1 //==-- llvm/Support/FileCheck.h ---------------------------*- C++ -*-==//
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 /// \file This file has some utilities to use FileCheck as an API
11 //
12 //===----------------------------------------------------------------------===//
13 
14 #ifndef LLVM_SUPPORT_FILECHECK_H
15 #define LLVM_SUPPORT_FILECHECK_H
16 
17 #include "llvm/ADT/StringMap.h"
18 #include "llvm/Support/MemoryBuffer.h"
19 #include "llvm/Support/Regex.h"
20 #include "llvm/Support/SourceMgr.h"
21 #include <vector>
22 #include <map>
23 
24 namespace llvm {
25 
26 /// Contains info about various FileCheck options.
27 struct FileCheckRequest {
28   std::vector<std::string> CheckPrefixes;
29   bool NoCanonicalizeWhiteSpace = false;
30   std::vector<std::string> ImplicitCheckNot;
31   std::vector<std::string> GlobalDefines;
32   bool AllowEmptyInput = false;
33   bool MatchFullLines = false;
34   bool EnableVarScope = false;
35   bool AllowDeprecatedDagOverlap = false;
36   bool Verbose = false;
37   bool VerboseVerbose = false;
38 };
39 
40 
41 //===----------------------------------------------------------------------===//
42 // Pattern Handling Code.
43 //===----------------------------------------------------------------------===//
44 
45 namespace Check {
46 
47 enum FileCheckKind {
48   CheckNone = 0,
49   CheckPlain,
50   CheckNext,
51   CheckSame,
52   CheckNot,
53   CheckDAG,
54   CheckLabel,
55   CheckEmpty,
56 
57   /// Indicates the pattern only matches the end of file. This is used for
58   /// trailing CHECK-NOTs.
59   CheckEOF,
60 
61   /// Marks when parsing found a -NOT check combined with another CHECK suffix.
62   CheckBadNot,
63 
64   /// Marks when parsing found a -COUNT directive with invalid count value.
65   CheckBadCount
66 };
67 
68 class FileCheckType {
69   FileCheckKind Kind;
70   int Count; ///< optional Count for some checks
71 
72 public:
Kind(Kind)73   FileCheckType(FileCheckKind Kind = CheckNone) : Kind(Kind), Count(1) {}
74   FileCheckType(const FileCheckType &) = default;
75 
FileCheckKind()76   operator FileCheckKind() const { return Kind; }
77 
getCount()78   int getCount() const { return Count; }
79   FileCheckType &setCount(int C);
80 
81   std::string getDescription(StringRef Prefix) const;
82 };
83 }
84 
85 struct FileCheckDiag;
86 
87 class FileCheckPattern {
88   SMLoc PatternLoc;
89 
90   /// A fixed string to match as the pattern or empty if this pattern requires
91   /// a regex match.
92   StringRef FixedStr;
93 
94   /// A regex string to match as the pattern or empty if this pattern requires
95   /// a fixed string to match.
96   std::string RegExStr;
97 
98   /// Entries in this vector map to uses of a variable in the pattern, e.g.
99   /// "foo[[bar]]baz".  In this case, the RegExStr will contain "foobaz" and
100   /// we'll get an entry in this vector that tells us to insert the value of
101   /// bar at offset 3.
102   std::vector<std::pair<StringRef, unsigned>> VariableUses;
103 
104   /// Maps definitions of variables to their parenthesized capture numbers.
105   ///
106   /// E.g. for the pattern "foo[[bar:.*]]baz", VariableDefs will map "bar" to
107   /// 1.
108   std::map<StringRef, unsigned> VariableDefs;
109 
110   Check::FileCheckType CheckTy;
111 
112   /// Contains the number of line this pattern is in.
113   unsigned LineNumber;
114 
115 public:
FileCheckPattern(Check::FileCheckType Ty)116   explicit FileCheckPattern(Check::FileCheckType Ty)
117       : CheckTy(Ty) {}
118 
119   /// Returns the location in source code.
getLoc()120   SMLoc getLoc() const { return PatternLoc; }
121 
122   bool ParsePattern(StringRef PatternStr, StringRef Prefix, SourceMgr &SM,
123                     unsigned LineNumber, const FileCheckRequest &Req);
124   size_t Match(StringRef Buffer, size_t &MatchLen,
125                StringMap<StringRef> &VariableTable) const;
126   void PrintVariableUses(const SourceMgr &SM, StringRef Buffer,
127                          const StringMap<StringRef> &VariableTable,
128                          SMRange MatchRange = None) const;
129   void PrintFuzzyMatch(const SourceMgr &SM, StringRef Buffer,
130                        const StringMap<StringRef> &VariableTable,
131                        std::vector<FileCheckDiag> *Diags) const;
132 
hasVariable()133   bool hasVariable() const {
134     return !(VariableUses.empty() && VariableDefs.empty());
135   }
136 
getCheckTy()137   Check::FileCheckType getCheckTy() const { return CheckTy; }
138 
getCount()139   int getCount() const { return CheckTy.getCount(); }
140 
141 private:
142   bool AddRegExToRegEx(StringRef RS, unsigned &CurParen, SourceMgr &SM);
143   void AddBackrefToRegEx(unsigned BackrefNum);
144   unsigned
145   ComputeMatchDistance(StringRef Buffer,
146                        const StringMap<StringRef> &VariableTable) const;
147   bool EvaluateExpression(StringRef Expr, std::string &Value) const;
148   size_t FindRegexVarEnd(StringRef Str, SourceMgr &SM);
149 };
150 
151 //===----------------------------------------------------------------------===//
152 /// Summary of a FileCheck diagnostic.
153 //===----------------------------------------------------------------------===//
154 
155 struct FileCheckDiag {
156   /// What is the FileCheck directive for this diagnostic?
157   Check::FileCheckType CheckTy;
158   /// Where is the FileCheck directive for this diagnostic?
159   unsigned CheckLine, CheckCol;
160   /// What type of match result does this diagnostic describe?
161   ///
162   /// A directive's supplied pattern is said to be either expected or excluded
163   /// depending on whether the pattern must have or must not have a match in
164   /// order for the directive to succeed.  For example, a CHECK directive's
165   /// pattern is expected, and a CHECK-NOT directive's pattern is excluded.
166   /// All match result types whose names end with "Excluded" are for excluded
167   /// patterns, and all others are for expected patterns.
168   ///
169   /// There might be more than one match result for a single pattern.  For
170   /// example, there might be several discarded matches
171   /// (MatchFoundButDiscarded) before either a good match
172   /// (MatchFoundAndExpected) or a failure to match (MatchNoneButExpected),
173   /// and there might be a fuzzy match (MatchFuzzy) after the latter.
174   enum MatchType {
175     /// Indicates a good match for an expected pattern.
176     MatchFoundAndExpected,
177     /// Indicates a match for an excluded pattern.
178     MatchFoundButExcluded,
179     /// Indicates a match for an expected pattern, but the match is on the
180     /// wrong line.
181     MatchFoundButWrongLine,
182     /// Indicates a discarded match for an expected pattern.
183     MatchFoundButDiscarded,
184     /// Indicates no match for an excluded pattern.
185     MatchNoneAndExcluded,
186     /// Indicates no match for an expected pattern, but this might follow good
187     /// matches when multiple matches are expected for the pattern, or it might
188     /// follow discarded matches for the pattern.
189     MatchNoneButExpected,
190     /// Indicates a fuzzy match that serves as a suggestion for the next
191     /// intended match for an expected pattern with too few or no good matches.
192     MatchFuzzy,
193   } MatchTy;
194   /// The search range if MatchTy is MatchNoneAndExcluded or
195   /// MatchNoneButExpected, or the match range otherwise.
196   unsigned InputStartLine;
197   unsigned InputStartCol;
198   unsigned InputEndLine;
199   unsigned InputEndCol;
200   FileCheckDiag(const SourceMgr &SM, const Check::FileCheckType &CheckTy,
201                 SMLoc CheckLoc, MatchType MatchTy, SMRange InputRange);
202 };
203 
204 //===----------------------------------------------------------------------===//
205 // Check Strings.
206 //===----------------------------------------------------------------------===//
207 
208 /// A check that we found in the input file.
209 struct FileCheckString {
210   /// The pattern to match.
211   FileCheckPattern Pat;
212 
213   /// Which prefix name this check matched.
214   StringRef Prefix;
215 
216   /// The location in the match file that the check string was specified.
217   SMLoc Loc;
218 
219   /// All of the strings that are disallowed from occurring between this match
220   /// string and the previous one (or start of file).
221   std::vector<FileCheckPattern> DagNotStrings;
222 
FileCheckStringFileCheckString223   FileCheckString(const FileCheckPattern &P, StringRef S, SMLoc L)
224       : Pat(P), Prefix(S), Loc(L) {}
225 
226   size_t Check(const SourceMgr &SM, StringRef Buffer, bool IsLabelScanMode,
227                size_t &MatchLen, StringMap<StringRef> &VariableTable,
228                FileCheckRequest &Req, std::vector<FileCheckDiag> *Diags) const;
229 
230   bool CheckNext(const SourceMgr &SM, StringRef Buffer) const;
231   bool CheckSame(const SourceMgr &SM, StringRef Buffer) const;
232   bool CheckNot(const SourceMgr &SM, StringRef Buffer,
233                 const std::vector<const FileCheckPattern *> &NotStrings,
234                 StringMap<StringRef> &VariableTable,
235                 const FileCheckRequest &Req,
236                 std::vector<FileCheckDiag> *Diags) const;
237   size_t CheckDag(const SourceMgr &SM, StringRef Buffer,
238                   std::vector<const FileCheckPattern *> &NotStrings,
239                   StringMap<StringRef> &VariableTable,
240                   const FileCheckRequest &Req,
241                   std::vector<FileCheckDiag> *Diags) const;
242 };
243 
244 /// FileCheck class takes the request and exposes various methods that
245 /// use information from the request.
246 class FileCheck {
247   FileCheckRequest Req;
248 
249 public:
FileCheck(FileCheckRequest Req)250   FileCheck(FileCheckRequest Req) : Req(Req) {}
251 
252   // Combines the check prefixes into a single regex so that we can efficiently
253   // scan for any of the set.
254   //
255   // The semantics are that the longest-match wins which matches our regex
256   // library.
257   Regex buildCheckPrefixRegex();
258 
259   /// Read the check file, which specifies the sequence of expected strings.
260   ///
261   /// The strings are added to the CheckStrings vector. Returns true in case of
262   /// an error, false otherwise.
263   bool ReadCheckFile(SourceMgr &SM, StringRef Buffer, Regex &PrefixRE,
264                      std::vector<FileCheckString> &CheckStrings);
265 
266   bool ValidateCheckPrefixes();
267 
268   /// Canonicalize whitespaces in the file. Line endings are replaced with
269   /// UNIX-style '\n'.
270   StringRef CanonicalizeFile(MemoryBuffer &MB,
271                              SmallVectorImpl<char> &OutputBuffer);
272 
273   /// Check the input to FileCheck provided in the \p Buffer against the \p
274   /// CheckStrings read from the check file.
275   ///
276   /// Returns false if the input fails to satisfy the checks.
277   bool CheckInput(SourceMgr &SM, StringRef Buffer,
278                   ArrayRef<FileCheckString> CheckStrings,
279                   std::vector<FileCheckDiag> *Diags = nullptr);
280 };
281 } // namespace llvm
282 #endif
283