1ee3c74fbSChris Lattner //===- FileCheck.cpp - Check that File's Contents match what is expected --===//
2ee3c74fbSChris Lattner //
3ee3c74fbSChris Lattner //                     The LLVM Compiler Infrastructure
4ee3c74fbSChris Lattner //
5ee3c74fbSChris Lattner // This file is distributed under the University of Illinois Open Source
6ee3c74fbSChris Lattner // License. See LICENSE.TXT for details.
7ee3c74fbSChris Lattner //
8ee3c74fbSChris Lattner //===----------------------------------------------------------------------===//
9ee3c74fbSChris Lattner //
10ee3c74fbSChris Lattner // FileCheck does a line-by line check of a file that validates whether it
11ee3c74fbSChris Lattner // contains the expected content.  This is useful for regression tests etc.
12ee3c74fbSChris Lattner //
13ee3c74fbSChris Lattner // This program exits with an error status of 2 on error, exit status of 0 if
14ee3c74fbSChris Lattner // the file matched the expected contents, and exit status of 1 if it did not
15ee3c74fbSChris Lattner // contain the expected contents.
16ee3c74fbSChris Lattner //
17ee3c74fbSChris Lattner //===----------------------------------------------------------------------===//
18ee3c74fbSChris Lattner 
1939a0ffc3SMichael J. Spencer #include "llvm/ADT/OwningPtr.h"
20ee3c74fbSChris Lattner #include "llvm/Support/CommandLine.h"
21ee3c74fbSChris Lattner #include "llvm/Support/MemoryBuffer.h"
22ee3c74fbSChris Lattner #include "llvm/Support/PrettyStackTrace.h"
23f08d2db9SChris Lattner #include "llvm/Support/Regex.h"
24ee3c74fbSChris Lattner #include "llvm/Support/SourceMgr.h"
25ee3c74fbSChris Lattner #include "llvm/Support/raw_ostream.h"
26447762daSMichael J. Spencer #include "llvm/Support/Signals.h"
277b6fef82SMichael J. Spencer #include "llvm/Support/system_error.h"
28e0ef65abSDaniel Dunbar #include "llvm/ADT/SmallString.h"
2992987fb3SAlexander Kornienko #include "llvm/ADT/StringExtras.h"
308879e06dSChris Lattner #include "llvm/ADT/StringMap.h"
318879e06dSChris Lattner #include <algorithm>
32ee3c74fbSChris Lattner using namespace llvm;
33ee3c74fbSChris Lattner 
34ee3c74fbSChris Lattner static cl::opt<std::string>
35ee3c74fbSChris Lattner CheckFilename(cl::Positional, cl::desc("<check-file>"), cl::Required);
36ee3c74fbSChris Lattner 
37ee3c74fbSChris Lattner static cl::opt<std::string>
38ee3c74fbSChris Lattner InputFilename("input-file", cl::desc("File to check (defaults to stdin)"),
39ee3c74fbSChris Lattner               cl::init("-"), cl::value_desc("filename"));
40ee3c74fbSChris Lattner 
41ee3c74fbSChris Lattner static cl::opt<std::string>
42ee3c74fbSChris Lattner CheckPrefix("check-prefix", cl::init("CHECK"),
43ee3c74fbSChris Lattner             cl::desc("Prefix to use from check file (defaults to 'CHECK')"));
44ee3c74fbSChris Lattner 
452c3e5cdfSChris Lattner static cl::opt<bool>
462c3e5cdfSChris Lattner NoCanonicalizeWhiteSpace("strict-whitespace",
472c3e5cdfSChris Lattner               cl::desc("Do not treat all horizontal whitespace as equivalent"));
482c3e5cdfSChris Lattner 
4974d50731SChris Lattner //===----------------------------------------------------------------------===//
5074d50731SChris Lattner // Pattern Handling Code.
5174d50731SChris Lattner //===----------------------------------------------------------------------===//
5274d50731SChris Lattner 
533b40b445SChris Lattner class Pattern {
540a4c44bdSChris Lattner   SMLoc PatternLoc;
550a4c44bdSChris Lattner 
56eba55822SJakob Stoklund Olesen   /// MatchEOF - When set, this pattern only matches the end of file. This is
57eba55822SJakob Stoklund Olesen   /// used for trailing CHECK-NOTs.
58eba55822SJakob Stoklund Olesen   bool MatchEOF;
59eba55822SJakob Stoklund Olesen 
60b16ab0c4SChris Lattner   /// FixedStr - If non-empty, this pattern is a fixed string match with the
61b16ab0c4SChris Lattner   /// specified fixed string.
62221460e0SChris Lattner   StringRef FixedStr;
63b16ab0c4SChris Lattner 
64b16ab0c4SChris Lattner   /// RegEx - If non-empty, this is a regex pattern.
65b16ab0c4SChris Lattner   std::string RegExStr;
668879e06dSChris Lattner 
6792987fb3SAlexander Kornienko   /// \brief Contains the number of line this pattern is in.
6892987fb3SAlexander Kornienko   unsigned LineNumber;
6992987fb3SAlexander Kornienko 
708879e06dSChris Lattner   /// VariableUses - Entries in this vector map to uses of a variable in the
718879e06dSChris Lattner   /// pattern, e.g. "foo[[bar]]baz".  In this case, the RegExStr will contain
728879e06dSChris Lattner   /// "foobaz" and we'll get an entry in this vector that tells us to insert the
738879e06dSChris Lattner   /// value of bar at offset 3.
748879e06dSChris Lattner   std::vector<std::pair<StringRef, unsigned> > VariableUses;
758879e06dSChris Lattner 
768879e06dSChris Lattner   /// VariableDefs - Entries in this vector map to definitions of a variable in
778879e06dSChris Lattner   /// the pattern, e.g. "foo[[bar:.*]]baz".  In this case, the RegExStr will
788879e06dSChris Lattner   /// contain "foo(.*)baz" and VariableDefs will contain the pair "bar",1.  The
798879e06dSChris Lattner   /// index indicates what parenthesized value captures the variable value.
808879e06dSChris Lattner   std::vector<std::pair<StringRef, unsigned> > VariableDefs;
818879e06dSChris Lattner 
823b40b445SChris Lattner public:
833b40b445SChris Lattner 
84eba55822SJakob Stoklund Olesen   Pattern(bool matchEOF = false) : MatchEOF(matchEOF) { }
8574d50731SChris Lattner 
8692987fb3SAlexander Kornienko   bool ParsePattern(StringRef PatternStr, SourceMgr &SM, unsigned LineNumber);
873b40b445SChris Lattner 
883b40b445SChris Lattner   /// Match - Match the pattern string against the input buffer Buffer.  This
893b40b445SChris Lattner   /// returns the position that is matched or npos if there is no match.  If
903b40b445SChris Lattner   /// there is a match, the size of the matched string is returned in MatchLen.
918879e06dSChris Lattner   ///
928879e06dSChris Lattner   /// The VariableTable StringMap provides the current values of filecheck
938879e06dSChris Lattner   /// variables and is updated if this match defines new values.
948879e06dSChris Lattner   size_t Match(StringRef Buffer, size_t &MatchLen,
958879e06dSChris Lattner                StringMap<StringRef> &VariableTable) const;
96b16ab0c4SChris Lattner 
97e0ef65abSDaniel Dunbar   /// PrintFailureInfo - Print additional information about a failure to match
98e0ef65abSDaniel Dunbar   /// involving this pattern.
99e0ef65abSDaniel Dunbar   void PrintFailureInfo(const SourceMgr &SM, StringRef Buffer,
100e0ef65abSDaniel Dunbar                         const StringMap<StringRef> &VariableTable) const;
101e0ef65abSDaniel Dunbar 
102b16ab0c4SChris Lattner private:
1038879e06dSChris Lattner   static void AddFixedStringToRegEx(StringRef FixedStr, std::string &TheStr);
1048879e06dSChris Lattner   bool AddRegExToRegEx(StringRef RegExStr, unsigned &CurParen, SourceMgr &SM);
105fd29d886SDaniel Dunbar 
106fd29d886SDaniel Dunbar   /// ComputeMatchDistance - Compute an arbitrary estimate for the quality of
107fd29d886SDaniel Dunbar   /// matching this pattern at the start of \arg Buffer; a distance of zero
108fd29d886SDaniel Dunbar   /// should correspond to a perfect match.
109fd29d886SDaniel Dunbar   unsigned ComputeMatchDistance(StringRef Buffer,
110fd29d886SDaniel Dunbar                                const StringMap<StringRef> &VariableTable) const;
11192987fb3SAlexander Kornienko 
11292987fb3SAlexander Kornienko   /// \brief Evaluates expression and stores the result to \p Value.
11392987fb3SAlexander Kornienko   /// \return true on success. false when the expression has invalid syntax.
11492987fb3SAlexander Kornienko   bool EvaluateExpression(StringRef Expr, std::string &Value) const;
1153b40b445SChris Lattner };
1163b40b445SChris Lattner 
1178879e06dSChris Lattner 
11892987fb3SAlexander Kornienko bool Pattern::ParsePattern(StringRef PatternStr, SourceMgr &SM,
11992987fb3SAlexander Kornienko                            unsigned LineNumber) {
12092987fb3SAlexander Kornienko   this->LineNumber = LineNumber;
1210a4c44bdSChris Lattner   PatternLoc = SMLoc::getFromPointer(PatternStr.data());
1220a4c44bdSChris Lattner 
12374d50731SChris Lattner   // Ignore trailing whitespace.
12474d50731SChris Lattner   while (!PatternStr.empty() &&
12574d50731SChris Lattner          (PatternStr.back() == ' ' || PatternStr.back() == '\t'))
12674d50731SChris Lattner     PatternStr = PatternStr.substr(0, PatternStr.size()-1);
12774d50731SChris Lattner 
12874d50731SChris Lattner   // Check that there is something on the line.
12974d50731SChris Lattner   if (PatternStr.empty()) {
13003b80a40SChris Lattner     SM.PrintMessage(PatternLoc, SourceMgr::DK_Error,
13103b80a40SChris Lattner                     "found empty check string with prefix '" +
13203b80a40SChris Lattner                     CheckPrefix+":'");
13374d50731SChris Lattner     return true;
13474d50731SChris Lattner   }
13574d50731SChris Lattner 
136221460e0SChris Lattner   // Check to see if this is a fixed string, or if it has regex pieces.
137d9466967STed Kremenek   if (PatternStr.size() < 2 ||
1388879e06dSChris Lattner       (PatternStr.find("{{") == StringRef::npos &&
1398879e06dSChris Lattner        PatternStr.find("[[") == StringRef::npos)) {
140221460e0SChris Lattner     FixedStr = PatternStr;
141221460e0SChris Lattner     return false;
142221460e0SChris Lattner   }
143221460e0SChris Lattner 
1448879e06dSChris Lattner   // Paren value #0 is for the fully matched string.  Any new parenthesized
14553e0679dSChris Lattner   // values add from there.
1468879e06dSChris Lattner   unsigned CurParen = 1;
1478879e06dSChris Lattner 
148b16ab0c4SChris Lattner   // Otherwise, there is at least one regex piece.  Build up the regex pattern
149b16ab0c4SChris Lattner   // by escaping scary characters in fixed strings, building up one big regex.
150f08d2db9SChris Lattner   while (!PatternStr.empty()) {
1518879e06dSChris Lattner     // RegEx matches.
15253e0679dSChris Lattner     if (PatternStr.startswith("{{")) {
15374d50731SChris Lattner 
154f08d2db9SChris Lattner       // Otherwise, this is the start of a regex match.  Scan for the }}.
155f08d2db9SChris Lattner       size_t End = PatternStr.find("}}");
156f08d2db9SChris Lattner       if (End == StringRef::npos) {
157f08d2db9SChris Lattner         SM.PrintMessage(SMLoc::getFromPointer(PatternStr.data()),
15803b80a40SChris Lattner                         SourceMgr::DK_Error,
15903b80a40SChris Lattner                         "found start of regex string with no end '}}'");
160f08d2db9SChris Lattner         return true;
161f08d2db9SChris Lattner       }
162f08d2db9SChris Lattner 
163e53c95f1SChris Lattner       // Enclose {{}} patterns in parens just like [[]] even though we're not
164e53c95f1SChris Lattner       // capturing the result for any purpose.  This is required in case the
165e53c95f1SChris Lattner       // expression contains an alternation like: CHECK:  abc{{x|z}}def.  We
166e53c95f1SChris Lattner       // want this to turn into: "abc(x|z)def" not "abcx|zdef".
167e53c95f1SChris Lattner       RegExStr += '(';
168e53c95f1SChris Lattner       ++CurParen;
169e53c95f1SChris Lattner 
1708879e06dSChris Lattner       if (AddRegExToRegEx(PatternStr.substr(2, End-2), CurParen, SM))
1718879e06dSChris Lattner         return true;
172e53c95f1SChris Lattner       RegExStr += ')';
17353e0679dSChris Lattner 
1748879e06dSChris Lattner       PatternStr = PatternStr.substr(End+2);
1758879e06dSChris Lattner       continue;
1768879e06dSChris Lattner     }
1778879e06dSChris Lattner 
1788879e06dSChris Lattner     // Named RegEx matches.  These are of two forms: [[foo:.*]] which matches .*
1798879e06dSChris Lattner     // (or some other regex) and assigns it to the FileCheck variable 'foo'. The
1808879e06dSChris Lattner     // second form is [[foo]] which is a reference to foo.  The variable name
18157cb733bSDaniel Dunbar     // itself must be of the form "[a-zA-Z_][0-9a-zA-Z_]*", otherwise we reject
1828879e06dSChris Lattner     // it.  This is to catch some common errors.
18353e0679dSChris Lattner     if (PatternStr.startswith("[[")) {
1848879e06dSChris Lattner       // Verify that it is terminated properly.
1858879e06dSChris Lattner       size_t End = PatternStr.find("]]");
1868879e06dSChris Lattner       if (End == StringRef::npos) {
1878879e06dSChris Lattner         SM.PrintMessage(SMLoc::getFromPointer(PatternStr.data()),
18803b80a40SChris Lattner                         SourceMgr::DK_Error,
18903b80a40SChris Lattner                         "invalid named regex reference, no ]] found");
190f08d2db9SChris Lattner         return true;
191f08d2db9SChris Lattner       }
192f08d2db9SChris Lattner 
1938879e06dSChris Lattner       StringRef MatchStr = PatternStr.substr(2, End-2);
194f08d2db9SChris Lattner       PatternStr = PatternStr.substr(End+2);
1958879e06dSChris Lattner 
1968879e06dSChris Lattner       // Get the regex name (e.g. "foo").
1978879e06dSChris Lattner       size_t NameEnd = MatchStr.find(':');
1988879e06dSChris Lattner       StringRef Name = MatchStr.substr(0, NameEnd);
1998879e06dSChris Lattner 
2008879e06dSChris Lattner       if (Name.empty()) {
20103b80a40SChris Lattner         SM.PrintMessage(SMLoc::getFromPointer(Name.data()), SourceMgr::DK_Error,
20203b80a40SChris Lattner                         "invalid name in named regex: empty name");
2038879e06dSChris Lattner         return true;
2048879e06dSChris Lattner       }
2058879e06dSChris Lattner 
20692987fb3SAlexander Kornienko       // Verify that the name/expression is well formed. FileCheck currently
20792987fb3SAlexander Kornienko       // supports @LINE, @LINE+number, @LINE-number expressions. The check here
20892987fb3SAlexander Kornienko       // is relaxed, more strict check is performed in \c EvaluateExpression.
20992987fb3SAlexander Kornienko       bool IsExpression = false;
21092987fb3SAlexander Kornienko       for (unsigned i = 0, e = Name.size(); i != e; ++i) {
21192987fb3SAlexander Kornienko         if (i == 0 && Name[i] == '@') {
21292987fb3SAlexander Kornienko           if (NameEnd != StringRef::npos) {
21392987fb3SAlexander Kornienko             SM.PrintMessage(SMLoc::getFromPointer(Name.data()),
21492987fb3SAlexander Kornienko                             SourceMgr::DK_Error,
21592987fb3SAlexander Kornienko                             "invalid name in named regex definition");
21692987fb3SAlexander Kornienko             return true;
21792987fb3SAlexander Kornienko           }
21892987fb3SAlexander Kornienko           IsExpression = true;
21992987fb3SAlexander Kornienko           continue;
22092987fb3SAlexander Kornienko         }
22192987fb3SAlexander Kornienko         if (Name[i] != '_' && !isalnum(Name[i]) &&
22292987fb3SAlexander Kornienko             (!IsExpression || (Name[i] != '+' && Name[i] != '-'))) {
2238879e06dSChris Lattner           SM.PrintMessage(SMLoc::getFromPointer(Name.data()+i),
22403b80a40SChris Lattner                           SourceMgr::DK_Error, "invalid name in named regex");
2258879e06dSChris Lattner           return true;
2268879e06dSChris Lattner         }
22792987fb3SAlexander Kornienko       }
2288879e06dSChris Lattner 
2298879e06dSChris Lattner       // Name can't start with a digit.
2308879e06dSChris Lattner       if (isdigit(Name[0])) {
23103b80a40SChris Lattner         SM.PrintMessage(SMLoc::getFromPointer(Name.data()), SourceMgr::DK_Error,
23203b80a40SChris Lattner                         "invalid name in named regex");
2338879e06dSChris Lattner         return true;
2348879e06dSChris Lattner       }
2358879e06dSChris Lattner 
2368879e06dSChris Lattner       // Handle [[foo]].
2378879e06dSChris Lattner       if (NameEnd == StringRef::npos) {
2388879e06dSChris Lattner         VariableUses.push_back(std::make_pair(Name, RegExStr.size()));
2398879e06dSChris Lattner         continue;
2408879e06dSChris Lattner       }
2418879e06dSChris Lattner 
2428879e06dSChris Lattner       // Handle [[foo:.*]].
2438879e06dSChris Lattner       VariableDefs.push_back(std::make_pair(Name, CurParen));
2448879e06dSChris Lattner       RegExStr += '(';
2458879e06dSChris Lattner       ++CurParen;
2468879e06dSChris Lattner 
2478879e06dSChris Lattner       if (AddRegExToRegEx(MatchStr.substr(NameEnd+1), CurParen, SM))
2488879e06dSChris Lattner         return true;
2498879e06dSChris Lattner 
2508879e06dSChris Lattner       RegExStr += ')';
2518879e06dSChris Lattner     }
2528879e06dSChris Lattner 
2538879e06dSChris Lattner     // Handle fixed string matches.
2548879e06dSChris Lattner     // Find the end, which is the start of the next regex.
2558879e06dSChris Lattner     size_t FixedMatchEnd = PatternStr.find("{{");
2568879e06dSChris Lattner     FixedMatchEnd = std::min(FixedMatchEnd, PatternStr.find("[["));
2578879e06dSChris Lattner     AddFixedStringToRegEx(PatternStr.substr(0, FixedMatchEnd), RegExStr);
2588879e06dSChris Lattner     PatternStr = PatternStr.substr(FixedMatchEnd);
259f08d2db9SChris Lattner   }
260f08d2db9SChris Lattner 
26174d50731SChris Lattner   return false;
26274d50731SChris Lattner }
26374d50731SChris Lattner 
2648879e06dSChris Lattner void Pattern::AddFixedStringToRegEx(StringRef FixedStr, std::string &TheStr) {
265b16ab0c4SChris Lattner   // Add the characters from FixedStr to the regex, escaping as needed.  This
266b16ab0c4SChris Lattner   // avoids "leaning toothpicks" in common patterns.
267b16ab0c4SChris Lattner   for (unsigned i = 0, e = FixedStr.size(); i != e; ++i) {
268b16ab0c4SChris Lattner     switch (FixedStr[i]) {
269b16ab0c4SChris Lattner     // These are the special characters matched in "p_ere_exp".
270b16ab0c4SChris Lattner     case '(':
271b16ab0c4SChris Lattner     case ')':
272b16ab0c4SChris Lattner     case '^':
273b16ab0c4SChris Lattner     case '$':
274b16ab0c4SChris Lattner     case '|':
275b16ab0c4SChris Lattner     case '*':
276b16ab0c4SChris Lattner     case '+':
277b16ab0c4SChris Lattner     case '?':
278b16ab0c4SChris Lattner     case '.':
279b16ab0c4SChris Lattner     case '[':
280b16ab0c4SChris Lattner     case '\\':
281b16ab0c4SChris Lattner     case '{':
2828879e06dSChris Lattner       TheStr += '\\';
283b16ab0c4SChris Lattner       // FALL THROUGH.
284b16ab0c4SChris Lattner     default:
2858879e06dSChris Lattner       TheStr += FixedStr[i];
286b16ab0c4SChris Lattner       break;
287b16ab0c4SChris Lattner     }
288b16ab0c4SChris Lattner   }
289b16ab0c4SChris Lattner }
290b16ab0c4SChris Lattner 
2918879e06dSChris Lattner bool Pattern::AddRegExToRegEx(StringRef RegexStr, unsigned &CurParen,
2928879e06dSChris Lattner                               SourceMgr &SM) {
2938879e06dSChris Lattner   Regex R(RegexStr);
2948879e06dSChris Lattner   std::string Error;
2958879e06dSChris Lattner   if (!R.isValid(Error)) {
29603b80a40SChris Lattner     SM.PrintMessage(SMLoc::getFromPointer(RegexStr.data()), SourceMgr::DK_Error,
29703b80a40SChris Lattner                     "invalid regex: " + Error);
2988879e06dSChris Lattner     return true;
2998879e06dSChris Lattner   }
3008879e06dSChris Lattner 
3018879e06dSChris Lattner   RegExStr += RegexStr.str();
3028879e06dSChris Lattner   CurParen += R.getNumMatches();
3038879e06dSChris Lattner   return false;
3048879e06dSChris Lattner }
305b16ab0c4SChris Lattner 
30692987fb3SAlexander Kornienko bool Pattern::EvaluateExpression(StringRef Expr, std::string &Value) const {
30792987fb3SAlexander Kornienko   // The only supported expression is @LINE([\+-]\d+)?
30892987fb3SAlexander Kornienko   if (!Expr.startswith("@LINE"))
30992987fb3SAlexander Kornienko     return false;
31092987fb3SAlexander Kornienko   Expr = Expr.substr(StringRef("@LINE").size());
31192987fb3SAlexander Kornienko   int Offset = 0;
31292987fb3SAlexander Kornienko   if (!Expr.empty()) {
31392987fb3SAlexander Kornienko     if (Expr[0] == '+')
31492987fb3SAlexander Kornienko       Expr = Expr.substr(1);
31592987fb3SAlexander Kornienko     else if (Expr[0] != '-')
31692987fb3SAlexander Kornienko       return false;
31792987fb3SAlexander Kornienko     if (Expr.getAsInteger(10, Offset))
31892987fb3SAlexander Kornienko       return false;
31992987fb3SAlexander Kornienko   }
32092987fb3SAlexander Kornienko   Value = llvm::itostr(LineNumber + Offset);
32192987fb3SAlexander Kornienko   return true;
32292987fb3SAlexander Kornienko }
32392987fb3SAlexander Kornienko 
324f08d2db9SChris Lattner /// Match - Match the pattern string against the input buffer Buffer.  This
325f08d2db9SChris Lattner /// returns the position that is matched or npos if there is no match.  If
326f08d2db9SChris Lattner /// there is a match, the size of the matched string is returned in MatchLen.
3278879e06dSChris Lattner size_t Pattern::Match(StringRef Buffer, size_t &MatchLen,
3288879e06dSChris Lattner                       StringMap<StringRef> &VariableTable) const {
329eba55822SJakob Stoklund Olesen   // If this is the EOF pattern, match it immediately.
330eba55822SJakob Stoklund Olesen   if (MatchEOF) {
331eba55822SJakob Stoklund Olesen     MatchLen = 0;
332eba55822SJakob Stoklund Olesen     return Buffer.size();
333eba55822SJakob Stoklund Olesen   }
334eba55822SJakob Stoklund Olesen 
335221460e0SChris Lattner   // If this is a fixed string pattern, just match it now.
336221460e0SChris Lattner   if (!FixedStr.empty()) {
337221460e0SChris Lattner     MatchLen = FixedStr.size();
338221460e0SChris Lattner     return Buffer.find(FixedStr);
339221460e0SChris Lattner   }
340221460e0SChris Lattner 
341b16ab0c4SChris Lattner   // Regex match.
3428879e06dSChris Lattner 
3438879e06dSChris Lattner   // If there are variable uses, we need to create a temporary string with the
3448879e06dSChris Lattner   // actual value.
3458879e06dSChris Lattner   StringRef RegExToMatch = RegExStr;
3468879e06dSChris Lattner   std::string TmpStr;
3478879e06dSChris Lattner   if (!VariableUses.empty()) {
3488879e06dSChris Lattner     TmpStr = RegExStr;
3498879e06dSChris Lattner 
3508879e06dSChris Lattner     unsigned InsertOffset = 0;
3518879e06dSChris Lattner     for (unsigned i = 0, e = VariableUses.size(); i != e; ++i) {
35292987fb3SAlexander Kornienko       std::string Value;
35392987fb3SAlexander Kornienko 
35492987fb3SAlexander Kornienko       if (VariableUses[i].first[0] == '@') {
35592987fb3SAlexander Kornienko         if (!EvaluateExpression(VariableUses[i].first, Value))
35692987fb3SAlexander Kornienko           return StringRef::npos;
35792987fb3SAlexander Kornienko       } else {
358e0ef65abSDaniel Dunbar         StringMap<StringRef>::iterator it =
359e0ef65abSDaniel Dunbar           VariableTable.find(VariableUses[i].first);
360e0ef65abSDaniel Dunbar         // If the variable is undefined, return an error.
361e0ef65abSDaniel Dunbar         if (it == VariableTable.end())
362e0ef65abSDaniel Dunbar           return StringRef::npos;
363e0ef65abSDaniel Dunbar 
3648879e06dSChris Lattner         // Look up the value and escape it so that we can plop it into the regex.
365e0ef65abSDaniel Dunbar         AddFixedStringToRegEx(it->second, Value);
36692987fb3SAlexander Kornienko       }
3678879e06dSChris Lattner 
3688879e06dSChris Lattner       // Plop it into the regex at the adjusted offset.
3698879e06dSChris Lattner       TmpStr.insert(TmpStr.begin()+VariableUses[i].second+InsertOffset,
3708879e06dSChris Lattner                     Value.begin(), Value.end());
3718879e06dSChris Lattner       InsertOffset += Value.size();
3728879e06dSChris Lattner     }
3738879e06dSChris Lattner 
3748879e06dSChris Lattner     // Match the newly constructed regex.
3758879e06dSChris Lattner     RegExToMatch = TmpStr;
3768879e06dSChris Lattner   }
3778879e06dSChris Lattner 
3788879e06dSChris Lattner 
379b16ab0c4SChris Lattner   SmallVector<StringRef, 4> MatchInfo;
3808879e06dSChris Lattner   if (!Regex(RegExToMatch, Regex::Newline).match(Buffer, &MatchInfo))
381f08d2db9SChris Lattner     return StringRef::npos;
382b16ab0c4SChris Lattner 
383b16ab0c4SChris Lattner   // Successful regex match.
384b16ab0c4SChris Lattner   assert(!MatchInfo.empty() && "Didn't get any match");
385b16ab0c4SChris Lattner   StringRef FullMatch = MatchInfo[0];
386b16ab0c4SChris Lattner 
3878879e06dSChris Lattner   // If this defines any variables, remember their values.
3888879e06dSChris Lattner   for (unsigned i = 0, e = VariableDefs.size(); i != e; ++i) {
3898879e06dSChris Lattner     assert(VariableDefs[i].second < MatchInfo.size() &&
3908879e06dSChris Lattner            "Internal paren error");
3918879e06dSChris Lattner     VariableTable[VariableDefs[i].first] = MatchInfo[VariableDefs[i].second];
3920a4c44bdSChris Lattner   }
3930a4c44bdSChris Lattner 
394b16ab0c4SChris Lattner   MatchLen = FullMatch.size();
395b16ab0c4SChris Lattner   return FullMatch.data()-Buffer.data();
396f08d2db9SChris Lattner }
397f08d2db9SChris Lattner 
398fd29d886SDaniel Dunbar unsigned Pattern::ComputeMatchDistance(StringRef Buffer,
399fd29d886SDaniel Dunbar                               const StringMap<StringRef> &VariableTable) const {
400fd29d886SDaniel Dunbar   // Just compute the number of matching characters. For regular expressions, we
401fd29d886SDaniel Dunbar   // just compare against the regex itself and hope for the best.
402fd29d886SDaniel Dunbar   //
403fd29d886SDaniel Dunbar   // FIXME: One easy improvement here is have the regex lib generate a single
404fd29d886SDaniel Dunbar   // example regular expression which matches, and use that as the example
405fd29d886SDaniel Dunbar   // string.
406fd29d886SDaniel Dunbar   StringRef ExampleString(FixedStr);
407fd29d886SDaniel Dunbar   if (ExampleString.empty())
408fd29d886SDaniel Dunbar     ExampleString = RegExStr;
409fd29d886SDaniel Dunbar 
410e9aa36c8SDaniel Dunbar   // Only compare up to the first line in the buffer, or the string size.
411e9aa36c8SDaniel Dunbar   StringRef BufferPrefix = Buffer.substr(0, ExampleString.size());
412e9aa36c8SDaniel Dunbar   BufferPrefix = BufferPrefix.split('\n').first;
413e9aa36c8SDaniel Dunbar   return BufferPrefix.edit_distance(ExampleString);
414fd29d886SDaniel Dunbar }
415fd29d886SDaniel Dunbar 
416e0ef65abSDaniel Dunbar void Pattern::PrintFailureInfo(const SourceMgr &SM, StringRef Buffer,
417e0ef65abSDaniel Dunbar                                const StringMap<StringRef> &VariableTable) const{
418e0ef65abSDaniel Dunbar   // If this was a regular expression using variables, print the current
419e0ef65abSDaniel Dunbar   // variable values.
420e0ef65abSDaniel Dunbar   if (!VariableUses.empty()) {
421e0ef65abSDaniel Dunbar     for (unsigned i = 0, e = VariableUses.size(); i != e; ++i) {
422e0ef65abSDaniel Dunbar       SmallString<256> Msg;
423e0ef65abSDaniel Dunbar       raw_svector_ostream OS(Msg);
42492987fb3SAlexander Kornienko       StringRef Var = VariableUses[i].first;
42592987fb3SAlexander Kornienko       if (Var[0] == '@') {
42692987fb3SAlexander Kornienko         std::string Value;
42792987fb3SAlexander Kornienko         if (EvaluateExpression(Var, Value)) {
42892987fb3SAlexander Kornienko           OS << "with expression \"";
42992987fb3SAlexander Kornienko           OS.write_escaped(Var) << "\" equal to \"";
43092987fb3SAlexander Kornienko           OS.write_escaped(Value) << "\"";
43192987fb3SAlexander Kornienko         } else {
43292987fb3SAlexander Kornienko           OS << "uses incorrect expression \"";
43392987fb3SAlexander Kornienko           OS.write_escaped(Var) << "\"";
43492987fb3SAlexander Kornienko         }
43592987fb3SAlexander Kornienko       } else {
43692987fb3SAlexander Kornienko         StringMap<StringRef>::const_iterator it = VariableTable.find(Var);
437e0ef65abSDaniel Dunbar 
438e0ef65abSDaniel Dunbar         // Check for undefined variable references.
439e0ef65abSDaniel Dunbar         if (it == VariableTable.end()) {
440e0ef65abSDaniel Dunbar           OS << "uses undefined variable \"";
44192987fb3SAlexander Kornienko           OS.write_escaped(Var) << "\"";
442e0ef65abSDaniel Dunbar         } else {
443e0ef65abSDaniel Dunbar           OS << "with variable \"";
444e0ef65abSDaniel Dunbar           OS.write_escaped(Var) << "\" equal to \"";
445e0ef65abSDaniel Dunbar           OS.write_escaped(it->second) << "\"";
446e0ef65abSDaniel Dunbar         }
44792987fb3SAlexander Kornienko       }
448e0ef65abSDaniel Dunbar 
44903b80a40SChris Lattner       SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()), SourceMgr::DK_Note,
45003b80a40SChris Lattner                       OS.str());
451e0ef65abSDaniel Dunbar     }
452e0ef65abSDaniel Dunbar   }
453fd29d886SDaniel Dunbar 
454fd29d886SDaniel Dunbar   // Attempt to find the closest/best fuzzy match.  Usually an error happens
455fd29d886SDaniel Dunbar   // because some string in the output didn't exactly match. In these cases, we
456fd29d886SDaniel Dunbar   // would like to show the user a best guess at what "should have" matched, to
457fd29d886SDaniel Dunbar   // save them having to actually check the input manually.
458fd29d886SDaniel Dunbar   size_t NumLinesForward = 0;
459fd29d886SDaniel Dunbar   size_t Best = StringRef::npos;
460fd29d886SDaniel Dunbar   double BestQuality = 0;
461fd29d886SDaniel Dunbar 
462fd29d886SDaniel Dunbar   // Use an arbitrary 4k limit on how far we will search.
4632bf486ebSDan Gohman   for (size_t i = 0, e = std::min(size_t(4096), Buffer.size()); i != e; ++i) {
464fd29d886SDaniel Dunbar     if (Buffer[i] == '\n')
465fd29d886SDaniel Dunbar       ++NumLinesForward;
466fd29d886SDaniel Dunbar 
467df22bbf7SDan Gohman     // Patterns have leading whitespace stripped, so skip whitespace when
468df22bbf7SDan Gohman     // looking for something which looks like a pattern.
469df22bbf7SDan Gohman     if (Buffer[i] == ' ' || Buffer[i] == '\t')
470df22bbf7SDan Gohman       continue;
471df22bbf7SDan Gohman 
472fd29d886SDaniel Dunbar     // Compute the "quality" of this match as an arbitrary combination of the
473fd29d886SDaniel Dunbar     // match distance and the number of lines skipped to get to this match.
474fd29d886SDaniel Dunbar     unsigned Distance = ComputeMatchDistance(Buffer.substr(i), VariableTable);
475fd29d886SDaniel Dunbar     double Quality = Distance + (NumLinesForward / 100.);
476fd29d886SDaniel Dunbar 
477fd29d886SDaniel Dunbar     if (Quality < BestQuality || Best == StringRef::npos) {
478fd29d886SDaniel Dunbar       Best = i;
479fd29d886SDaniel Dunbar       BestQuality = Quality;
480fd29d886SDaniel Dunbar     }
481fd29d886SDaniel Dunbar   }
482fd29d886SDaniel Dunbar 
483fd29d886SDaniel Dunbar   // Print the "possible intended match here" line if we found something
484c069cc8eSDaniel Dunbar   // reasonable and not equal to what we showed in the "scanning from here"
485c069cc8eSDaniel Dunbar   // line.
486c069cc8eSDaniel Dunbar   if (Best && Best != StringRef::npos && BestQuality < 50) {
487fd29d886SDaniel Dunbar       SM.PrintMessage(SMLoc::getFromPointer(Buffer.data() + Best),
48803b80a40SChris Lattner                       SourceMgr::DK_Note, "possible intended match here");
489fd29d886SDaniel Dunbar 
490fd29d886SDaniel Dunbar     // FIXME: If we wanted to be really friendly we would show why the match
491fd29d886SDaniel Dunbar     // failed, as it can be hard to spot simple one character differences.
492fd29d886SDaniel Dunbar   }
493e0ef65abSDaniel Dunbar }
49474d50731SChris Lattner 
49574d50731SChris Lattner //===----------------------------------------------------------------------===//
49674d50731SChris Lattner // Check Strings.
49774d50731SChris Lattner //===----------------------------------------------------------------------===//
4983b40b445SChris Lattner 
4993b40b445SChris Lattner /// CheckString - This is a check that we found in the input file.
5003b40b445SChris Lattner struct CheckString {
5013b40b445SChris Lattner   /// Pat - The pattern to match.
5023b40b445SChris Lattner   Pattern Pat;
50326cccfe1SChris Lattner 
50426cccfe1SChris Lattner   /// Loc - The location in the match file that the check string was specified.
50526cccfe1SChris Lattner   SMLoc Loc;
50626cccfe1SChris Lattner 
507da108b4eSChris Lattner   /// IsCheckNext - This is true if this is a CHECK-NEXT: directive (as opposed
508da108b4eSChris Lattner   /// to a CHECK: directive.
509da108b4eSChris Lattner   bool IsCheckNext;
510da108b4eSChris Lattner 
511236d2d5eSChris Lattner   /// NotStrings - These are all of the strings that are disallowed from
512236d2d5eSChris Lattner   /// occurring between this match string and the previous one (or start of
513236d2d5eSChris Lattner   /// file).
51474d50731SChris Lattner   std::vector<std::pair<SMLoc, Pattern> > NotStrings;
515236d2d5eSChris Lattner 
5163b40b445SChris Lattner   CheckString(const Pattern &P, SMLoc L, bool isCheckNext)
5173b40b445SChris Lattner     : Pat(P), Loc(L), IsCheckNext(isCheckNext) {}
51826cccfe1SChris Lattner };
51926cccfe1SChris Lattner 
520a2f8fc5aSChris Lattner /// CanonicalizeInputFile - Remove duplicate horizontal space from the specified
521a2f8fc5aSChris Lattner /// memory buffer, free it, and return a new one.
522a2f8fc5aSChris Lattner static MemoryBuffer *CanonicalizeInputFile(MemoryBuffer *MB) {
5230e45d24aSChris Lattner   SmallString<128> NewFile;
524a2f8fc5aSChris Lattner   NewFile.reserve(MB->getBufferSize());
525a2f8fc5aSChris Lattner 
526a2f8fc5aSChris Lattner   for (const char *Ptr = MB->getBufferStart(), *End = MB->getBufferEnd();
527a2f8fc5aSChris Lattner        Ptr != End; ++Ptr) {
528fd781bf0SNAKAMURA Takumi     // Eliminate trailing dosish \r.
529fd781bf0SNAKAMURA Takumi     if (Ptr <= End - 2 && Ptr[0] == '\r' && Ptr[1] == '\n') {
530fd781bf0SNAKAMURA Takumi       continue;
531fd781bf0SNAKAMURA Takumi     }
532fd781bf0SNAKAMURA Takumi 
5336e9fcd62SDmitri Gribenko     // If current char is not a horizontal whitespace, dump it to output as is.
534a2f8fc5aSChris Lattner     if (*Ptr != ' ' && *Ptr != '\t') {
535a2f8fc5aSChris Lattner       NewFile.push_back(*Ptr);
536a2f8fc5aSChris Lattner       continue;
537a2f8fc5aSChris Lattner     }
538a2f8fc5aSChris Lattner 
539a2f8fc5aSChris Lattner     // Otherwise, add one space and advance over neighboring space.
540a2f8fc5aSChris Lattner     NewFile.push_back(' ');
541a2f8fc5aSChris Lattner     while (Ptr+1 != End &&
542a2f8fc5aSChris Lattner            (Ptr[1] == ' ' || Ptr[1] == '\t'))
543a2f8fc5aSChris Lattner       ++Ptr;
544a2f8fc5aSChris Lattner   }
545a2f8fc5aSChris Lattner 
546a2f8fc5aSChris Lattner   // Free the old buffer and return a new one.
547a2f8fc5aSChris Lattner   MemoryBuffer *MB2 =
5480e45d24aSChris Lattner     MemoryBuffer::getMemBufferCopy(NewFile.str(), MB->getBufferIdentifier());
549a2f8fc5aSChris Lattner 
550a2f8fc5aSChris Lattner   delete MB;
551a2f8fc5aSChris Lattner   return MB2;
552a2f8fc5aSChris Lattner }
553a2f8fc5aSChris Lattner 
554ee3c74fbSChris Lattner 
555ee3c74fbSChris Lattner /// ReadCheckFile - Read the check file, which specifies the sequence of
556ee3c74fbSChris Lattner /// expected strings.  The strings are added to the CheckStrings vector.
557ee3c74fbSChris Lattner static bool ReadCheckFile(SourceMgr &SM,
55826cccfe1SChris Lattner                           std::vector<CheckString> &CheckStrings) {
559ee3c74fbSChris Lattner   // Open the check file, and tell SourceMgr about it.
56039a0ffc3SMichael J. Spencer   OwningPtr<MemoryBuffer> File;
56139a0ffc3SMichael J. Spencer   if (error_code ec =
56239a0ffc3SMichael J. Spencer         MemoryBuffer::getFileOrSTDIN(CheckFilename.c_str(), File)) {
563ee3c74fbSChris Lattner     errs() << "Could not open check file '" << CheckFilename << "': "
5647b6fef82SMichael J. Spencer            << ec.message() << '\n';
565ee3c74fbSChris Lattner     return true;
566ee3c74fbSChris Lattner   }
56739a0ffc3SMichael J. Spencer   MemoryBuffer *F = File.take();
568a2f8fc5aSChris Lattner 
569a2f8fc5aSChris Lattner   // If we want to canonicalize whitespace, strip excess whitespace from the
570a2f8fc5aSChris Lattner   // buffer containing the CHECK lines.
571a2f8fc5aSChris Lattner   if (!NoCanonicalizeWhiteSpace)
572a2f8fc5aSChris Lattner     F = CanonicalizeInputFile(F);
573a2f8fc5aSChris Lattner 
574ee3c74fbSChris Lattner   SM.AddNewSourceBuffer(F, SMLoc());
575ee3c74fbSChris Lattner 
57610f10cedSChris Lattner   // Find all instances of CheckPrefix followed by : in the file.
577caa5fc0cSChris Lattner   StringRef Buffer = F->getBuffer();
578ee3c74fbSChris Lattner 
57974d50731SChris Lattner   std::vector<std::pair<SMLoc, Pattern> > NotMatches;
580236d2d5eSChris Lattner 
58192987fb3SAlexander Kornienko   unsigned LineNumber = 1;
58292987fb3SAlexander Kornienko 
583ee3c74fbSChris Lattner   while (1) {
584ee3c74fbSChris Lattner     // See if Prefix occurs in the memory buffer.
58592987fb3SAlexander Kornienko     size_t PrefixLoc = Buffer.find(CheckPrefix);
586ee3c74fbSChris Lattner     // If we didn't find a match, we're done.
58792987fb3SAlexander Kornienko     if (PrefixLoc == StringRef::npos)
588ee3c74fbSChris Lattner       break;
589ee3c74fbSChris Lattner 
59092987fb3SAlexander Kornienko     // Recalculate line number.
59192987fb3SAlexander Kornienko     LineNumber += Buffer.substr(0, PrefixLoc).count('\n');
59292987fb3SAlexander Kornienko 
59392987fb3SAlexander Kornienko     Buffer = Buffer.substr(PrefixLoc);
59492987fb3SAlexander Kornienko 
595caa5fc0cSChris Lattner     const char *CheckPrefixStart = Buffer.data();
596da108b4eSChris Lattner 
597da108b4eSChris Lattner     // When we find a check prefix, keep track of whether we find CHECK: or
598da108b4eSChris Lattner     // CHECK-NEXT:
599236d2d5eSChris Lattner     bool IsCheckNext = false, IsCheckNot = false;
600da108b4eSChris Lattner 
60110f10cedSChris Lattner     // Verify that the : is present after the prefix.
602caa5fc0cSChris Lattner     if (Buffer[CheckPrefix.size()] == ':') {
603caa5fc0cSChris Lattner       Buffer = Buffer.substr(CheckPrefix.size()+1);
604caa5fc0cSChris Lattner     } else if (Buffer.size() > CheckPrefix.size()+6 &&
605caa5fc0cSChris Lattner                memcmp(Buffer.data()+CheckPrefix.size(), "-NEXT:", 6) == 0) {
606ef878a83SBenjamin Kramer       Buffer = Buffer.substr(CheckPrefix.size()+6);
607da108b4eSChris Lattner       IsCheckNext = true;
608236d2d5eSChris Lattner     } else if (Buffer.size() > CheckPrefix.size()+5 &&
609236d2d5eSChris Lattner                memcmp(Buffer.data()+CheckPrefix.size(), "-NOT:", 5) == 0) {
610ef878a83SBenjamin Kramer       Buffer = Buffer.substr(CheckPrefix.size()+5);
611236d2d5eSChris Lattner       IsCheckNot = true;
612da108b4eSChris Lattner     } else {
613caa5fc0cSChris Lattner       Buffer = Buffer.substr(1);
61410f10cedSChris Lattner       continue;
61510f10cedSChris Lattner     }
61610f10cedSChris Lattner 
617ee3c74fbSChris Lattner     // Okay, we found the prefix, yay.  Remember the rest of the line, but
618ee3c74fbSChris Lattner     // ignore leading and trailing whitespace.
619236d2d5eSChris Lattner     Buffer = Buffer.substr(Buffer.find_first_not_of(" \t"));
620ee3c74fbSChris Lattner 
621ee3c74fbSChris Lattner     // Scan ahead to the end of line.
622caa5fc0cSChris Lattner     size_t EOL = Buffer.find_first_of("\n\r");
623ee3c74fbSChris Lattner 
624838fb09aSDan Gohman     // Remember the location of the start of the pattern, for diagnostics.
625838fb09aSDan Gohman     SMLoc PatternLoc = SMLoc::getFromPointer(Buffer.data());
626838fb09aSDan Gohman 
62774d50731SChris Lattner     // Parse the pattern.
62874d50731SChris Lattner     Pattern P;
62992987fb3SAlexander Kornienko     if (P.ParsePattern(Buffer.substr(0, EOL), SM, LineNumber))
630ee3c74fbSChris Lattner       return true;
631ee3c74fbSChris Lattner 
632236d2d5eSChris Lattner     Buffer = Buffer.substr(EOL);
63374d50731SChris Lattner 
634236d2d5eSChris Lattner 
635da108b4eSChris Lattner     // Verify that CHECK-NEXT lines have at least one CHECK line before them.
636da108b4eSChris Lattner     if (IsCheckNext && CheckStrings.empty()) {
637da108b4eSChris Lattner       SM.PrintMessage(SMLoc::getFromPointer(CheckPrefixStart),
63803b80a40SChris Lattner                       SourceMgr::DK_Error,
639da108b4eSChris Lattner                       "found '"+CheckPrefix+"-NEXT:' without previous '"+
64003b80a40SChris Lattner                       CheckPrefix+ ": line");
641da108b4eSChris Lattner       return true;
642da108b4eSChris Lattner     }
643da108b4eSChris Lattner 
64474d50731SChris Lattner     // Handle CHECK-NOT.
64574d50731SChris Lattner     if (IsCheckNot) {
64674d50731SChris Lattner       NotMatches.push_back(std::make_pair(SMLoc::getFromPointer(Buffer.data()),
64774d50731SChris Lattner                                           P));
64874d50731SChris Lattner       continue;
64974d50731SChris Lattner     }
65074d50731SChris Lattner 
6513b40b445SChris Lattner 
652ee3c74fbSChris Lattner     // Okay, add the string we captured to the output vector and move on.
6533b40b445SChris Lattner     CheckStrings.push_back(CheckString(P,
654838fb09aSDan Gohman                                        PatternLoc,
655da108b4eSChris Lattner                                        IsCheckNext));
656236d2d5eSChris Lattner     std::swap(NotMatches, CheckStrings.back().NotStrings);
657ee3c74fbSChris Lattner   }
658ee3c74fbSChris Lattner 
659eba55822SJakob Stoklund Olesen   // Add an EOF pattern for any trailing CHECK-NOTs.
660eba55822SJakob Stoklund Olesen   if (!NotMatches.empty()) {
661eba55822SJakob Stoklund Olesen     CheckStrings.push_back(CheckString(Pattern(true),
662eba55822SJakob Stoklund Olesen                                        SMLoc::getFromPointer(Buffer.data()),
663eba55822SJakob Stoklund Olesen                                        false));
664eba55822SJakob Stoklund Olesen     std::swap(NotMatches, CheckStrings.back().NotStrings);
665eba55822SJakob Stoklund Olesen   }
666eba55822SJakob Stoklund Olesen 
667ee3c74fbSChris Lattner   if (CheckStrings.empty()) {
66810f10cedSChris Lattner     errs() << "error: no check strings found with prefix '" << CheckPrefix
66910f10cedSChris Lattner            << ":'\n";
670ee3c74fbSChris Lattner     return true;
671ee3c74fbSChris Lattner   }
672ee3c74fbSChris Lattner 
673ee3c74fbSChris Lattner   return false;
674ee3c74fbSChris Lattner }
675ee3c74fbSChris Lattner 
676da108b4eSChris Lattner static void PrintCheckFailed(const SourceMgr &SM, const CheckString &CheckStr,
677e0ef65abSDaniel Dunbar                              StringRef Buffer,
678e0ef65abSDaniel Dunbar                              StringMap<StringRef> &VariableTable) {
679da108b4eSChris Lattner   // Otherwise, we have an error, emit an error message.
68003b80a40SChris Lattner   SM.PrintMessage(CheckStr.Loc, SourceMgr::DK_Error,
68103b80a40SChris Lattner                   "expected string not found in input");
682da108b4eSChris Lattner 
683da108b4eSChris Lattner   // Print the "scanning from here" line.  If the current position is at the
684da108b4eSChris Lattner   // end of a line, advance to the start of the next line.
685caa5fc0cSChris Lattner   Buffer = Buffer.substr(Buffer.find_first_not_of(" \t\n\r"));
686da108b4eSChris Lattner 
68703b80a40SChris Lattner   SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()), SourceMgr::DK_Note,
68803b80a40SChris Lattner                   "scanning from here");
689e0ef65abSDaniel Dunbar 
690e0ef65abSDaniel Dunbar   // Allow the pattern to print additional information if desired.
691e0ef65abSDaniel Dunbar   CheckStr.Pat.PrintFailureInfo(SM, Buffer, VariableTable);
692da108b4eSChris Lattner }
693da108b4eSChris Lattner 
69437183584SChris Lattner /// CountNumNewlinesBetween - Count the number of newlines in the specified
69537183584SChris Lattner /// range.
69637183584SChris Lattner static unsigned CountNumNewlinesBetween(StringRef Range) {
697da108b4eSChris Lattner   unsigned NumNewLines = 0;
69837183584SChris Lattner   while (1) {
699da108b4eSChris Lattner     // Scan for newline.
70037183584SChris Lattner     Range = Range.substr(Range.find_first_of("\n\r"));
70137183584SChris Lattner     if (Range.empty()) return NumNewLines;
702da108b4eSChris Lattner 
703da108b4eSChris Lattner     ++NumNewLines;
704da108b4eSChris Lattner 
705da108b4eSChris Lattner     // Handle \n\r and \r\n as a single newline.
70637183584SChris Lattner     if (Range.size() > 1 &&
70737183584SChris Lattner         (Range[1] == '\n' || Range[1] == '\r') &&
70837183584SChris Lattner         (Range[0] != Range[1]))
70937183584SChris Lattner       Range = Range.substr(1);
71037183584SChris Lattner     Range = Range.substr(1);
711da108b4eSChris Lattner   }
712da108b4eSChris Lattner }
713da108b4eSChris Lattner 
714ee3c74fbSChris Lattner int main(int argc, char **argv) {
715ee3c74fbSChris Lattner   sys::PrintStackTraceOnErrorSignal();
716ee3c74fbSChris Lattner   PrettyStackTraceProgram X(argc, argv);
717ee3c74fbSChris Lattner   cl::ParseCommandLineOptions(argc, argv);
718ee3c74fbSChris Lattner 
719ee3c74fbSChris Lattner   SourceMgr SM;
720ee3c74fbSChris Lattner 
721ee3c74fbSChris Lattner   // Read the expected strings from the check file.
72226cccfe1SChris Lattner   std::vector<CheckString> CheckStrings;
723ee3c74fbSChris Lattner   if (ReadCheckFile(SM, CheckStrings))
724ee3c74fbSChris Lattner     return 2;
725ee3c74fbSChris Lattner 
726ee3c74fbSChris Lattner   // Open the file to check and add it to SourceMgr.
72739a0ffc3SMichael J. Spencer   OwningPtr<MemoryBuffer> File;
72839a0ffc3SMichael J. Spencer   if (error_code ec =
72939a0ffc3SMichael J. Spencer         MemoryBuffer::getFileOrSTDIN(InputFilename.c_str(), File)) {
730ee3c74fbSChris Lattner     errs() << "Could not open input file '" << InputFilename << "': "
7317b6fef82SMichael J. Spencer            << ec.message() << '\n';
732*8e1c6477SEli Bendersky     return 2;
733ee3c74fbSChris Lattner   }
73439a0ffc3SMichael J. Spencer   MemoryBuffer *F = File.take();
7352c3e5cdfSChris Lattner 
736b692bed7SChris Lattner   if (F->getBufferSize() == 0) {
737b692bed7SChris Lattner     errs() << "FileCheck error: '" << InputFilename << "' is empty.\n";
738*8e1c6477SEli Bendersky     return 2;
739b692bed7SChris Lattner   }
740b692bed7SChris Lattner 
7412c3e5cdfSChris Lattner   // Remove duplicate spaces in the input file if requested.
7422c3e5cdfSChris Lattner   if (!NoCanonicalizeWhiteSpace)
7432c3e5cdfSChris Lattner     F = CanonicalizeInputFile(F);
7442c3e5cdfSChris Lattner 
745ee3c74fbSChris Lattner   SM.AddNewSourceBuffer(F, SMLoc());
746ee3c74fbSChris Lattner 
7478879e06dSChris Lattner   /// VariableTable - This holds all the current filecheck variables.
7488879e06dSChris Lattner   StringMap<StringRef> VariableTable;
7498879e06dSChris Lattner 
750ee3c74fbSChris Lattner   // Check that we have all of the expected strings, in order, in the input
751ee3c74fbSChris Lattner   // file.
752caa5fc0cSChris Lattner   StringRef Buffer = F->getBuffer();
753ee3c74fbSChris Lattner 
754236d2d5eSChris Lattner   const char *LastMatch = Buffer.data();
755236d2d5eSChris Lattner 
756ee3c74fbSChris Lattner   for (unsigned StrNo = 0, e = CheckStrings.size(); StrNo != e; ++StrNo) {
75726cccfe1SChris Lattner     const CheckString &CheckStr = CheckStrings[StrNo];
758ee3c74fbSChris Lattner 
759caa5fc0cSChris Lattner     StringRef SearchFrom = Buffer;
760caa5fc0cSChris Lattner 
761ee3c74fbSChris Lattner     // Find StrNo in the file.
7623b40b445SChris Lattner     size_t MatchLen = 0;
763eba55822SJakob Stoklund Olesen     size_t MatchPos = CheckStr.Pat.Match(Buffer, MatchLen, VariableTable);
764eba55822SJakob Stoklund Olesen     Buffer = Buffer.substr(MatchPos);
765ee3c74fbSChris Lattner 
766da108b4eSChris Lattner     // If we didn't find a match, reject the input.
767eba55822SJakob Stoklund Olesen     if (MatchPos == StringRef::npos) {
768e0ef65abSDaniel Dunbar       PrintCheckFailed(SM, CheckStr, SearchFrom, VariableTable);
769da108b4eSChris Lattner       return 1;
770ee3c74fbSChris Lattner     }
771ee3c74fbSChris Lattner 
77237183584SChris Lattner     StringRef SkippedRegion(LastMatch, Buffer.data()-LastMatch);
77337183584SChris Lattner 
774da108b4eSChris Lattner     // If this check is a "CHECK-NEXT", verify that the previous match was on
775da108b4eSChris Lattner     // the previous line (i.e. that there is one newline between them).
776da108b4eSChris Lattner     if (CheckStr.IsCheckNext) {
777da108b4eSChris Lattner       // Count the number of newlines between the previous match and this one.
778236d2d5eSChris Lattner       assert(LastMatch != F->getBufferStart() &&
779236d2d5eSChris Lattner              "CHECK-NEXT can't be the first check in a file");
780da108b4eSChris Lattner 
78137183584SChris Lattner       unsigned NumNewLines = CountNumNewlinesBetween(SkippedRegion);
782da108b4eSChris Lattner       if (NumNewLines == 0) {
78303b80a40SChris Lattner         SM.PrintMessage(CheckStr.Loc, SourceMgr::DK_Error,
78403b80a40SChris Lattner                     CheckPrefix+"-NEXT: is on the same line as previous match");
785caa5fc0cSChris Lattner         SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()),
78603b80a40SChris Lattner                         SourceMgr::DK_Note, "'next' match was here");
78703b80a40SChris Lattner         SM.PrintMessage(SMLoc::getFromPointer(LastMatch), SourceMgr::DK_Note,
78803b80a40SChris Lattner                         "previous match was here");
789aedd8185SChris Lattner         return 1;
790ee3c74fbSChris Lattner       }
791ee3c74fbSChris Lattner 
792da108b4eSChris Lattner       if (NumNewLines != 1) {
79303b80a40SChris Lattner         SM.PrintMessage(CheckStr.Loc, SourceMgr::DK_Error, CheckPrefix+
79403b80a40SChris Lattner                         "-NEXT: is not on the line after the previous match");
795caa5fc0cSChris Lattner         SM.PrintMessage(SMLoc::getFromPointer(Buffer.data()),
79603b80a40SChris Lattner                         SourceMgr::DK_Note, "'next' match was here");
79703b80a40SChris Lattner         SM.PrintMessage(SMLoc::getFromPointer(LastMatch), SourceMgr::DK_Note,
79803b80a40SChris Lattner                         "previous match was here");
799da108b4eSChris Lattner         return 1;
800da108b4eSChris Lattner       }
801da108b4eSChris Lattner     }
802da108b4eSChris Lattner 
803236d2d5eSChris Lattner     // If this match had "not strings", verify that they don't exist in the
804236d2d5eSChris Lattner     // skipped region.
8058879e06dSChris Lattner     for (unsigned ChunkNo = 0, e = CheckStr.NotStrings.size();
8068879e06dSChris Lattner          ChunkNo != e; ++ChunkNo) {
80774d50731SChris Lattner       size_t MatchLen = 0;
8088879e06dSChris Lattner       size_t Pos = CheckStr.NotStrings[ChunkNo].second.Match(SkippedRegion,
8098879e06dSChris Lattner                                                              MatchLen,
8108879e06dSChris Lattner                                                              VariableTable);
811236d2d5eSChris Lattner       if (Pos == StringRef::npos) continue;
812236d2d5eSChris Lattner 
81303b80a40SChris Lattner       SM.PrintMessage(SMLoc::getFromPointer(LastMatch+Pos), SourceMgr::DK_Error,
81403b80a40SChris Lattner                       CheckPrefix+"-NOT: string occurred!");
81503b80a40SChris Lattner       SM.PrintMessage(CheckStr.NotStrings[ChunkNo].first, SourceMgr::DK_Note,
81603b80a40SChris Lattner                       CheckPrefix+"-NOT: pattern specified here");
817236d2d5eSChris Lattner       return 1;
818236d2d5eSChris Lattner     }
819236d2d5eSChris Lattner 
820236d2d5eSChris Lattner 
821b9f2bf46SChris Lattner     // Otherwise, everything is good.  Step over the matched text and remember
822b9f2bf46SChris Lattner     // the position after the match as the end of the last match.
8233b40b445SChris Lattner     Buffer = Buffer.substr(MatchLen);
824b9f2bf46SChris Lattner     LastMatch = Buffer.data();
825da108b4eSChris Lattner   }
826da108b4eSChris Lattner 
827ee3c74fbSChris Lattner   return 0;
828ee3c74fbSChris Lattner }
829