.
65  std::string RegExStr;
66
67  /// \brief Contains the number of line this pattern is in.
68  unsigned LineNumber;
69
70  /// VariableUses - Entries in this vector map to uses of a variable in the
71  /// pattern, e.g. "foo[[bar]]baz".  In this case, the RegExStr will contain
72  /// "foobaz" and we'll get an entry in this vector that tells us to insert the
73  /// value of bar at offset 3.
74  std::vector<std::pair<StringRef, unsigned> > VariableUses;
75
76  /// VariableDefs - Entries in this vector map to definitions of a variable in
77  /// the pattern, e.g. "foo[[bar:.*]]baz".  In this case, the RegExStr will
78  /// contain "foo(.*)baz" and VariableDefs will contain the pair "bar",1.  The
79  /// index indicates what parenthesized value captures the variable value.
80  std::vector<std::pair<StringRef, unsigned> > VariableDefs;
81
82public:
83
84  Pattern(bool matchEOF = false) : MatchEOF(matchEOF) { }
85
86  bool ParsePattern(StringRef PatternStr, SourceMgr &SM, unsigned LineNumber);
87
88  /// Match - Match the pattern string against the input buffer Buffer.  This
89  /// returns the position that is matched or npos if there is no match.  If
90  /// there is a match, the size of the matched string is returned in MatchLen.
91  ///
92  /// The VariableTable StringMap provides the current values of filecheck
93  /// variables and is updated if this match defines new values.
94  size_t Match(StringRef Buffer, size_t &MatchLen,
95               StringMap<StringRef> &VariableTable) const;
96
97  /// PrintFailureInfo - Print additional information about a failure to match
98  /// involving this pattern.
99  void PrintFailureInfo(const SourceMgr &SM, StringRef Buffer,
100                        const StringMap<StringRef> &VariableTable) const;
101
102private:
103  static void AddFixedStringToRegEx(StringRef FixedStr, std::string &TheStr);
104  bool AddRegExToRegEx(StringRef RegExStr, unsigned &CurParen, SourceMgr &SM);
105
106  /// ComputeMatchDistance - Compute an arbitrary estimate for the quality of
107  /// matching this pattern at the start of \arg Buffer; a distance of zero
108  /// should correspond to a perfect match.
109  unsigned ComputeMatchDistance(StringRef Buffer,
110                               const StringMap<StringRef> &VariableTable) const;
111
112  /// \brief Evaluates expression and stores the result to \p Value.
113  /// \return true on success. false when the expression has invalid syntax.
114  bool EvaluateExpression(StringRef Expr, std::string &Value) const;
115};
116
117
118bool Pattern::ParsePattern(StringRef PatternStr, SourceMgr &SM,
119                           unsigned LineNumber) {
120  this->LineNumber = LineNumber;
121  PatternLoc = SMLoc::getFromPointer(PatternStr.data());
122
123  // Ignore trailing whitespace.
124  while (!PatternStr.empty() &&
125         (PatternStr.back() == ' ' || PatternStr.back() == '\t'))
126    PatternStr = PatternStr.substr(0, PatternStr.size()-1);
127
128  // Check that there is something on the line.
129  if (PatternStr.empty()) {
130    SM.PrintMessage(PatternLoc, SourceMgr::DK_Error,
131                    "found empty check string with prefix '" +
132                    CheckPrefix+":'");
133    return true;
134  }
135
136  // Check to see if this is a fixed string, or if it has regex pieces.
137  if (PatternStr.size() < 2 ||
138      (PatternStr.find("{{") == StringRef::npos &&
139       PatternStr.find("[[") == StringRef::npos)) {
140    FixedStr = PatternStr;
141    return false;
142  }
143
144  // Paren value #0 is for the fully matched string.  Any new parenthesized
145  // values add from there.
146  unsigned CurParen = 1;
147
148  // Otherwise, there is at least one regex piece.  Build up the regex pattern
149  // by escaping scary characters in fixed strings, building up one big regex.
150  while (!PatternStr.empty()) {
151    // RegEx matches.
152    if (PatternStr.startswith("{{")) {
153
154      // Otherwise, this is the start of a regex match.  Scan for the }}.
155      size_t End = PatternStr.find("}}");
156      if (End == StringRef::npos) {
157        SM.PrintMessage(SMLoc::getFromPointer(PatternStr.data()),
158                        SourceMgr::DK_Error,
159                        "found start of regex string with no end '}}'");
160        return true;
161      }
162
163      // Enclose {{}} patterns in parens just like [[]] even though we're not
164      // capturing the result for any purpose.  This is required in case the
165      // expression contains an alternation like: CHECK:  abc{{x|z}}def.  We
166      // want this to turn into: "abc(x|z)def" not "abcx|zdef".
167      RegExStr += '(';
168      ++CurParen;
169
170      if (AddRegExToRegEx(PatternStr.substr(2, End-2), CurParen, SM))
171        return true;
172      RegExStr += ')';
173
174      PatternStr = PatternStr.substr(End+2);
175      continue;
176    }
177
178    // Named RegEx matches.  These are of two forms: [[foo:.*]] which matches .*
179    // (or some other regex) and assigns it to the FileCheck variable 'foo'. The
180    // second form is [[foo]] which is a reference to foo.  The variable name
181    // itself must be of the form "[a-zA-Z_][0-9a-zA-Z_]*", otherwise we reject
182    // it.  This is to catch some common errors.
183    if (PatternStr.startswith("[[")) {
184      // Verify that it is terminated properly.
185      size_t End = PatternStr.find("]]");
186      if (End == StringRef::npos) {
187        SM.PrintMessage(SMLoc::getFromPointer(PatternStr.data()),
188                        SourceMgr::DK_Error,
189                        "invalid named regex reference, no ]] found");
190        return true;
191      }
192
193      StringRef MatchStr = PatternStr.substr(2, End-2);
194      PatternStr = PatternStr.substr(End+2);
195
196      // Get the regex name (e.g. "foo").
197      size_t NameEnd = MatchStr.find(':');
198      StringRef Name = MatchStr.substr(0, NameEnd);
199
200      if (Name.empty()) {
201        SM.PrintMessage(SMLoc::getFromPointer(Name.data()), SourceMgr::DK_Error,
202                        "invalid name in named regex: empty name");
203        return true;
204      }
205
206      // Verify that the name/expression is well formed. FileCheck currently
207      // supports @LINE, @LINE+number, @LINE-number expressions. The check here
208      // is relaxed, more strict check is performed in \c EvaluateExpression.
209      bool IsExpression = false;
210      for (unsigned i = 0, e = Name.size(); i != e; ++i) {
211        if (i == 0 && Name[i] == '@') {
212          if (NameEnd != StringRef::npos) {
213            SM.PrintMessage(SMLoc::getFromPointer(Name.data()),
214                            SourceMgr::DK_Error,
215                            "invalid name in named regex definition");
216            return true;
217          }
218          IsExpression = true;
219          continue;
220        }
221        if (Name[i] != '_' && !isalnum(Name[i]) &&
222            (!IsExpression || (Name[i] != '+' && Name[i] != '-'))) {
223          SM.PrintMessage(SMLoc::getFromPointer(Name.data()+i),
224                          SourceMgr::DK_Error, "invalid name in named regex");
225          return true;
226        }
227      }
228
229      // Name can't start with a digit.
230      if (isdigit(Name[0])) {
231        SM.PrintMessage(SMLoc::getFromPointer(Name.data()), SourceMgr::DK_Error,
232                        "invalid name in named regex");
233        return true;
234      }
235
236      // Handle [[foo]].
237      if (NameEnd == StringRef::npos) {
238        VariableUses.push_back(std::make_pair(Name, RegExStr.size()));
239        continue;
240      }
241
242      // Handle [[foo:.*]].
243      VariableDefs.push_back(std::make_pair(Name, CurParen));
244      RegExStr += '(';
245      ++CurParen;
246
247      if (AddRegExToRegEx(MatchStr.substr(NameEnd+1), CurParen, SM))
248        return true;
249
250      RegExStr += ')';
251    }
252
253    // Handle fixed string matches.
254    // Find the end, which is the start of the next regex.
255    size_t FixedMatchEnd = PatternStr.find("{{");
256    FixedMatchEnd = std::min(FixedMatchEnd, PatternStr.find("[["));
257    AddFixedStringToRegEx(PatternStr.substr(0, FixedMatchEnd), RegExStr);
258    PatternStr = PatternStr.substr(FixedMatchEnd);
259  }
260
261  return false;
262}
263
264void Pattern::AddFixedStringToRegEx(StringRef FixedStr, std::string &TheStr) {
265  // Add the characters from FixedStr to the regex, escaping as needed.  This
266  // avoids "leaning toothpicks" in common patterns.
267  for (unsigned i = 0, e = FixedStr.size(); i != e; ++i) {
268    switch (FixedStr[i]) {
269    // These are the special characters matched in "p_ere_exp".
270    case '(':
271    case ')':
272    case '^':
273    case '$':
274    case '|':
275    case '*':
276    case '+':
277    case '?':
278    case '.':
279    case '[':
280    case '\\':
281    case '{':
282      TheStr += '\\';
283      // FALL THROUGH.
284    default:
285      TheStr += FixedStr[i];
286      break;
287    }
288  }
289}
290
291bool Pattern::AddRegExToRegEx(StringRef RegexStr, unsigned &CurParen,
292                              SourceMgr &SM) {
293  Regex R(RegexStr);
294  std::string Error;
295  if (!R.isValid(Error)) {
296    SM.PrintMessage(SMLoc::getFromPointer(RegexStr.data()), SourceMgr::DK_Error,
297                    "invalid regex: " + Error);
298    return true;
299  }
300
301  RegExStr += RegexStr.str();
302  CurParen += R.getNumMatches();
303  return false;
304}
305
306bool Pattern::EvaluateExpression(StringRef Expr, std::string &Value) const {
307  // The only supported expression is @LINE([\+-]\d+)?
308  if (!Expr.startswith("@LINE"))
309    return false;
310  Expr = Expr.substr(StringRef("@LINE").size());
311  int Offset = 0;
312  if (!Expr.empty()) {
313    if (Expr[0] == '+')
314      Expr = Expr.substr(1);
315    else if (Expr[0] != '-')
316      return false;
317    if (Expr.getAsInteger(10, Offset))
318      return false;
319  }
320  Value = llvm::itostr(LineNumber + Offset);
321  return true;
322}
323
324/// Match - Match the pattern string against the input buffer Buffer.  This
325/// returns the position that is matched or npos if there is no match.  If
326/// there is a match, the size of the matched string is returned in MatchLen.
327size_t Pattern::Match(StringRef Buffer, size_t &MatchLen,
328                      StringMap<StringRef> &VariableTable) const {
329  // If this is the EOF pattern, match it immediately.
330  if (MatchEOF) {
331    MatchLen = 0;
332    return Buffer.size();
333  }
334
335  // If this is a fixed string pattern, just match it now.
336  if (!FixedStr.empty()) {
337    MatchLen = FixedStr.size();
338    return Buffer.find(FixedStr);
339  }
340
341  // Regex match.
342
343  // If there are variable uses, we need to create a temporary string with the
344  // actual value.
345  StringRef RegExToMatch = RegExStr;
346  std::string TmpStr;
347  if (!VariableUses.empty()) {
348    TmpStr = RegExStr;
349
350    unsigned InsertOffset = 0;
351    for (unsigned i = 0, e = VariableUses.size(); i != e; ++i) {
352      std::string Value;
353
354      if (VariableUses[i].first[0] == '@') {
355        if (!EvaluateExpression(VariableUses[i].first, Value))
356          return StringRef::npos;
357      } else {
358        StringMap<StringRef>::iterator it =
359          VariableTable.find(VariableUses[i].first);
360        // If the variable is undefined, return an error.
361        if (it == VariableTable.end())
362          return StringRef::npos;
363
364        // Look up the value and escape it so that we can plop it into the regex.
365        AddFixedStringToRegEx(it->second, Value);
366      }
367
368      // Plop it into the regex at the adjusted offset.
369      TmpStr.insert(TmpStr.begin()+VariableUses[i].second+InsertOffset,
370                    Value.begin(), Value.end());
371      InsertOffset += Value.size();
372    }
373
374    // Match the newly constructed regex.
375    RegExToMatch = TmpStr;
376  }
377
378
379  SmallVector<StringRef, 4> MatchInfo;
380  if (!Regex(RegExToMatch, Regex::Newline).match(Buffer, &MatchInfo))
381    return StringRef::npos;
382
383  // Successful regex match.
384  assert(!MatchInfo.empty() && "Didn't get any match");
385  StringRef FullMatch = MatchInfo[0];
386
387  // If this defines any variables, remember their values.
388  for (unsigned i = 0, e = VariableDefs.size(); i != e; ++i) {
389    assert(VariableDefs[i].second < MatchInfo.size() &&
390           "Internal paren error");
391    VariableTable[VariableDefs[i].first] = MatchInfo[VariableDefs[i].second];
392  }
393
394  MatchLen = FullMatch.size();
395  return FullMatch.data()-Buffer.data();
396}
397
398unsigned Pattern::ComputeMatchDistance(StringRef Buffer,
399                              const StringMap<StringRef> &VariableTable) const {
400  // Just compute the number of matching characters. For regular expressions, we
401  // just compare against the regex itself and hope for the best.
402  //
403  // FIXME: One easy improvement here is have the regex lib generate a single
404  // example regular expression which matches, and use that as the example
405  // string.
406  StringRef ExampleString(FixedStr);
407  if (ExampleString.empty())
408    ExampleString = RegExStr;
409
410  // Only compare up to the first line in the buffer, or the string size.
411  StringRef BufferPrefix = Buffer.substr(0, ExampleString.size());
412  BufferPrefix = BufferPrefix.split('\n').first;
413  return BufferPrefix.edit_distance(ExampleString);
414}
415
416void Pattern::PrintFailureInfo(const SourceMgr &SM, StringRef Buffer,
417                               const StringMap<StringRef> &VariableTable) const{
418  // If this was a regular expression using variables, print the current
419  // variable values.
420  if (!VariableUses.empty()) {
421    for (unsigned i = 0, e = VariableUses.size(); i != e; ++i) {
422      SmallString<256> Msg;
423      raw_svector_ostream OS(Msg);
424      StringRef Var = VariableUses[i].first;
425      if (Var[0] == '@') {
426        std::string Value;
427        if (EvaluateExpression(Var, Value)) {
428          OS << "with expression \"";
429          OS.write_escaped(Var) << "\" equal to \"";
430          OS.write_escaped(Value) << "\"";
431        } else {
432          OS << "uses incorrect expression \"";
433          OS.write_escaped(Var) << "\"";
434        }
435      } else {
436        StringMap<StringRef>::const_iterator it = VariableTable.find(Var);
437
438        // Check for undefined variable references.
439        if (it == VariableTable.end()) {
440          OS << "uses undefined variable \"";
441          OS.write_escaped(Var) << "\"";
442        } else {
443          OS << "with variable \"";
444          OS.write_escaped(Var) << "\" equal to \"";
445          OS.write_escaped(it->second) << "\"";
446        }
447      }
448
449      SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()), SourceMgr::DK_Note,
450                      OS.str());
451    }
452  }
453
454  // Attempt to find the closest/best fuzzy match.  Usually an error happens
455  // because some string in the output didn't exactly match. In these cases, we
456  // would like to show the user a best guess at what "should have" matched, to
457  // save them having to actually check the input manually.
458  size_t NumLinesForward = 0;
459  size_t Best = StringRef::npos;
460  double BestQuality = 0;
461
462  // Use an arbitrary 4k limit on how far we will search.
463  for (size_t i = 0, e = std::min(size_t(4096), Buffer.size()); i != e; ++i) {
464    if (Buffer[i] == '\n')
465      ++NumLinesForward;
466
467    // Patterns have leading whitespace stripped, so skip whitespace when
468    // looking for something which looks like a pattern.
469    if (Buffer[i] == ' ' || Buffer[i] == '\t')
470      continue;
471
472    // Compute the "quality" of this match as an arbitrary combination of the
473    // match distance and the number of lines skipped to get to this match.
474    unsigned Distance = ComputeMatchDistance(Buffer.substr(i), VariableTable);
475    double Quality = Distance + (NumLinesForward / 100.);
476
477    if (Quality < BestQuality || Best == StringRef::npos) {
478      Best = i;
479      BestQuality = Quality;
480    }
481  }
482
483  // Print the "possible intended match here" line if we found something
484  // reasonable and not equal to what we showed in the "scanning from here"
485  // line.
486  if (Best && Best != StringRef::npos && BestQuality < 50) {
487      SM.PrintMessage(SMLoc::getFromPointer(Buffer.data() + Best),
488                      SourceMgr::DK_Note, "possible intended match here");
489
490    // FIXME: If we wanted to be really friendly we would show why the match
491    // failed, as it can be hard to spot simple one character differences.
492  }
493}
494
495//===----------------------------------------------------------------------===//
496// Check Strings.
497//===----------------------------------------------------------------------===//
498
499/// CheckString - This is a check that we found in the input file.
500struct CheckString {
501  /// Pat - The pattern to match.
502  Pattern Pat;
503
504  /// Loc - The location in the match file that the check string was specified.
505  SMLoc Loc;
506
507  /// IsCheckNext - This is true if this is a CHECK-NEXT: directive (as opposed
508  /// to a CHECK: directive.
509  bool IsCheckNext;
510
511  /// NotStrings - These are all of the strings that are disallowed from
512  /// occurring between this match string and the previous one (or start of
513  /// file).
514  std::vector<std::pair<SMLoc, Pattern> > NotStrings;
515
516  CheckString(const Pattern &P, SMLoc L, bool isCheckNext)
517    : Pat(P), Loc(L), IsCheckNext(isCheckNext) {}
518};
519
520/// CanonicalizeInputFile - Remove duplicate horizontal space from the specified
521/// memory buffer, free it, and return a new one.
522static MemoryBuffer *CanonicalizeInputFile(MemoryBuffer *MB) {
523  SmallString<128> NewFile;
524  NewFile.reserve(MB->getBufferSize());
525
526  for (const char *Ptr = MB->getBufferStart(), *End = MB->getBufferEnd();
527       Ptr != End; ++Ptr) {
528    // Eliminate trailing dosish \r.
529    if (Ptr <= End - 2 && Ptr[0] == '\r' && Ptr[1] == '\n') {
530      continue;
531    }
532
533    // If current char is not a horizontal whitespace, dump it to output as is.
534    if (*Ptr != ' ' && *Ptr != '\t') {
535      NewFile.push_back(*Ptr);
536      continue;
537    }
538
539    // Otherwise, add one space and advance over neighboring space.
540    NewFile.push_back(' ');
541    while (Ptr+1 != End &&
542           (Ptr[1] == ' ' || Ptr[1] == '\t'))
543      ++Ptr;
544  }
545
546  // Free the old buffer and return a new one.
547  MemoryBuffer *MB2 =
548    MemoryBuffer::getMemBufferCopy(NewFile.str(), MB->getBufferIdentifier());
549
550  delete MB;
551  return MB2;
552}
553
554
555/// ReadCheckFile - Read the check file, which specifies the sequence of
556/// expected strings.  The strings are added to the CheckStrings vector.
557static bool ReadCheckFile(SourceMgr &SM,
558                          std::vector<CheckString> &CheckStrings) {
559  // Open the check file, and tell SourceMgr about it.
560  OwningPtr<MemoryBuffer> File;
561  if (error_code ec =
562        MemoryBuffer::getFileOrSTDIN(CheckFilename.c_str(), File)) {
563    errs() << "Could not open check file '" << CheckFilename << "': "
564           << ec.message() << '\n';
565    return true;
566  }
567  MemoryBuffer *F = File.take();
568
569  // If we want to canonicalize whitespace, strip excess whitespace from the
570  // buffer containing the CHECK lines.
571  if (!NoCanonicalizeWhiteSpace)
572    F = CanonicalizeInputFile(F);
573
574  SM.AddNewSourceBuffer(F, SMLoc());
575
576  // Find all instances of CheckPrefix followed by : in the file.
577  StringRef Buffer = F->getBuffer();
578
579  std::vector<std::pair<SMLoc, Pattern> > NotMatches;
580
581  unsigned LineNumber = 1;
582
583  while (1) {
584    // See if Prefix occurs in the memory buffer.
585    size_t PrefixLoc = Buffer.find(CheckPrefix);
586    // If we didn't find a match, we're done.
587    if (PrefixLoc == StringRef::npos)
588      break;
589
590    // Recalculate line number.
591    LineNumber += Buffer.substr(0, PrefixLoc).count('\n');
592
593    Buffer = Buffer.substr(PrefixLoc);
594
595    const char *CheckPrefixStart = Buffer.data();
596
597    // When we find a check prefix, keep track of whether we find CHECK: or
598    // CHECK-NEXT:
599    bool IsCheckNext = false, IsCheckNot = false;
600
601    // Verify that the : is present after the prefix.
602    if (Buffer[CheckPrefix.size()] == ':') {
603      Buffer = Buffer.substr(CheckPrefix.size()+1);
604    } else if (Buffer.size() > CheckPrefix.size()+6 &&
605               memcmp(Buffer.data()+CheckPrefix.size(), "-NEXT:", 6) == 0) {
606      Buffer = Buffer.substr(CheckPrefix.size()+6);
607      IsCheckNext = true;
608    } else if (Buffer.size() > CheckPrefix.size()+5 &&
609               memcmp(Buffer.data()+CheckPrefix.size(), "-NOT:", 5) == 0) {
610      Buffer = Buffer.substr(CheckPrefix.size()+5);
611      IsCheckNot = true;
612    } else {
613      Buffer = Buffer.substr(1);
614      continue;
615    }
616
617    // Okay, we found the prefix, yay.  Remember the rest of the line, but
618    // ignore leading and trailing whitespace.
619    Buffer = Buffer.substr(Buffer.find_first_not_of(" \t"));
620
621    // Scan ahead to the end of line.
622    size_t EOL = Buffer.find_first_of("\n\r");
623
624    // Remember the location of the start of the pattern, for diagnostics.
625    SMLoc PatternLoc = SMLoc::getFromPointer(Buffer.data());
626
627    // Parse the pattern.
628    Pattern P;
629    if (P.ParsePattern(Buffer.substr(0, EOL), SM, LineNumber))
630      return true;
631
632    Buffer = Buffer.substr(EOL);
633
634
635    // Verify that CHECK-NEXT lines have at least one CHECK line before them.
636    if (IsCheckNext && CheckStrings.empty()) {
637      SM.PrintMessage(SMLoc::getFromPointer(CheckPrefixStart),
638                      SourceMgr::DK_Error,
639                      "found '"+CheckPrefix+"-NEXT:' without previous '"+
640                      CheckPrefix+ ": line");
641      return true;
642    }
643
644    // Handle CHECK-NOT.
645    if (IsCheckNot) {
646      NotMatches.push_back(std::make_pair(SMLoc::getFromPointer(Buffer.data()),
647                                          P));
648      continue;
649    }
650
651
652    // Okay, add the string we captured to the output vector and move on.
653    CheckStrings.push_back(CheckString(P,
654                                       PatternLoc,
655                                       IsCheckNext));
656    std::swap(NotMatches, CheckStrings.back().NotStrings);
657  }
658
659  // Add an EOF pattern for any trailing CHECK-NOTs.
660  if (!NotMatches.empty()) {
661    CheckStrings.push_back(CheckString(Pattern(true),
662                                       SMLoc::getFromPointer(Buffer.data()),
663                                       false));
664    std::swap(NotMatches, CheckStrings.back().NotStrings);
665  }
666
667  if (CheckStrings.empty()) {
668    errs() << "error: no check strings found with prefix '" << CheckPrefix
669           << ":'\n";
670    return true;
671  }
672
673  return false;
674}
675
676static void PrintCheckFailed(const SourceMgr &SM, const CheckString &CheckStr,
677                             StringRef Buffer,
678                             StringMap<StringRef> &VariableTable) {
679  // Otherwise, we have an error, emit an error message.
680  SM.PrintMessage(CheckStr.Loc, SourceMgr::DK_Error,
681                  "expected string not found in input");
682
683  // Print the "scanning from here" line.  If the current position is at the
684  // end of a line, advance to the start of the next line.
685  Buffer = Buffer.substr(Buffer.find_first_not_of(" \t\n\r"));
686
687  SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()), SourceMgr::DK_Note,
688                  "scanning from here");
689
690  // Allow the pattern to print additional information if desired.
691  CheckStr.Pat.PrintFailureInfo(SM, Buffer, VariableTable);
692}
693
694/// CountNumNewlinesBetween - Count the number of newlines in the specified
695/// range.
696static unsigned CountNumNewlinesBetween(StringRef Range) {
697  unsigned NumNewLines = 0;
698  while (1) {
699    // Scan for newline.
700    Range = Range.substr(Range.find_first_of("\n\r"));
701    if (Range.empty()) return NumNewLines;
702
703    ++NumNewLines;
704
705    // Handle \n\r and \r\n as a single newline.
706    if (Range.size() > 1 &&
707        (Range[1] == '\n' || Range[1] == '\r') &&
708        (Range[0] != Range[1]))
709      Range = Range.substr(1);
710    Range = Range.substr(1);
711  }
712}
713
714int main(int argc, char **argv) {
715  sys::PrintStackTraceOnErrorSignal();
716  PrettyStackTraceProgram X(argc, argv);
717  cl::ParseCommandLineOptions(argc, argv);
718
719  SourceMgr SM;
720
721  // Read the expected strings from the check file.
722  std::vector<CheckString> CheckStrings;
723  if (ReadCheckFile(SM, CheckStrings))
724    return 2;
725
726  // Open the file to check and add it to SourceMgr.
727  OwningPtr<MemoryBuffer> File;
728  if (error_code ec =
729        MemoryBuffer::getFileOrSTDIN(InputFilename.c_str(), File)) {
730    errs() << "Could not open input file '" << InputFilename << "': "
731           << ec.message() << '\n';
732    return 2;
733  }
734  MemoryBuffer *F = File.take();
735
736  if (F->getBufferSize() == 0) {
737    errs() << "FileCheck error: '" << InputFilename << "' is empty.\n";
738    return 2;
739  }
740
741  // Remove duplicate spaces in the input file if requested.
742  if (!NoCanonicalizeWhiteSpace)
743    F = CanonicalizeInputFile(F);
744
745  SM.AddNewSourceBuffer(F, SMLoc());
746
747  /// VariableTable - This holds all the current filecheck variables.
748  StringMap<StringRef> VariableTable;
749
750  // Check that we have all of the expected strings, in order, in the input
751  // file.
752  StringRef Buffer = F->getBuffer();
753
754  const char *LastMatch = Buffer.data();
755
756  for (unsigned StrNo = 0, e = CheckStrings.size(); StrNo != e; ++StrNo) {
757    const CheckString &CheckStr = CheckStrings[StrNo];
758
759    StringRef SearchFrom = Buffer;
760
761    // Find StrNo in the file.
762    size_t MatchLen = 0;
763    size_t MatchPos = CheckStr.Pat.Match(Buffer, MatchLen, VariableTable);
764    Buffer = Buffer.substr(MatchPos);
765
766    // If we didn't find a match, reject the input.
767    if (MatchPos == StringRef::npos) {
768      PrintCheckFailed(SM, CheckStr, SearchFrom, VariableTable);
769      return 1;
770    }
771
772    StringRef SkippedRegion(LastMatch, Buffer.data()-LastMatch);
773
774    // If this check is a "CHECK-NEXT", verify that the previous match was on
775    // the previous line (i.e. that there is one newline between them).
776    if (CheckStr.IsCheckNext) {
777      // Count the number of newlines between the previous match and this one.
778      assert(LastMatch != F->getBufferStart() &&
779             "CHECK-NEXT can't be the first check in a file");
780
781      unsigned NumNewLines = CountNumNewlinesBetween(SkippedRegion);
782      if (NumNewLines == 0) {
783        SM.PrintMessage(CheckStr.Loc, SourceMgr::DK_Error,
784                    CheckPrefix+"-NEXT: is on the same line as previous match");
785        SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()),
786                        SourceMgr::DK_Note, "'next' match was here");
787        SM.PrintMessage(SMLoc::getFromPointer(LastMatch), SourceMgr::DK_Note,
788                        "previous match was here");
789        return 1;
790      }
791
792      if (NumNewLines != 1) {
793        SM.PrintMessage(CheckStr.Loc, SourceMgr::DK_Error, CheckPrefix+
794                        "-NEXT: is not on the line after the previous match");
795        SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()),
796                        SourceMgr::DK_Note, "'next' match was here");
797        SM.PrintMessage(SMLoc::getFromPointer(LastMatch), SourceMgr::DK_Note,
798                        "previous match was here");
799        return 1;
800      }
801    }
802
803    // If this match had "not strings", verify that they don't exist in the
804    // skipped region.
805    for (unsigned ChunkNo = 0, e = CheckStr.NotStrings.size();
806         ChunkNo != e; ++ChunkNo) {
807      size_t MatchLen = 0;
808      size_t Pos = CheckStr.NotStrings[ChunkNo].second.Match(SkippedRegion,
809                                                             MatchLen,
810                                                             VariableTable);
811      if (Pos == StringRef::npos) continue;
812
813      SM.PrintMessage(SMLoc::getFromPointer(LastMatch+Pos), SourceMgr::DK_Error,
814                      CheckPrefix+"-NOT: string occurred!");
815      SM.PrintMessage(CheckStr.NotStrings[ChunkNo].first, SourceMgr::DK_Note,
816                      CheckPrefix+"-NOT: pattern specified here");
817      return 1;
818    }
819
820
821    // Otherwise, everything is good.  Step over the matched text and remember
822    // the position after the match as the end of the last match.
823    Buffer = Buffer.substr(MatchLen);
824    LastMatch = Buffer.data();
825  }
826
827  return 0;
828}
829