add regex support for -verify mode. You can now do things like

expected-error-re {{someregex}}

Patch by mike-m!



git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@102516 91177308-0d34-0410-b5e6-96231b3b80d8

include/clang/Basic/DiagnosticFrontendKinds.td

@@ -72,12 +72,12 @@ def err_fe_pth_file_has_no_source_header : Error<
 def warn_fe_macro_contains_embedded_newline : Warning<
     "macro '%0' contains embedded newline, text after the newline is ignored.">;
 
-def err_verify_bogus_characters : Error<
-    "bogus characters before '{{' in expected string">;
 def err_verify_missing_start : Error<
-    "cannot find start ('{{') of expected string">;
+    "cannot find start ('{{') of expected %0">;
 def err_verify_missing_end : Error<
-    "cannot find end ('}}') of expected string">;
+    "cannot find end ('}}') of expected %0">;
+def err_verify_invalid_content : Error<
+    "invalid expected %0: %1">;
 def err_verify_inconsistent_diags : Error<
     "'%0' diagnostics %select{expected|seen}1 but not %select{seen|expected}1: %2">;
 

include/clang/Frontend/VerifyDiagnosticsClient.h

@@ -25,7 +25,10 @@ class TextDiagnosticBuffer;
 /// USING THE DIAGNOSTIC CHECKER:
 ///
 /// Indicating that a line expects an error or a warning is simple. Put a
-/// comment on the line that has the diagnostic, use "expected-{error,warning}"
+/// comment on the line that has the diagnostic, use:
+///
+///     expected-{error,warning,note}
+///
 /// to tag if it's an expected error or warning, and place the expected text
 /// between {{ and }} markers. The full text doesn't have to be included, only
 /// enough to ensure that the correct diagnostic was emitted.
@@ -45,6 +48,20 @@ class TextDiagnosticBuffer;
 ///
 ///   void f(); // expected-note 2 {{previous declaration is here}}
 ///
+/// Regex matching mode may be selected by appending '-re' to type. Example:
+///
+///   expected-error-re
+///
+/// Examples matching error: "variable has incomplete type 'struct s'"
+///
+///   // expected-error {{variable has incomplete type 'struct s'}}
+///   // expected-error {{variable has incomplete type}}
+///
+///   // expected-error-re {{variable has has type 'struct .'}}
+///   // expected-error-re {{variable has has type 'struct .*'}}
+///   // expected-error-re {{variable has has type 'struct (.*)'}}
+///   // expected-error-re {{variable has has type 'struct[[:space:]](.*)'}}
+///
 class VerifyDiagnosticsClient : public DiagnosticClient {
 public:
   Diagnostic &Diags;

lib/Frontend/VerifyDiagnosticsClient.cpp

@@ -16,6 +16,7 @@
 #include "clang/Frontend/TextDiagnosticBuffer.h"
 #include "clang/Lex/Preprocessor.h"
 #include "llvm/ADT/SmallString.h"
+#include "llvm/Support/Regex.h"
 #include "llvm/Support/raw_ostream.h"
 using namespace clang;
 
@@ -71,97 +72,267 @@ bool VerifyDiagnosticsClient::HadErrors() {
 typedef TextDiagnosticBuffer::DiagList DiagList;
 typedef TextDiagnosticBuffer::const_iterator const_diag_iterator;
 
-/// FindDiagnostics - Go through the comment and see if it indicates expected
-/// diagnostics. If so, then put them in a diagnostic list.
+namespace {
+
+/// Directive - Abstract class representing a parsed verify directive.
 ///
-static void FindDiagnostics(const char *CommentStart, unsigned CommentLen,
-                            DiagList &ExpectedDiags,
-                            Preprocessor &PP, SourceLocation Pos,
-                            const char *ExpectedStr) {
-  const char *CommentEnd = CommentStart+CommentLen;
-  unsigned ExpectedStrLen = strlen(ExpectedStr);
+class Directive {
+public:
+  static Directive* Create(bool RegexKind, const SourceLocation &Location,
+                           const std::string &Text, unsigned Count);
+public:
+  SourceLocation Location;
+  const std::string Text;
+  unsigned Count;
 
-  // Find all expected-foo diagnostics in the string and add them to
-  // ExpectedDiags.
-  while (CommentStart != CommentEnd) {
-    CommentStart = std::find(CommentStart, CommentEnd, 'e');
-    if (unsigned(CommentEnd-CommentStart) < ExpectedStrLen) return;
+  virtual ~Directive() { }
 
-    // If this isn't expected-foo, ignore it.
-    if (memcmp(CommentStart, ExpectedStr, ExpectedStrLen)) {
-      ++CommentStart;
+  // Returns true if directive text is valid.
+  // Otherwise returns false and populates E.
+  virtual bool isValid(std::string &Error) = 0;
+
+  // Returns true on match.
+  virtual bool Match(const std::string &S) = 0;
+
+protected:
+  Directive(const SourceLocation &Location, const std::string &Text,
+            unsigned Count)
+    : Location(Location), Text(Text), Count(Count) { }
+
+private:
+  Directive(const Directive&); // DO NOT IMPLEMENT
+  void operator=(const Directive&); // DO NOT IMPLEMENT
+};
+
+/// StandardDirective - Directive with string matching.
+///
+class StandardDirective : public Directive {
+public:
+  StandardDirective(const SourceLocation &Location, const std::string &Text,
+                    unsigned Count)
+    : Directive(Location, Text, Count) { }
+
+  virtual bool isValid(std::string &Error) {
+    // all strings are considered valid; even empty ones
+    return true;
+  }
+
+  virtual bool Match(const std::string &S) {
+    return S.find(Text) != std::string::npos ||
+           Text.find(S) != std::string::npos;
+  }
+};
+
+/// RegexDirective - Directive with regular-expression matching.
+///
+class RegexDirective : public Directive {
+public:
+  RegexDirective(const SourceLocation &Location, const std::string &Text,
+                 unsigned Count)
+    : Directive(Location, Text, Count), Regex(Text) { }
+
+  virtual bool isValid(std::string &Error) {
+    if (Regex.isValid(Error))
+      return true;
+    return false;
+  }
+
+  virtual bool Match(const std::string &S) {
+    return Regex.match(S);
+  }
+
+private:
+  llvm::Regex Regex;
+};
+
+typedef std::vector<Directive*> DirectiveList;
+
+/// ExpectedData - owns directive objects and deletes on destructor.
+///
+struct ExpectedData {
+  DirectiveList Errors;
+  DirectiveList Warnings;
+  DirectiveList Notes;
+
+  ~ExpectedData() {
+    DirectiveList* Lists[] = { &Errors, &Warnings, &Notes, 0 };
+    for (DirectiveList **PL = Lists; *PL; ++PL) {
+      DirectiveList * const L = *PL;
+      for (DirectiveList::iterator I = L->begin(), E = L->end(); I != E; ++I)
+        delete *I;
+    }
+  }
+};  
+
+class ParseHelper
+{
+public:
+  ParseHelper(const char *Begin, const char *End)
+    : Begin(Begin), End(End), C(Begin), P(Begin), PEnd(NULL) { }
+
+  // Return true if string literal is next.
+  bool Next(const std::string &S) {
+    std::string::size_type LEN = S.length();
+    P = C;
+    PEnd = C + LEN;
+    if (PEnd > End)
+      return false;
+    return !memcmp(P, S.c_str(), LEN);
+  }
+
+  // Return true if number is next.
+  // Output N only if number is next.
+  bool Next(unsigned &N) {
+    unsigned TMP = 0;
+    P = C;
+    for (; P < End && P[0] >= '0' && P[0] <= '9'; ++P) {
+      TMP *= 10;
+      TMP += P[0] - '0';
+    }
+    if (P == C)
+      return false;
+    PEnd = P;
+    N = TMP;
+    return true;
+  }
+
+  // Return true if string literal is found.
+  // When true, P marks begin-position of S in content.
+  bool Search(const std::string &S) {
+    P = std::search(C, End, S.begin(), S.end());
+    PEnd = P + S.length();
+    return P != End;
+  }
+
+  // Advance 1-past previous next/search.
+  // Behavior is undefined if previous next/search failed.
+  bool Advance() {
+    C = PEnd;
+    return C < End;
+  }
+
+  // Skip zero or more whitespace.
+  void SkipWhitespace() {
+    for (; C < End && isspace(*C); ++C)
+      ;
+  }
+
+  // Return true if EOF reached.
+  bool Done() {
+    return !(C < End);
+  }
+
+  const char * const Begin; // beginning of expected content
+  const char * const End;   // end of expected content (1-past)
+  const char *C;            // position of next char in content
+  const char *P;
+
+private:
+  const char *PEnd; // previous next/search subject end (1-past)
+};
+
+} // namespace anonymous
+
+/// ParseDirective - Go through the comment and see if it indicates expected
+/// diagnostics. If so, then put them in the appropriate directive list.
+///
+static void ParseDirective(const char *CommentStart, unsigned CommentLen,
+                           ExpectedData &ED, Preprocessor &PP,
+                           SourceLocation Pos) {
+  // A single comment may contain multiple directives.
+  for (ParseHelper PH(CommentStart, CommentStart+CommentLen); !PH.Done();) {
+    // search for token: expected
+    if (!PH.Search("expected"))
+      break;
+    PH.Advance();
+
+    // next token: -
+    if (!PH.Next("-"))
+      continue;
+    PH.Advance();
+
+    // next token: { error | warning | note }
+    DirectiveList* DL = NULL;
+    if (PH.Next("error"))
+      DL = &ED.Errors;
+    else if (PH.Next("warning"))
+      DL = &ED.Warnings;
+    else if (PH.Next("note"))
+      DL = &ED.Notes;
+    else
+      continue;
+    PH.Advance();
+
+    // default directive kind
+    bool RegexKind = false;
+    const char* KindStr = "string";
+
+    // next optional token: -
+    if (PH.Next("-re")) {
+      PH.Advance();
+      RegexKind = true;
+      KindStr = "regex";
+    }
+
+    // skip optional whitespace
+    PH.SkipWhitespace();
+
+    // next optional token: positive integer
+    unsigned Count = 1;
+    if (PH.Next(Count))
+      PH.Advance();
+
+    // skip optional whitespace
+    PH.SkipWhitespace();
+
+    // next token: {{
+    if (!PH.Next("{{")) {
+      PP.Diag(Pos.getFileLocWithOffset(PH.C-PH.Begin),
+              diag::err_verify_missing_start) << KindStr;
       continue;
     }
+    PH.Advance();
+    const char* const ContentBegin = PH.C; // mark content begin
 
-    CommentStart += ExpectedStrLen;
-
-    // Skip whitespace.
-    while (CommentStart != CommentEnd &&
-           isspace(CommentStart[0]))
-      ++CommentStart;
-
-    // Default, if we find the '{' now, is 1 time.
-    int Times = 1;
-    int Temp = 0;
-    // In extended syntax, there could be a digit now.
-    while (CommentStart != CommentEnd &&
-           CommentStart[0] >= '0' && CommentStart[0] <= '9') {
-      Temp *= 10;
-      Temp += CommentStart[0] - '0';
-      ++CommentStart;
+    // search for token: }}
+    if (!PH.Search("}}")) {
+      PP.Diag(Pos.getFileLocWithOffset(PH.C-PH.Begin),
+              diag::err_verify_missing_end) << KindStr;
+      continue;
     }
-    if (Temp > 0)
-      Times = Temp;
+    const char* const ContentEnd = PH.P; // mark content end
+    PH.Advance();
 
-    // Skip whitespace again.
-    while (CommentStart != CommentEnd &&
-           isspace(CommentStart[0]))
-      ++CommentStart;
-
-    // We should have a {{ now.
-    if (CommentEnd-CommentStart < 2 ||
-        CommentStart[0] != '{' || CommentStart[1] != '{') {
-      if (std::find(CommentStart, CommentEnd, '{') != CommentEnd)
-        PP.Diag(Pos, diag::err_verify_bogus_characters);
-      else
-        PP.Diag(Pos, diag::err_verify_missing_start);
-      return;
+    // build directive text; convert \n to newlines
+    std::string Text;
+    llvm::StringRef NewlineStr = "\\n";
+    llvm::StringRef Content(ContentBegin, ContentEnd-ContentBegin);
+    size_t CPos = 0;
+    size_t FPos;
+    while ((FPos = Content.find(NewlineStr, CPos)) != llvm::StringRef::npos) {
+      Text += Content.substr(CPos, FPos-CPos);
+      Text += '\n';
+      CPos = FPos + NewlineStr.size();
     }
-    CommentStart += 2;
+    if (Text.empty())
+      Text.assign(ContentBegin, ContentEnd);
 
-    // Find the }}.
-    const char *ExpectedEnd = CommentStart;
-    while (1) {
-      ExpectedEnd = std::find(ExpectedEnd, CommentEnd, '}');
-      if (CommentEnd-ExpectedEnd < 2) {
-        PP.Diag(Pos, diag::err_verify_missing_end);
-        return;
+    // construct new directive
+    Directive *D = Directive::Create(RegexKind, Pos, Text, Count);
+    std::string Error;
+    if (D->isValid(Error))
+      DL->push_back(D);
+    else {
+      PP.Diag(Pos.getFileLocWithOffset(ContentBegin-PH.Begin),
+              diag::err_verify_invalid_content)
+        << KindStr << Error;
     }
-
-      if (ExpectedEnd[1] == '}')
-        break;
-
-      ++ExpectedEnd;  // Skip over singular }'s
-    }
-
-    std::string Msg(CommentStart, ExpectedEnd);
-    std::string::size_type FindPos;
-    while ((FindPos = Msg.find("\\n")) != std::string::npos)
-      Msg.replace(FindPos, 2, "\n");
-    // Add is possibly multiple times.
-    for (int i = 0; i < Times; ++i)
-      ExpectedDiags.push_back(std::make_pair(Pos, Msg));
-
-    CommentStart = ExpectedEnd;
   }
 }
 
 /// FindExpectedDiags - Lex the main source file to find all of the
 //   expected errors and warnings.
-static void FindExpectedDiags(Preprocessor &PP,
-                              DiagList &ExpectedErrors,
-                              DiagList &ExpectedWarnings,
-                              DiagList &ExpectedNotes) {
+static void FindExpectedDiags(Preprocessor &PP, ExpectedData &ED) {
   // Create a raw lexer to pull all the comments out of the main file.  We don't
   // want to look in #include'd headers for expected-error strings.
   SourceManager &SM = PP.getSourceManager();
@@ -185,17 +356,8 @@ static void FindExpectedDiags(Preprocessor &PP,
     std::string Comment = PP.getSpelling(Tok);
     if (Comment.empty()) continue;
 
-    // Find all expected errors.
-    FindDiagnostics(&Comment[0], Comment.size(), ExpectedErrors, PP,
-                    Tok.getLocation(), "expected-error");
-
-    // Find all expected warnings.
-    FindDiagnostics(&Comment[0], Comment.size(), ExpectedWarnings, PP,
-                    Tok.getLocation(), "expected-warning");
-
-    // Find all expected notes.
-    FindDiagnostics(&Comment[0], Comment.size(), ExpectedNotes, PP,
-                    Tok.getLocation(), "expected-note");
+    // Find all expected errors/warnings/notes.
+    ParseDirective(&Comment[0], Comment.size(), ED, PP, Tok.getLocation());
   };
 }
 
@@ -225,35 +387,54 @@ static unsigned PrintProblem(Diagnostic &Diags, SourceManager *SourceMgr,
   return std::distance(diag_begin, diag_end);
 }
 
-/// CompareDiagLists - Compare two diagnostic lists and return the difference
-/// between them.
+static unsigned PrintProblem(Diagnostic &Diags, SourceManager *SourceMgr,
+                             DirectiveList &DL, const char *Kind,
+                             bool Expected) {
+  if (DL.empty())
+    return 0;
+
+  llvm::SmallString<256> Fmt;
+  llvm::raw_svector_ostream OS(Fmt);
+  for (DirectiveList::iterator I = DL.begin(), E = DL.end(); I != E; ++I) {
+    Directive& D = **I;
+    if (D.Location.isInvalid() || !SourceMgr)
+      OS << "\n  (frontend)";
+    else
+      OS << "\n  Line " << SourceMgr->getInstantiationLineNumber(D.Location);
+    OS << ": " << D.Text;
+  }
+
+  Diags.Report(diag::err_verify_inconsistent_diags)
+    << Kind << !Expected << OS.str();
+  return DL.size();
+}
+
+/// CheckLists - Compare expected to seen diagnostic lists and return the
+/// the difference between them.
 ///
-static unsigned CompareDiagLists(Diagnostic &Diags,
-                                 SourceManager &SourceMgr,
-                                 const_diag_iterator d1_begin,
-                                 const_diag_iterator d1_end,
+static unsigned CheckLists(Diagnostic &Diags, SourceManager &SourceMgr,
+                           const char *Label,
+                           DirectiveList &Left,
                            const_diag_iterator d2_begin,
-                                 const_diag_iterator d2_end,
-                                 const char *Label) {
-  DiagList LeftOnly;
-  DiagList Left(d1_begin, d1_end);
+                           const_diag_iterator d2_end) {
+  DirectiveList LeftOnly;
   DiagList Right(d2_begin, d2_end);
 
-  for (const_diag_iterator I = Left.begin(), E = Left.end(); I != E; ++I) {
-    unsigned LineNo1 = SourceMgr.getInstantiationLineNumber(I->first);
-    const std::string &Diag1 = I->second;
+  for (DirectiveList::iterator I = Left.begin(), E = Left.end(); I != E; ++I) {
+    Directive& D = **I;
+    unsigned LineNo1 = SourceMgr.getInstantiationLineNumber(D.Location);
 
+    for (unsigned i = 0; i < D.Count; ++i) {
       DiagList::iterator II, IE;
       for (II = Right.begin(), IE = Right.end(); II != IE; ++II) {
         unsigned LineNo2 = SourceMgr.getInstantiationLineNumber(II->first);
-      if (LineNo1 != LineNo2) continue;
+        if (LineNo1 != LineNo2)
+          continue;
 
-      const std::string &Diag2 = II->second;
-      if (Diag2.find(Diag1) != std::string::npos ||
-          Diag1.find(Diag2) != std::string::npos) {
+        const std::string &RightText = II->second;
+        if (D.Match(RightText))
           break;
       }
-    }
       if (II == IE) {
         // Not found.
         LeftOnly.push_back(*I);
@@ -262,12 +443,12 @@ static unsigned CompareDiagLists(Diagnostic &Diags,
         Right.erase(II);
       }
     }
+  }
   // Now all that's left in Right are those that were not matched.
 
-  return (PrintProblem(Diags, &SourceMgr,
-                      LeftOnly.begin(), LeftOnly.end(), Label, true) +
-          PrintProblem(Diags, &SourceMgr,
-                       Right.begin(), Right.end(), Label, false));
+  return (PrintProblem(Diags, &SourceMgr, LeftOnly, Label, true) +
+          PrintProblem(Diags, &SourceMgr, Right.begin(), Right.end(),
+                       Label, false));
 }
 
 /// CheckResults - This compares the expected results to those that
@@ -276,9 +457,7 @@ static unsigned CompareDiagLists(Diagnostic &Diags,
 ///
 static unsigned CheckResults(Diagnostic &Diags, SourceManager &SourceMgr,
                              const TextDiagnosticBuffer &Buffer,
-                             const DiagList &ExpectedErrors,
-                             const DiagList &ExpectedWarnings,
-                             const DiagList &ExpectedNotes) {
+                             ExpectedData &ED) {
   // We want to capture the delta between what was expected and what was
   // seen.
   //
@@ -287,31 +466,22 @@ static unsigned CheckResults(Diagnostic &Diags, SourceManager &SourceMgr,
   unsigned NumProblems = 0;
 
   // See if there are error mismatches.
-  NumProblems += CompareDiagLists(Diags, SourceMgr,
-                                  ExpectedErrors.begin(), ExpectedErrors.end(),
-                                  Buffer.err_begin(), Buffer.err_end(),
-                                  "error");
+  NumProblems += CheckLists(Diags, SourceMgr, "error", ED.Errors,
+                            Buffer.err_begin(), Buffer.err_end());
 
   // See if there are warning mismatches.
-  NumProblems += CompareDiagLists(Diags, SourceMgr,
-                                  ExpectedWarnings.begin(),
-                                  ExpectedWarnings.end(),
-                                  Buffer.warn_begin(), Buffer.warn_end(),
-                                  "warning");
+  NumProblems += CheckLists(Diags, SourceMgr, "warning", ED.Warnings,
+                            Buffer.warn_begin(), Buffer.warn_end());
 
   // See if there are note mismatches.
-  NumProblems += CompareDiagLists(Diags, SourceMgr,
-                                  ExpectedNotes.begin(),
-                                  ExpectedNotes.end(),
-                                  Buffer.note_begin(), Buffer.note_end(),
-                                  "note");
+  NumProblems += CheckLists(Diags, SourceMgr, "note", ED.Notes,
+                            Buffer.note_begin(), Buffer.note_end());
 
   return NumProblems;
 }
 
-
 void VerifyDiagnosticsClient::CheckDiagnostics() {
-  DiagList ExpectedErrors, ExpectedWarnings, ExpectedNotes;
+  ExpectedData ED;
 
   // Ensure any diagnostics go to the primary client.
   DiagnosticClient *CurClient = Diags.getClient();
@@ -320,13 +490,11 @@ void VerifyDiagnosticsClient::CheckDiagnostics() {
   // If we have a preprocessor, scan the source for expected diagnostic
   // markers. If not then any diagnostics are unexpected.
   if (CurrentPreprocessor) {
-    FindExpectedDiags(*CurrentPreprocessor, ExpectedErrors, ExpectedWarnings,
-                      ExpectedNotes);
+    FindExpectedDiags(*CurrentPreprocessor, ED);
 
     // Check that the expected diagnostics occurred.
     NumErrors += CheckResults(Diags, CurrentPreprocessor->getSourceManager(),
-                              *Buffer,
-                              ExpectedErrors, ExpectedWarnings, ExpectedNotes);
+                              *Buffer, ED);
   } else {
     NumErrors += (PrintProblem(Diags, 0,
                                Buffer->err_begin(), Buffer->err_end(),
@@ -344,3 +512,10 @@ void VerifyDiagnosticsClient::CheckDiagnostics() {
   // Reset the buffer, we have processed all the diagnostics in it.
   Buffer.reset(new TextDiagnosticBuffer());
 }
+
+Directive* Directive::Create(bool RegexKind, const SourceLocation &Location,
+                             const std::string &Text, unsigned Count) {
+  if (RegexKind)
+    return new RegexDirective(Location, Text, Count);
+  return new StandardDirective(Location, Text, Count);
+}

test/Misc/verify.c

@@ -0,0 +1,14 @@
+// RUN: %clang_cc1 -fsyntax-only -verify %s
+
+struct s; // expected-note 7 {{forward declaration of 'struct s'}}
+
+// standard string matching
+struct s s1; // expected-error {{tentative definition has type 'struct s' that is never completed}}
+struct s s2; // expected-error {{tentative definition has type}}
+
+// regex matching
+struct s r1; // expected-error-re {{tentative definition has type 'struct s' that is never completed}}
+struct s r2; // expected-error-re {{tentative definition has type '.*[[:space:]]*.*' that is never completed}}
+struct s r3; // expected-error-re {{tentative definition has type '(.*)[[:space:]]*(.*)' that is never completed}}
+struct s r4; // expected-error-re {{^tentative}}
+struct s r5; // expected-error-re {{completed$}}