FileCheck.cpp revision ee4f5eae1c416eddbecbb2d742e2bb8dc0032bf6 
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/ADT/SmallString.h"
21#include "llvm/ADT/StringExtras.h"
22#include "llvm/ADT/StringMap.h"
23#include "llvm/ADT/StringSet.h"
24#include "llvm/Support/CommandLine.h"
25#include "llvm/Support/MemoryBuffer.h"
26#include "llvm/Support/PrettyStackTrace.h"
27#include "llvm/Support/Regex.h"
28#include "llvm/Support/Signals.h"
29#include "llvm/Support/SourceMgr.h"
30#include "llvm/Support/raw_ostream.h"
31#include "llvm/Support/system_error.h"
32#include <algorithm>
33#include <cctype>
34#include <map>
35#include <string>
36#include <vector>
37using namespace llvm;
38
39static cl::opt<std::string>
40CheckFilename(cl::Positional, cl::desc("<check-file>"), cl::Required);
41
42static cl::opt<std::string>
43InputFilename("input-file", cl::desc("File to check (defaults to stdin)"),
44              cl::init("-"), cl::value_desc("filename"));
45
46static cl::list<std::string>
47CheckPrefixes("check-prefix",
48              cl::desc("Prefix to use from check file (defaults to 'CHECK')"));
49
50static cl::opt<bool>
51NoCanonicalizeWhiteSpace("strict-whitespace",
52              cl::desc("Do not treat all horizontal whitespace as equivalent"));
53
54typedef cl::list<std::string>::const_iterator prefix_iterator;
55
56//===----------------------------------------------------------------------===//
57// Pattern Handling Code.
58//===----------------------------------------------------------------------===//
59
60namespace Check {
61  enum CheckType {
62    CheckNone = 0,
63    CheckPlain,
64    CheckNext,
65    CheckNot,
66    CheckDAG,
67    CheckLabel,
68
69    /// MatchEOF - When set, this pattern only matches the end of file. This is
70    /// used for trailing CHECK-NOTs.
71    CheckEOF
72  };
73}
74
75class Pattern {
76  SMLoc PatternLoc;
77
78  Check::CheckType CheckTy;
79
80  /// FixedStr - If non-empty, this pattern is a fixed string match with the
81  /// specified fixed string.
82  StringRef FixedStr;
83
84  /// RegEx - If non-empty, this is a regex pattern.
85  std::string RegExStr;
86
87  /// \brief Contains the number of line this pattern is in.
88  unsigned LineNumber;
89
90  /// VariableUses - Entries in this vector map to uses of a variable in the
91  /// pattern, e.g. "foo[[bar]]baz".  In this case, the RegExStr will contain
92  /// "foobaz" and we'll get an entry in this vector that tells us to insert the
93  /// value of bar at offset 3.
94  std::vector<std::pair<StringRef, unsigned> > VariableUses;
95
96  /// VariableDefs - Maps definitions of variables to their parenthesized
97  /// capture numbers.
98  /// E.g. for the pattern "foo[[bar:.*]]baz", VariableDefs will map "bar" to 1.
99  std::map<StringRef, unsigned> VariableDefs;
100
101public:
102
103  Pattern(Check::CheckType Ty)
104    : CheckTy(Ty) { }
105
106  /// getLoc - Return the location in source code.
107  SMLoc getLoc() const { return PatternLoc; }
108
109  /// ParsePattern - Parse the given string into the Pattern. Prefix provides
110  /// which prefix is being matched, SM provides the SourceMgr used for error
111  /// reports, and LineNumber is the line number in the input file from which
112  /// the pattern string was read.  Returns true in case of an error, false
113  /// otherwise.
114  bool ParsePattern(StringRef PatternStr,
115                    StringRef Prefix,
116                    SourceMgr &SM,
117                    unsigned LineNumber);
118
119  /// Match - Match the pattern string against the input buffer Buffer.  This
120  /// returns the position that is matched or npos if there is no match.  If
121  /// there is a match, the size of the matched string is returned in MatchLen.
122  ///
123  /// The VariableTable StringMap provides the current values of filecheck
124  /// variables and is updated if this match defines new values.
125  size_t Match(StringRef Buffer, size_t &MatchLen,
126               StringMap<StringRef> &VariableTable) const;
127
128  /// PrintFailureInfo - Print additional information about a failure to match
129  /// involving this pattern.
130  void PrintFailureInfo(const SourceMgr &SM, StringRef Buffer,
131                        const StringMap<StringRef> &VariableTable) const;
132
133  bool hasVariable() const { return !(VariableUses.empty() &&
134                                      VariableDefs.empty()); }
135
136  Check::CheckType getCheckTy() const { return CheckTy; }
137
138private:
139  static void AddFixedStringToRegEx(StringRef FixedStr, std::string &TheStr);
140  bool AddRegExToRegEx(StringRef RS, unsigned &CurParen, SourceMgr &SM);
141  void AddBackrefToRegEx(unsigned BackrefNum);
142
143  /// ComputeMatchDistance - Compute an arbitrary estimate for the quality of
144  /// matching this pattern at the start of \arg Buffer; a distance of zero
145  /// should correspond to a perfect match.
146  unsigned ComputeMatchDistance(StringRef Buffer,
147                               const StringMap<StringRef> &VariableTable) const;
148
149  /// \brief Evaluates expression and stores the result to \p Value.
150  /// \return true on success. false when the expression has invalid syntax.
151  bool EvaluateExpression(StringRef Expr, std::string &Value) const;
152
153  /// \brief Finds the closing sequence of a regex variable usage or
154  /// definition. Str has to point in the beginning of the definition
155  /// (right after the opening sequence).
156  /// \return offset of the closing sequence within Str, or npos if it was not
157  /// found.
158  size_t FindRegexVarEnd(StringRef Str);
159};
160
161
162bool Pattern::ParsePattern(StringRef PatternStr,
163                           StringRef Prefix,
164                           SourceMgr &SM,
165                           unsigned LineNumber) {
166  this->LineNumber = LineNumber;
167  PatternLoc = SMLoc::getFromPointer(PatternStr.data());
168
169  // Ignore trailing whitespace.
170  while (!PatternStr.empty() &&
171         (PatternStr.back() == ' ' || PatternStr.back() == '\t'))
172    PatternStr = PatternStr.substr(0, PatternStr.size()-1);
173
174  // Check that there is something on the line.
175  if (PatternStr.empty()) {
176    SM.PrintMessage(PatternLoc, SourceMgr::DK_Error,
177                    "found empty check string with prefix '" +
178                    Prefix + ":'");
179    return true;
180  }
181
182  // Check to see if this is a fixed string, or if it has regex pieces.
183  if (PatternStr.size() < 2 ||
184      (PatternStr.find("{{") == StringRef::npos &&
185       PatternStr.find("[[") == StringRef::npos)) {
186    FixedStr = PatternStr;
187    return false;
188  }
189
190  // Paren value #0 is for the fully matched string.  Any new parenthesized
191  // values add from there.
192  unsigned CurParen = 1;
193
194  // Otherwise, there is at least one regex piece.  Build up the regex pattern
195  // by escaping scary characters in fixed strings, building up one big regex.
196  while (!PatternStr.empty()) {
197    // RegEx matches.
198    if (PatternStr.startswith("{{")) {
199      // This is the start of a regex match.  Scan for the }}.
200      size_t End = PatternStr.find("}}");
201      if (End == StringRef::npos) {
202        SM.PrintMessage(SMLoc::getFromPointer(PatternStr.data()),
203                        SourceMgr::DK_Error,
204                        "found start of regex string with no end '}}'");
205        return true;
206      }
207
208      // Enclose {{}} patterns in parens just like [[]] even though we're not
209      // capturing the result for any purpose.  This is required in case the
210      // expression contains an alternation like: CHECK:  abc{{x|z}}def.  We
211      // want this to turn into: "abc(x|z)def" not "abcx|zdef".
212      RegExStr += '(';
213      ++CurParen;
214
215      if (AddRegExToRegEx(PatternStr.substr(2, End-2), CurParen, SM))
216        return true;
217      RegExStr += ')';
218
219      PatternStr = PatternStr.substr(End+2);
220      continue;
221    }
222
223    // Named RegEx matches.  These are of two forms: [[foo:.*]] which matches .*
224    // (or some other regex) and assigns it to the FileCheck variable 'foo'. The
225    // second form is [[foo]] which is a reference to foo.  The variable name
226    // itself must be of the form "[a-zA-Z_][0-9a-zA-Z_]*", otherwise we reject
227    // it.  This is to catch some common errors.
228    if (PatternStr.startswith("[[")) {
229      // Find the closing bracket pair ending the match.  End is going to be an
230      // offset relative to the beginning of the match string.
231      size_t End = FindRegexVarEnd(PatternStr.substr(2));
232
233      if (End == StringRef::npos) {
234        SM.PrintMessage(SMLoc::getFromPointer(PatternStr.data()),
235                        SourceMgr::DK_Error,
236                        "invalid named regex reference, no ]] found");
237        return true;
238      }
239
240      StringRef MatchStr = PatternStr.substr(2, End);
241      PatternStr = PatternStr.substr(End+4);
242
243      // Get the regex name (e.g. "foo").
244      size_t NameEnd = MatchStr.find(':');
245      StringRef Name = MatchStr.substr(0, NameEnd);
246
247      if (Name.empty()) {
248        SM.PrintMessage(SMLoc::getFromPointer(Name.data()), SourceMgr::DK_Error,
249                        "invalid name in named regex: empty name");
250        return true;
251      }
252
253      // Verify that the name/expression is well formed. FileCheck currently
254      // supports @LINE, @LINE+number, @LINE-number expressions. The check here
255      // is relaxed, more strict check is performed in \c EvaluateExpression.
256      bool IsExpression = false;
257      for (unsigned i = 0, e = Name.size(); i != e; ++i) {
258        if (i == 0 && Name[i] == '@') {
259          if (NameEnd != StringRef::npos) {
260            SM.PrintMessage(SMLoc::getFromPointer(Name.data()),
261                            SourceMgr::DK_Error,
262                            "invalid name in named regex definition");
263            return true;
264          }
265          IsExpression = true;
266          continue;
267        }
268        if (Name[i] != '_' && !isalnum(Name[i]) &&
269            (!IsExpression || (Name[i] != '+' && Name[i] != '-'))) {
270          SM.PrintMessage(SMLoc::getFromPointer(Name.data()+i),
271                          SourceMgr::DK_Error, "invalid name in named regex");
272          return true;
273        }
274      }
275
276      // Name can't start with a digit.
277      if (isdigit(static_cast<unsigned char>(Name[0]))) {
278        SM.PrintMessage(SMLoc::getFromPointer(Name.data()), SourceMgr::DK_Error,
279                        "invalid name in named regex");
280        return true;
281      }
282
283      // Handle [[foo]].
284      if (NameEnd == StringRef::npos) {
285        // Handle variables that were defined earlier on the same line by
286        // emitting a backreference.
287        if (VariableDefs.find(Name) != VariableDefs.end()) {
288          unsigned VarParenNum = VariableDefs[Name];
289          if (VarParenNum < 1 || VarParenNum > 9) {
290            SM.PrintMessage(SMLoc::getFromPointer(Name.data()),
291                            SourceMgr::DK_Error,
292                            "Can't back-reference more than 9 variables");
293            return true;
294          }
295          AddBackrefToRegEx(VarParenNum);
296        } else {
297          VariableUses.push_back(std::make_pair(Name, RegExStr.size()));
298        }
299        continue;
300      }
301
302      // Handle [[foo:.*]].
303      VariableDefs[Name] = CurParen;
304      RegExStr += '(';
305      ++CurParen;
306
307      if (AddRegExToRegEx(MatchStr.substr(NameEnd+1), CurParen, SM))
308        return true;
309
310      RegExStr += ')';
311    }
312
313    // Handle fixed string matches.
314    // Find the end, which is the start of the next regex.
315    size_t FixedMatchEnd = PatternStr.find("{{");
316    FixedMatchEnd = std::min(FixedMatchEnd, PatternStr.find("[["));
317    AddFixedStringToRegEx(PatternStr.substr(0, FixedMatchEnd), RegExStr);
318    PatternStr = PatternStr.substr(FixedMatchEnd);
319  }
320
321  return false;
322}
323
324void Pattern::AddFixedStringToRegEx(StringRef FixedStr, std::string &TheStr) {
325  // Add the characters from FixedStr to the regex, escaping as needed.  This
326  // avoids "leaning toothpicks" in common patterns.
327  for (unsigned i = 0, e = FixedStr.size(); i != e; ++i) {
328    switch (FixedStr[i]) {
329    // These are the special characters matched in "p_ere_exp".
330    case '(':
331    case ')':
332    case '^':
333    case '$':
334    case '|':
335    case '*':
336    case '+':
337    case '?':
338    case '.':
339    case '[':
340    case '\\':
341    case '{':
342      TheStr += '\\';
343      // FALL THROUGH.
344    default:
345      TheStr += FixedStr[i];
346      break;
347    }
348  }
349}
350
351bool Pattern::AddRegExToRegEx(StringRef RS, unsigned &CurParen,
352                              SourceMgr &SM) {
353  Regex R(RS);
354  std::string Error;
355  if (!R.isValid(Error)) {
356    SM.PrintMessage(SMLoc::getFromPointer(RS.data()), SourceMgr::DK_Error,
357                    "invalid regex: " + Error);
358    return true;
359  }
360
361  RegExStr += RS.str();
362  CurParen += R.getNumMatches();
363  return false;
364}
365
366void Pattern::AddBackrefToRegEx(unsigned BackrefNum) {
367  assert(BackrefNum >= 1 && BackrefNum <= 9 && "Invalid backref number");
368  std::string Backref = std::string("\\") +
369                        std::string(1, '0' + BackrefNum);
370  RegExStr += Backref;
371}
372
373bool Pattern::EvaluateExpression(StringRef Expr, std::string &Value) const {
374  // The only supported expression is @LINE([\+-]\d+)?
375  if (!Expr.startswith("@LINE"))
376    return false;
377  Expr = Expr.substr(StringRef("@LINE").size());
378  int Offset = 0;
379  if (!Expr.empty()) {
380    if (Expr[0] == '+')
381      Expr = Expr.substr(1);
382    else if (Expr[0] != '-')
383      return false;
384    if (Expr.getAsInteger(10, Offset))
385      return false;
386  }
387  Value = llvm::itostr(LineNumber + Offset);
388  return true;
389}
390
391/// Match - Match the pattern string against the input buffer Buffer.  This
392/// returns the position that is matched or npos if there is no match.  If
393/// there is a match, the size of the matched string is returned in MatchLen.
394size_t Pattern::Match(StringRef Buffer, size_t &MatchLen,
395                      StringMap<StringRef> &VariableTable) const {
396  // If this is the EOF pattern, match it immediately.
397  if (CheckTy == Check::CheckEOF) {
398    MatchLen = 0;
399    return Buffer.size();
400  }
401
402  // If this is a fixed string pattern, just match it now.
403  if (!FixedStr.empty()) {
404    MatchLen = FixedStr.size();
405    return Buffer.find(FixedStr);
406  }
407
408  // Regex match.
409
410  // If there are variable uses, we need to create a temporary string with the
411  // actual value.
412  StringRef RegExToMatch = RegExStr;
413  std::string TmpStr;
414  if (!VariableUses.empty()) {
415    TmpStr = RegExStr;
416
417    unsigned InsertOffset = 0;
418    for (unsigned i = 0, e = VariableUses.size(); i != e; ++i) {
419      std::string Value;
420
421      if (VariableUses[i].first[0] == '@') {
422        if (!EvaluateExpression(VariableUses[i].first, Value))
423          return StringRef::npos;
424      } else {
425        StringMap<StringRef>::iterator it =
426          VariableTable.find(VariableUses[i].first);
427        // If the variable is undefined, return an error.
428        if (it == VariableTable.end())
429          return StringRef::npos;
430
431        // Look up the value and escape it so that we can plop it into the regex.
432        AddFixedStringToRegEx(it->second, Value);
433      }
434
435      // Plop it into the regex at the adjusted offset.
436      TmpStr.insert(TmpStr.begin()+VariableUses[i].second+InsertOffset,
437                    Value.begin(), Value.end());
438      InsertOffset += Value.size();
439    }
440
441    // Match the newly constructed regex.
442    RegExToMatch = TmpStr;
443  }
444
445
446  SmallVector<StringRef, 4> MatchInfo;
447  if (!Regex(RegExToMatch, Regex::Newline).match(Buffer, &MatchInfo))
448    return StringRef::npos;
449
450  // Successful regex match.
451  assert(!MatchInfo.empty() && "Didn't get any match");
452  StringRef FullMatch = MatchInfo[0];
453
454  // If this defines any variables, remember their values.
455  for (std::map<StringRef, unsigned>::const_iterator I = VariableDefs.begin(),
456                                                     E = VariableDefs.end();
457       I != E; ++I) {
458    assert(I->second < MatchInfo.size() && "Internal paren error");
459    VariableTable[I->first] = MatchInfo[I->second];
460  }
461
462  MatchLen = FullMatch.size();
463  return FullMatch.data()-Buffer.data();
464}
465
466unsigned Pattern::ComputeMatchDistance(StringRef Buffer,
467                              const StringMap<StringRef> &VariableTable) const {
468  // Just compute the number of matching characters. For regular expressions, we
469  // just compare against the regex itself and hope for the best.
470  //
471  // FIXME: One easy improvement here is have the regex lib generate a single
472  // example regular expression which matches, and use that as the example
473  // string.
474  StringRef ExampleString(FixedStr);
475  if (ExampleString.empty())
476    ExampleString = RegExStr;
477
478  // Only compare up to the first line in the buffer, or the string size.
479  StringRef BufferPrefix = Buffer.substr(0, ExampleString.size());
480  BufferPrefix = BufferPrefix.split('\n').first;
481  return BufferPrefix.edit_distance(ExampleString);
482}
483
484void Pattern::PrintFailureInfo(const SourceMgr &SM, StringRef Buffer,
485                               const StringMap<StringRef> &VariableTable) const{
486  // If this was a regular expression using variables, print the current
487  // variable values.
488  if (!VariableUses.empty()) {
489    for (unsigned i = 0, e = VariableUses.size(); i != e; ++i) {
490      SmallString<256> Msg;
491      raw_svector_ostream OS(Msg);
492      StringRef Var = VariableUses[i].first;
493      if (Var[0] == '@') {
494        std::string Value;
495        if (EvaluateExpression(Var, Value)) {
496          OS << "with expression \"";
497          OS.write_escaped(Var) << "\" equal to \"";
498          OS.write_escaped(Value) << "\"";
499        } else {
500          OS << "uses incorrect expression \"";
501          OS.write_escaped(Var) << "\"";
502        }
503      } else {
504        StringMap<StringRef>::const_iterator it = VariableTable.find(Var);
505
506        // Check for undefined variable references.
507        if (it == VariableTable.end()) {
508          OS << "uses undefined variable \"";
509          OS.write_escaped(Var) << "\"";
510        } else {
511          OS << "with variable \"";
512          OS.write_escaped(Var) << "\" equal to \"";
513          OS.write_escaped(it->second) << "\"";
514        }
515      }
516
517      SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()), SourceMgr::DK_Note,
518                      OS.str());
519    }
520  }
521
522  // Attempt to find the closest/best fuzzy match.  Usually an error happens
523  // because some string in the output didn't exactly match. In these cases, we
524  // would like to show the user a best guess at what "should have" matched, to
525  // save them having to actually check the input manually.
526  size_t NumLinesForward = 0;
527  size_t Best = StringRef::npos;
528  double BestQuality = 0;
529
530  // Use an arbitrary 4k limit on how far we will search.
531  for (size_t i = 0, e = std::min(size_t(4096), Buffer.size()); i != e; ++i) {
532    if (Buffer[i] == '\n')
533      ++NumLinesForward;
534
535    // Patterns have leading whitespace stripped, so skip whitespace when
536    // looking for something which looks like a pattern.
537    if (Buffer[i] == ' ' || Buffer[i] == '\t')
538      continue;
539
540    // Compute the "quality" of this match as an arbitrary combination of the
541    // match distance and the number of lines skipped to get to this match.
542    unsigned Distance = ComputeMatchDistance(Buffer.substr(i), VariableTable);
543    double Quality = Distance + (NumLinesForward / 100.);
544
545    if (Quality < BestQuality || Best == StringRef::npos) {
546      Best = i;
547      BestQuality = Quality;
548    }
549  }
550
551  // Print the "possible intended match here" line if we found something
552  // reasonable and not equal to what we showed in the "scanning from here"
553  // line.
554  if (Best && Best != StringRef::npos && BestQuality < 50) {
555      SM.PrintMessage(SMLoc::getFromPointer(Buffer.data() + Best),
556                      SourceMgr::DK_Note, "possible intended match here");
557
558    // FIXME: If we wanted to be really friendly we would show why the match
559    // failed, as it can be hard to spot simple one character differences.
560  }
561}
562
563size_t Pattern::FindRegexVarEnd(StringRef Str) {
564  // Offset keeps track of the current offset within the input Str
565  size_t Offset = 0;
566  // [...] Nesting depth
567  size_t BracketDepth = 0;
568
569  while (!Str.empty()) {
570    if (Str.startswith("]]") && BracketDepth == 0)
571      return Offset;
572    if (Str[0] == '\\') {
573      // Backslash escapes the next char within regexes, so skip them both.
574      Str = Str.substr(2);
575      Offset += 2;
576    } else {
577      switch (Str[0]) {
578        default:
579          break;
580        case '[':
581          BracketDepth++;
582          break;
583        case ']':
584          assert(BracketDepth > 0 && "Invalid regex");
585          BracketDepth--;
586          break;
587      }
588      Str = Str.substr(1);
589      Offset++;
590    }
591  }
592
593  return StringRef::npos;
594}
595
596
597//===----------------------------------------------------------------------===//
598// Check Strings.
599//===----------------------------------------------------------------------===//
600
601/// CheckString - This is a check that we found in the input file.
602struct CheckString {
603  /// Pat - The pattern to match.
604  Pattern Pat;
605
606  /// Prefix - Which prefix name this check matched.
607  StringRef Prefix;
608
609  /// Loc - The location in the match file that the check string was specified.
610  SMLoc Loc;
611
612  /// CheckTy - Specify what kind of check this is. e.g. CHECK-NEXT: directive,
613  /// as opposed to a CHECK: directive.
614  Check::CheckType CheckTy;
615
616  /// DagNotStrings - These are all of the strings that are disallowed from
617  /// occurring between this match string and the previous one (or start of
618  /// file).
619  std::vector<Pattern> DagNotStrings;
620
621
622  CheckString(const Pattern &P,
623              StringRef S,
624              SMLoc L,
625              Check::CheckType Ty)
626    : Pat(P), Prefix(S), Loc(L), CheckTy(Ty) {}
627
628  /// Check - Match check string and its "not strings" and/or "dag strings".
629  size_t Check(const SourceMgr &SM, StringRef Buffer, bool IsLabelScanMode,
630               size_t &MatchLen, StringMap<StringRef> &VariableTable) const;
631
632  /// CheckNext - Verify there is a single line in the given buffer.
633  bool CheckNext(const SourceMgr &SM, StringRef Buffer) const;
634
635  /// CheckNot - Verify there's no "not strings" in the given buffer.
636  bool CheckNot(const SourceMgr &SM, StringRef Buffer,
637                const std::vector<const Pattern *> &NotStrings,
638                StringMap<StringRef> &VariableTable) const;
639
640  /// CheckDag - Match "dag strings" and their mixed "not strings".
641  size_t CheckDag(const SourceMgr &SM, StringRef Buffer,
642                  std::vector<const Pattern *> &NotStrings,
643                  StringMap<StringRef> &VariableTable) const;
644};
645
646/// Canonicalize whitespaces in the input file. Line endings are replaced
647/// with UNIX-style '\n'.
648///
649/// \param PreserveHorizontal Don't squash consecutive horizontal whitespace
650/// characters to a single space.
651static MemoryBuffer *CanonicalizeInputFile(MemoryBuffer *MB,
652                                           bool PreserveHorizontal) {
653  SmallString<128> NewFile;
654  NewFile.reserve(MB->getBufferSize());
655
656  for (const char *Ptr = MB->getBufferStart(), *End = MB->getBufferEnd();
657       Ptr != End; ++Ptr) {
658    // Eliminate trailing dosish \r.
659    if (Ptr <= End - 2 && Ptr[0] == '\r' && Ptr[1] == '\n') {
660      continue;
661    }
662
663    // If current char is not a horizontal whitespace or if horizontal
664    // whitespace canonicalization is disabled, dump it to output as is.
665    if (PreserveHorizontal || (*Ptr != ' ' && *Ptr != '\t')) {
666      NewFile.push_back(*Ptr);
667      continue;
668    }
669
670    // Otherwise, add one space and advance over neighboring space.
671    NewFile.push_back(' ');
672    while (Ptr+1 != End &&
673           (Ptr[1] == ' ' || Ptr[1] == '\t'))
674      ++Ptr;
675  }
676
677  // Free the old buffer and return a new one.
678  MemoryBuffer *MB2 =
679    MemoryBuffer::getMemBufferCopy(NewFile.str(), MB->getBufferIdentifier());
680
681  delete MB;
682  return MB2;
683}
684
685static bool IsPartOfWord(char c) {
686  return (isalnum(c) || c == '-' || c == '_');
687}
688
689// Get the size of the prefix extension.
690static size_t CheckTypeSize(Check::CheckType Ty) {
691  switch (Ty) {
692  case Check::CheckNone:
693    return 0;
694
695  case Check::CheckPlain:
696    return sizeof(":") - 1;
697
698  case Check::CheckNext:
699    return sizeof("-NEXT:") - 1;
700
701  case Check::CheckNot:
702    return sizeof("-NOT:") - 1;
703
704  case Check::CheckDAG:
705    return sizeof("-DAG:") - 1;
706
707  case Check::CheckLabel:
708    return sizeof("-LABEL:") - 1;
709
710  case Check::CheckEOF:
711    llvm_unreachable("Should not be using EOF size");
712  }
713
714  llvm_unreachable("Bad check type");
715}
716
717static Check::CheckType FindCheckType(StringRef Buffer, StringRef Prefix) {
718  char NextChar = Buffer[Prefix.size()];
719
720  // Verify that the : is present after the prefix.
721  if (NextChar == ':')
722    return Check::CheckPlain;
723
724  if (NextChar != '-')
725    return Check::CheckNone;
726
727  StringRef Rest = Buffer.drop_front(Prefix.size() + 1);
728  if (Rest.startswith("NEXT:"))
729    return Check::CheckNext;
730
731  if (Rest.startswith("NOT:"))
732    return Check::CheckNot;
733
734  if (Rest.startswith("DAG:"))
735    return Check::CheckDAG;
736
737  if (Rest.startswith("LABEL:"))
738    return Check::CheckLabel;
739
740  return Check::CheckNone;
741}
742
743// From the given position, find the next character after the word.
744static size_t SkipWord(StringRef Str, size_t Loc) {
745  while (Loc < Str.size() && IsPartOfWord(Str[Loc]))
746    ++Loc;
747  return Loc;
748}
749
750// Try to find the first match in buffer for any prefix. If a valid match is
751// found, return that prefix and set its type and location.  If there are almost
752// matches (e.g. the actual prefix string is found, but is not an actual check
753// string), but no valid match, return an empty string and set the position to
754// resume searching from. If no partial matches are found, return an empty
755// string and the location will be StringRef::npos. If one prefix is a substring
756// of another, the maximal match should be found. e.g. if "A" and "AA" are
757// prefixes then AA-CHECK: should match the second one.
758static StringRef FindFirstCandidateMatch(StringRef &Buffer,
759                                         Check::CheckType &CheckTy,
760                                         size_t &CheckLoc) {
761  StringRef FirstPrefix;
762  size_t FirstLoc = StringRef::npos;
763  size_t SearchLoc = StringRef::npos;
764  Check::CheckType FirstTy = Check::CheckNone;
765
766  CheckTy = Check::CheckNone;
767  CheckLoc = StringRef::npos;
768
769  for (prefix_iterator I = CheckPrefixes.begin(), E = CheckPrefixes.end();
770       I != E; ++I) {
771    StringRef Prefix(*I);
772    size_t PrefixLoc = Buffer.find(Prefix);
773
774    if (PrefixLoc == StringRef::npos)
775      continue;
776
777    // Track where we are searching for invalid prefixes that look almost right.
778    // We need to only advance to the first partial match on the next attempt
779    // since a partial match could be a substring of a later, valid prefix.
780    // Need to skip to the end of the word, otherwise we could end up
781    // matching a prefix in a substring later.
782    if (PrefixLoc < SearchLoc)
783      SearchLoc = SkipWord(Buffer, PrefixLoc);
784
785    // We only want to find the first match to avoid skipping some.
786    if (PrefixLoc > FirstLoc)
787      continue;
788
789    StringRef Rest = Buffer.drop_front(PrefixLoc);
790    // Make sure we have actually found the prefix, and not a word containing
791    // it. This should also prevent matching the wrong prefix when one is a
792    // substring of another.
793    if (PrefixLoc != 0 && IsPartOfWord(Buffer[PrefixLoc - 1]))
794      continue;
795
796    Check::CheckType Ty = FindCheckType(Rest, Prefix);
797    if (Ty == Check::CheckNone)
798      continue;
799
800    FirstLoc = PrefixLoc;
801    FirstTy = Ty;
802    FirstPrefix = Prefix;
803  }
804
805  if (FirstPrefix.empty()) {
806    CheckLoc = SearchLoc;
807  } else {
808    CheckTy = FirstTy;
809    CheckLoc = FirstLoc;
810  }
811
812  return FirstPrefix;
813}
814
815static StringRef FindFirstMatchingPrefix(StringRef &Buffer,
816                                         unsigned &LineNumber,
817                                         Check::CheckType &CheckTy,
818                                         size_t &CheckLoc) {
819  while (!Buffer.empty()) {
820    StringRef Prefix = FindFirstCandidateMatch(Buffer, CheckTy, CheckLoc);
821    // If we found a real match, we are done.
822    if (!Prefix.empty()) {
823      LineNumber += Buffer.substr(0, CheckLoc).count('\n');
824      return Prefix;
825    }
826
827    // We didn't find any almost matches either, we are also done.
828    if (CheckLoc == StringRef::npos)
829      return StringRef();
830
831    LineNumber += Buffer.substr(0, CheckLoc + 1).count('\n');
832
833    // Advance to the last possible match we found and try again.
834    Buffer = Buffer.drop_front(CheckLoc + 1);
835  }
836
837  return StringRef();
838}
839
840/// ReadCheckFile - Read the check file, which specifies the sequence of
841/// expected strings.  The strings are added to the CheckStrings vector.
842/// Returns true in case of an error, false otherwise.
843static bool ReadCheckFile(SourceMgr &SM,
844                          std::vector<CheckString> &CheckStrings) {
845  OwningPtr<MemoryBuffer> File;
846  if (error_code ec =
847        MemoryBuffer::getFileOrSTDIN(CheckFilename, File)) {
848    errs() << "Could not open check file '" << CheckFilename << "': "
849           << ec.message() << '\n';
850    return true;
851  }
852
853  // If we want to canonicalize whitespace, strip excess whitespace from the
854  // buffer containing the CHECK lines. Remove DOS style line endings.
855  MemoryBuffer *F =
856    CanonicalizeInputFile(File.take(), NoCanonicalizeWhiteSpace);
857
858  SM.AddNewSourceBuffer(F, SMLoc());
859
860  // Find all instances of CheckPrefix followed by : in the file.
861  StringRef Buffer = F->getBuffer();
862  std::vector<Pattern> DagNotMatches;
863
864  // LineNumber keeps track of the line on which CheckPrefix instances are
865  // found.
866  unsigned LineNumber = 1;
867
868  while (1) {
869    Check::CheckType CheckTy;
870    size_t PrefixLoc;
871
872    // See if a prefix occurs in the memory buffer.
873    StringRef UsedPrefix = FindFirstMatchingPrefix(Buffer,
874                                                   LineNumber,
875                                                   CheckTy,
876                                                   PrefixLoc);
877    if (UsedPrefix.empty())
878      break;
879
880    Buffer = Buffer.drop_front(PrefixLoc);
881
882    // Location to use for error messages.
883    const char *UsedPrefixStart = Buffer.data() + (PrefixLoc == 0 ? 0 : 1);
884
885    // PrefixLoc is to the start of the prefix. Skip to the end.
886    Buffer = Buffer.drop_front(UsedPrefix.size() + CheckTypeSize(CheckTy));
887
888    // Okay, we found the prefix, yay. Remember the rest of the line, but ignore
889    // leading and trailing whitespace.
890    Buffer = Buffer.substr(Buffer.find_first_not_of(" \t"));
891
892    // Scan ahead to the end of line.
893    size_t EOL = Buffer.find_first_of("\n\r");
894
895    // Remember the location of the start of the pattern, for diagnostics.
896    SMLoc PatternLoc = SMLoc::getFromPointer(Buffer.data());
897
898    // Parse the pattern.
899    Pattern P(CheckTy);
900    if (P.ParsePattern(Buffer.substr(0, EOL), UsedPrefix, SM, LineNumber))
901      return true;
902
903    // Verify that CHECK-LABEL lines do not define or use variables
904    if ((CheckTy == Check::CheckLabel) && P.hasVariable()) {
905      SM.PrintMessage(SMLoc::getFromPointer(UsedPrefixStart),
906                      SourceMgr::DK_Error,
907                      "found '" + UsedPrefix + "-LABEL:'"
908                      " with variable definition or use");
909      return true;
910    }
911
912    Buffer = Buffer.substr(EOL);
913
914    // Verify that CHECK-NEXT lines have at least one CHECK line before them.
915    if ((CheckTy == Check::CheckNext) && CheckStrings.empty()) {
916      SM.PrintMessage(SMLoc::getFromPointer(UsedPrefixStart),
917                      SourceMgr::DK_Error,
918                      "found '" + UsedPrefix + "-NEXT:' without previous '"
919                      + UsedPrefix + ": line");
920      return true;
921    }
922
923    // Handle CHECK-DAG/-NOT.
924    if (CheckTy == Check::CheckDAG || CheckTy == Check::CheckNot) {
925      DagNotMatches.push_back(P);
926      continue;
927    }
928
929    // Okay, add the string we captured to the output vector and move on.
930    CheckStrings.push_back(CheckString(P,
931                                       UsedPrefix,
932                                       PatternLoc,
933                                       CheckTy));
934    std::swap(DagNotMatches, CheckStrings.back().DagNotStrings);
935  }
936
937  // Add an EOF pattern for any trailing CHECK-DAG/-NOTs, and use the first
938  // prefix as a filler for the error message.
939  if (!DagNotMatches.empty()) {
940    CheckStrings.push_back(CheckString(Pattern(Check::CheckEOF),
941                                       CheckPrefixes[0],
942                                       SMLoc::getFromPointer(Buffer.data()),
943                                       Check::CheckEOF));
944    std::swap(DagNotMatches, CheckStrings.back().DagNotStrings);
945  }
946
947  if (CheckStrings.empty()) {
948    errs() << "error: no check strings found with prefix"
949           << (CheckPrefixes.size() > 1 ? "es " : " ");
950    for (size_t I = 0, N = CheckPrefixes.size(); I != N; ++I) {
951      StringRef Prefix(CheckPrefixes[I]);
952      errs() << '\'' << Prefix << ":'";
953      if (I != N - 1)
954        errs() << ", ";
955    }
956
957    errs() << '\n';
958    return true;
959  }
960
961  return false;
962}
963
964static void PrintCheckFailed(const SourceMgr &SM, const SMLoc &Loc,
965                             const Pattern &Pat, StringRef Buffer,
966                             StringMap<StringRef> &VariableTable) {
967  // Otherwise, we have an error, emit an error message.
968  SM.PrintMessage(Loc, SourceMgr::DK_Error,
969                  "expected string not found in input");
970
971  // Print the "scanning from here" line.  If the current position is at the
972  // end of a line, advance to the start of the next line.
973  Buffer = Buffer.substr(Buffer.find_first_not_of(" \t\n\r"));
974
975  SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()), SourceMgr::DK_Note,
976                  "scanning from here");
977
978  // Allow the pattern to print additional information if desired.
979  Pat.PrintFailureInfo(SM, Buffer, VariableTable);
980}
981
982static void PrintCheckFailed(const SourceMgr &SM, const CheckString &CheckStr,
983                             StringRef Buffer,
984                             StringMap<StringRef> &VariableTable) {
985  PrintCheckFailed(SM, CheckStr.Loc, CheckStr.Pat, Buffer, VariableTable);
986}
987
988/// CountNumNewlinesBetween - Count the number of newlines in the specified
989/// range.
990static unsigned CountNumNewlinesBetween(StringRef Range) {
991  unsigned NumNewLines = 0;
992  while (1) {
993    // Scan for newline.
994    Range = Range.substr(Range.find_first_of("\n\r"));
995    if (Range.empty()) return NumNewLines;
996
997    ++NumNewLines;
998
999    // Handle \n\r and \r\n as a single newline.
1000    if (Range.size() > 1 &&
1001        (Range[1] == '\n' || Range[1] == '\r') &&
1002        (Range[0] != Range[1]))
1003      Range = Range.substr(1);
1004    Range = Range.substr(1);
1005  }
1006}
1007
1008size_t CheckString::Check(const SourceMgr &SM, StringRef Buffer,
1009                          bool IsLabelScanMode, size_t &MatchLen,
1010                          StringMap<StringRef> &VariableTable) const {
1011  size_t LastPos = 0;
1012  std::vector<const Pattern *> NotStrings;
1013
1014  // IsLabelScanMode is true when we are scanning forward to find CHECK-LABEL
1015  // bounds; we have not processed variable definitions within the bounded block
1016  // yet so cannot handle any final CHECK-DAG yet; this is handled when going
1017  // over the block again (including the last CHECK-LABEL) in normal mode.
1018  if (!IsLabelScanMode) {
1019    // Match "dag strings" (with mixed "not strings" if any).
1020    LastPos = CheckDag(SM, Buffer, NotStrings, VariableTable);
1021    if (LastPos == StringRef::npos)
1022      return StringRef::npos;
1023  }
1024
1025  // Match itself from the last position after matching CHECK-DAG.
1026  StringRef MatchBuffer = Buffer.substr(LastPos);
1027  size_t MatchPos = Pat.Match(MatchBuffer, MatchLen, VariableTable);
1028  if (MatchPos == StringRef::npos) {
1029    PrintCheckFailed(SM, *this, MatchBuffer, VariableTable);
1030    return StringRef::npos;
1031  }
1032  MatchPos += LastPos;
1033
1034  // Similar to the above, in "label-scan mode" we can't yet handle CHECK-NEXT
1035  // or CHECK-NOT
1036  if (!IsLabelScanMode) {
1037    StringRef SkippedRegion = Buffer.substr(LastPos, MatchPos);
1038
1039    // If this check is a "CHECK-NEXT", verify that the previous match was on
1040    // the previous line (i.e. that there is one newline between them).
1041    if (CheckNext(SM, SkippedRegion))
1042      return StringRef::npos;
1043
1044    // If this match had "not strings", verify that they don't exist in the
1045    // skipped region.
1046    if (CheckNot(SM, SkippedRegion, NotStrings, VariableTable))
1047      return StringRef::npos;
1048  }
1049
1050  return MatchPos;
1051}
1052
1053bool CheckString::CheckNext(const SourceMgr &SM, StringRef Buffer) const {
1054  if (CheckTy != Check::CheckNext)
1055    return false;
1056
1057  // Count the number of newlines between the previous match and this one.
1058  assert(Buffer.data() !=
1059         SM.getMemoryBuffer(
1060           SM.FindBufferContainingLoc(
1061             SMLoc::getFromPointer(Buffer.data())))->getBufferStart() &&
1062         "CHECK-NEXT can't be the first check in a file");
1063
1064  unsigned NumNewLines = CountNumNewlinesBetween(Buffer);
1065
1066  if (NumNewLines == 0) {
1067    SM.PrintMessage(Loc, SourceMgr::DK_Error, Prefix +
1068                    "-NEXT: is on the same line as previous match");
1069    SM.PrintMessage(SMLoc::getFromPointer(Buffer.end()),
1070                    SourceMgr::DK_Note, "'next' match was here");
1071    SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()), SourceMgr::DK_Note,
1072                    "previous match ended here");
1073    return true;
1074  }
1075
1076  if (NumNewLines != 1) {
1077    SM.PrintMessage(Loc, SourceMgr::DK_Error, Prefix +
1078                    "-NEXT: is not on the line after the previous match");
1079    SM.PrintMessage(SMLoc::getFromPointer(Buffer.end()),
1080                    SourceMgr::DK_Note, "'next' match was here");
1081    SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()), SourceMgr::DK_Note,
1082                    "previous match ended here");
1083    return true;
1084  }
1085
1086  return false;
1087}
1088
1089bool CheckString::CheckNot(const SourceMgr &SM, StringRef Buffer,
1090                           const std::vector<const Pattern *> &NotStrings,
1091                           StringMap<StringRef> &VariableTable) const {
1092  for (unsigned ChunkNo = 0, e = NotStrings.size();
1093       ChunkNo != e; ++ChunkNo) {
1094    const Pattern *Pat = NotStrings[ChunkNo];
1095    assert((Pat->getCheckTy() == Check::CheckNot) && "Expect CHECK-NOT!");
1096
1097    size_t MatchLen = 0;
1098    size_t Pos = Pat->Match(Buffer, MatchLen, VariableTable);
1099
1100    if (Pos == StringRef::npos) continue;
1101
1102    SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()+Pos),
1103                    SourceMgr::DK_Error,
1104                    Prefix + "-NOT: string occurred!");
1105    SM.PrintMessage(Pat->getLoc(), SourceMgr::DK_Note,
1106                    Prefix + "-NOT: pattern specified here");
1107    return true;
1108  }
1109
1110  return false;
1111}
1112
1113size_t CheckString::CheckDag(const SourceMgr &SM, StringRef Buffer,
1114                             std::vector<const Pattern *> &NotStrings,
1115                             StringMap<StringRef> &VariableTable) const {
1116  if (DagNotStrings.empty())
1117    return 0;
1118
1119  size_t LastPos = 0;
1120  size_t StartPos = LastPos;
1121
1122  for (unsigned ChunkNo = 0, e = DagNotStrings.size();
1123       ChunkNo != e; ++ChunkNo) {
1124    const Pattern &Pat = DagNotStrings[ChunkNo];
1125
1126    assert((Pat.getCheckTy() == Check::CheckDAG ||
1127            Pat.getCheckTy() == Check::CheckNot) &&
1128           "Invalid CHECK-DAG or CHECK-NOT!");
1129
1130    if (Pat.getCheckTy() == Check::CheckNot) {
1131      NotStrings.push_back(&Pat);
1132      continue;
1133    }
1134
1135    assert((Pat.getCheckTy() == Check::CheckDAG) && "Expect CHECK-DAG!");
1136
1137    size_t MatchLen = 0, MatchPos;
1138
1139    // CHECK-DAG always matches from the start.
1140    StringRef MatchBuffer = Buffer.substr(StartPos);
1141    MatchPos = Pat.Match(MatchBuffer, MatchLen, VariableTable);
1142    // With a group of CHECK-DAGs, a single mismatching means the match on
1143    // that group of CHECK-DAGs fails immediately.
1144    if (MatchPos == StringRef::npos) {
1145      PrintCheckFailed(SM, Pat.getLoc(), Pat, MatchBuffer, VariableTable);
1146      return StringRef::npos;
1147    }
1148    // Re-calc it as the offset relative to the start of the original string.
1149    MatchPos += StartPos;
1150
1151    if (!NotStrings.empty()) {
1152      if (MatchPos < LastPos) {
1153        // Reordered?
1154        SM.PrintMessage(SMLoc::getFromPointer(Buffer.data() + MatchPos),
1155                        SourceMgr::DK_Error,
1156                        Prefix + "-DAG: found a match of CHECK-DAG"
1157                        " reordering across a CHECK-NOT");
1158        SM.PrintMessage(SMLoc::getFromPointer(Buffer.data() + LastPos),
1159                        SourceMgr::DK_Note,
1160                        Prefix + "-DAG: the farthest match of CHECK-DAG"
1161                        " is found here");
1162        SM.PrintMessage(NotStrings[0]->getLoc(), SourceMgr::DK_Note,
1163                        Prefix + "-NOT: the crossed pattern specified"
1164                        " here");
1165        SM.PrintMessage(Pat.getLoc(), SourceMgr::DK_Note,
1166                        Prefix + "-DAG: the reordered pattern specified"
1167                        " here");
1168        return StringRef::npos;
1169      }
1170      // All subsequent CHECK-DAGs should be matched from the farthest
1171      // position of all precedent CHECK-DAGs (including this one.)
1172      StartPos = LastPos;
1173      // If there's CHECK-NOTs between two CHECK-DAGs or from CHECK to
1174      // CHECK-DAG, verify that there's no 'not' strings occurred in that
1175      // region.
1176      StringRef SkippedRegion = Buffer.substr(LastPos, MatchPos);
1177      if (CheckNot(SM, SkippedRegion, NotStrings, VariableTable))
1178        return StringRef::npos;
1179      // Clear "not strings".
1180      NotStrings.clear();
1181    }
1182
1183    // Update the last position with CHECK-DAG matches.
1184    LastPos = std::max(MatchPos + MatchLen, LastPos);
1185  }
1186
1187  return LastPos;
1188}
1189
1190// A check prefix must contain only alphanumeric, hyphens and underscores.
1191static bool ValidateCheckPrefix(StringRef CheckPrefix) {
1192  Regex Validator("^[a-zA-Z0-9_-]*$");
1193  return Validator.match(CheckPrefix);
1194}
1195
1196static bool ValidateCheckPrefixes() {
1197  StringSet<> PrefixSet;
1198
1199  for (prefix_iterator I = CheckPrefixes.begin(), E = CheckPrefixes.end();
1200       I != E; ++I) {
1201    StringRef Prefix(*I);
1202
1203    if (!PrefixSet.insert(Prefix))
1204      return false;
1205
1206    if (!ValidateCheckPrefix(Prefix))
1207      return false;
1208  }
1209
1210  return true;
1211}
1212
1213// I don't think there's a way to specify an initial value for cl::list,
1214// so if nothing was specified, add the default
1215static void AddCheckPrefixIfNeeded() {
1216  if (CheckPrefixes.empty())
1217    CheckPrefixes.push_back("CHECK");
1218}
1219
1220int main(int argc, char **argv) {
1221  sys::PrintStackTraceOnErrorSignal();
1222  PrettyStackTraceProgram X(argc, argv);
1223  cl::ParseCommandLineOptions(argc, argv);
1224
1225  if (!ValidateCheckPrefixes()) {
1226    errs() << "Supplied check-prefix is invalid! Prefixes must be unique and "
1227              "start with a letter and contain only alphanumeric characters, "
1228              "hyphens and underscores\n";
1229    return 2;
1230  }
1231
1232  AddCheckPrefixIfNeeded();
1233
1234  SourceMgr SM;
1235
1236  // Read the expected strings from the check file.
1237  std::vector<CheckString> CheckStrings;
1238  if (ReadCheckFile(SM, CheckStrings))
1239    return 2;
1240
1241  // Open the file to check and add it to SourceMgr.
1242  OwningPtr<MemoryBuffer> File;
1243  if (error_code ec =
1244        MemoryBuffer::getFileOrSTDIN(InputFilename, File)) {
1245    errs() << "Could not open input file '" << InputFilename << "': "
1246           << ec.message() << '\n';
1247    return 2;
1248  }
1249
1250  if (File->getBufferSize() == 0) {
1251    errs() << "FileCheck error: '" << InputFilename << "' is empty.\n";
1252    return 2;
1253  }
1254
1255  // Remove duplicate spaces in the input file if requested.
1256  // Remove DOS style line endings.
1257  MemoryBuffer *F =
1258    CanonicalizeInputFile(File.take(), NoCanonicalizeWhiteSpace);
1259
1260  SM.AddNewSourceBuffer(F, SMLoc());
1261
1262  /// VariableTable - This holds all the current filecheck variables.
1263  StringMap<StringRef> VariableTable;
1264
1265  // Check that we have all of the expected strings, in order, in the input
1266  // file.
1267  StringRef Buffer = F->getBuffer();
1268
1269  bool hasError = false;
1270
1271  unsigned i = 0, j = 0, e = CheckStrings.size();
1272
1273  while (true) {
1274    StringRef CheckRegion;
1275    if (j == e) {
1276      CheckRegion = Buffer;
1277    } else {
1278      const CheckString &CheckLabelStr = CheckStrings[j];
1279      if (CheckLabelStr.CheckTy != Check::CheckLabel) {
1280        ++j;
1281        continue;
1282      }
1283
1284      // Scan to next CHECK-LABEL match, ignoring CHECK-NOT and CHECK-DAG
1285      size_t MatchLabelLen = 0;
1286      size_t MatchLabelPos = CheckLabelStr.Check(SM, Buffer, true,
1287                                                 MatchLabelLen, VariableTable);
1288      if (MatchLabelPos == StringRef::npos) {
1289        hasError = true;
1290        break;
1291      }
1292
1293      CheckRegion = Buffer.substr(0, MatchLabelPos + MatchLabelLen);
1294      Buffer = Buffer.substr(MatchLabelPos + MatchLabelLen);
1295      ++j;
1296    }
1297
1298    for ( ; i != j; ++i) {
1299      const CheckString &CheckStr = CheckStrings[i];
1300
1301      // Check each string within the scanned region, including a second check
1302      // of any final CHECK-LABEL (to verify CHECK-NOT and CHECK-DAG)
1303      size_t MatchLen = 0;
1304      size_t MatchPos = CheckStr.Check(SM, CheckRegion, false, MatchLen,
1305                                       VariableTable);
1306
1307      if (MatchPos == StringRef::npos) {
1308        hasError = true;
1309        i = j;
1310        break;
1311      }
1312
1313      CheckRegion = CheckRegion.substr(MatchPos + MatchLen);
1314    }
1315
1316    if (j == e)
1317      break;
1318  }
1319
1320  return hasError ? 1 : 0;
1321}
1322