[NFC][SpecialCaseList] Split Matcher into RegexMatcher and GlobMatcher

clang/lib/Basic/Diagnostic.cpp

@@ -517,12 +517,6 @@ class WarningsSpecialCaseList : public llvm::SpecialCaseList {
                         const SourceManager &SM) const;
 
 private:
-  // Find the longest glob pattern that matches FilePath amongst
-  // CategoriesToMatchers, return true iff the match exists and belongs to a
-  // positive category.
-  bool globsMatches(const llvm::StringMap<Matcher> &CategoriesToMatchers,
-                    StringRef FilePath) const;
-
   llvm::DenseMap<diag::kind, const Section *> DiagToSection;
 };
 } // namespace
@@ -538,7 +532,7 @@ WarningsSpecialCaseList::create(const llvm::MemoryBuffer &Input,
 
 void WarningsSpecialCaseList::processSections(DiagnosticsEngine &Diags) {
   static constexpr auto WarningFlavor = clang::diag::Flavor::WarningOrError;
-  for (const auto &SectionEntry : Sections) {
+  for (const auto &SectionEntry : sections()) {
     StringRef DiagGroup = SectionEntry.SectionStr;
     if (DiagGroup == "*") {
       // Drop the default section introduced by special case list, we only
@@ -584,43 +578,24 @@ void DiagnosticsEngine::setDiagSuppressionMapping(llvm::MemoryBuffer &Input) {
 bool WarningsSpecialCaseList::isDiagSuppressed(diag::kind DiagId,
                                                SourceLocation DiagLoc,
                                                const SourceManager &SM) const {
+  PresumedLoc PLoc = SM.getPresumedLoc(DiagLoc);
+  if (!PLoc.isValid())
+    return false;
   const Section *DiagSection = DiagToSection.lookup(DiagId);
   if (!DiagSection)
     return false;
-  const SectionEntries &EntityTypeToCategories = DiagSection->Entries;
-  auto SrcEntriesIt = EntityTypeToCategories.find("src");
-  if (SrcEntriesIt == EntityTypeToCategories.end())
+
+  StringRef F = llvm::sys::path::remove_leading_dotslash(PLoc.getFilename());
+
+  StringRef LongestSup = DiagSection->getLongestMatch("src", F, "");
+  if (LongestSup.empty())
     return false;
-  const llvm::StringMap<llvm::SpecialCaseList::Matcher> &CategoriesToMatchers =
-      SrcEntriesIt->getValue();
-  // We also use presumed locations here to improve reproducibility for
-  // preprocessed inputs.
-  if (PresumedLoc PLoc = SM.getPresumedLoc(DiagLoc); PLoc.isValid())
-    return globsMatches(
-        CategoriesToMatchers,
-        llvm::sys::path::remove_leading_dotslash(PLoc.getFilename()));
-  return false;
-}
 
-bool WarningsSpecialCaseList::globsMatches(
-    const llvm::StringMap<Matcher> &CategoriesToMatchers,
-    StringRef FilePath) const {
-  StringRef LongestMatch;
-  bool LongestIsPositive = false;
-  for (const auto &Entry : CategoriesToMatchers) {
-    StringRef Category = Entry.getKey();
-    const llvm::SpecialCaseList::Matcher &Matcher = Entry.getValue();
-    bool IsPositive = Category != "emit";
-    for (const auto &Glob : Matcher.Globs) {
-      if (Glob->Name.size() < LongestMatch.size())
-        continue;
-      if (!Glob->Pattern.match(FilePath))
-        continue;
-      LongestMatch = Glob->Name;
-      LongestIsPositive = IsPositive;
-    }
-  }
-  return LongestIsPositive;
+  StringRef LongestEmit = DiagSection->getLongestMatch("src", F, "emit");
+  if (LongestEmit.empty())
+    return true;
+
+  return LongestSup.size() > LongestEmit.size();
 }
 
 bool DiagnosticsEngine::isSuppressedViaMapping(diag::kind DiagId,

clang/lib/Basic/ProfileList.cpp

@@ -32,10 +32,10 @@ class ProfileSpecialCaseList : public llvm::SpecialCaseList {
   createOrDie(const std::vector<std::string> &Paths,
               llvm::vfs::FileSystem &VFS);
 
-  bool isEmpty() const { return Sections.empty(); }
+  bool isEmpty() const { return sections().empty(); }
 
   bool hasPrefix(StringRef Prefix) const {
-    for (const auto &It : Sections)
+    for (const auto &It : sections())
       if (It.Entries.count(Prefix) > 0)
         return true;
     return false;

clang/lib/Basic/SanitizerSpecialCaseList.cpp

@@ -38,7 +38,7 @@ SanitizerSpecialCaseList::createOrDie(const std::vector<std::string> &Paths,
 }
 
 void SanitizerSpecialCaseList::createSanitizerSections() {
-  for (const auto &S : Sections) {
+  for (const auto &S : sections()) {
     SanitizerMask Mask;
 
 #define SANITIZER(NAME, ID)                                                    \

llvm/include/llvm/Support/SpecialCaseList.h

@@ -12,13 +12,15 @@
 #ifndef LLVM_SUPPORT_SPECIALCASELIST_H
 #define LLVM_SUPPORT_SPECIALCASELIST_H
 
+#include "llvm/ADT/ArrayRef.h"
 #include "llvm/ADT/StringMap.h"
 #include "llvm/Support/Compiler.h"
 #include "llvm/Support/GlobPattern.h"
 #include "llvm/Support/Regex.h"
 #include <memory>
 #include <string>
 #include <utility>
+#include <variant>
 #include <vector>
 
 namespace llvm {
@@ -118,20 +120,34 @@ class SpecialCaseList {
   SpecialCaseList(SpecialCaseList const &) = delete;
   SpecialCaseList &operator=(SpecialCaseList const &) = delete;
 
-  /// Represents a set of globs and their line numbers
-  class Matcher {
+private:
+  // Lagacy v1 matcher.
+  class RegexMatcher {
   public:
-    LLVM_ABI Error insert(StringRef Pattern, unsigned LineNumber,
-                          bool UseRegex);
+    LLVM_ABI Error insert(StringRef Pattern, unsigned LineNumber);
     LLVM_ABI void
     match(StringRef Query,
           llvm::function_ref<void(StringRef Rule, unsigned LineNo)> Cb) const;
 
-    LLVM_ABI bool matchAny(StringRef Query) const {
-      bool R = false;
-      match(Query, [&](StringRef, unsigned) { R = true; });
-      return R;
-    }
+    struct Reg {
+      Reg(StringRef Name, unsigned LineNo, Regex &&Rg)
+          : Name(Name), LineNo(LineNo), Rg(std::move(Rg)) {}
+      std::string Name;
+      unsigned LineNo;
+      Regex Rg;
+      Reg(Reg &&) = delete;
+      Reg() = default;
+    };
+
+    std::vector<std::unique_ptr<Reg>> RegExes;
+  };
+
+  class GlobMatcher {
+  public:
+    LLVM_ABI Error insert(StringRef Pattern, unsigned LineNumber);
+    LLVM_ABI void
+    match(StringRef Query,
+          llvm::function_ref<void(StringRef Rule, unsigned LineNo)> Cb) const;
 
     struct Glob {
       Glob(StringRef Name, unsigned LineNo) : Name(Name), LineNo(LineNo) {}
@@ -144,25 +160,35 @@ class SpecialCaseList {
       Glob() = default;
     };
 
-    struct Reg {
-      Reg(StringRef Name, unsigned LineNo, Regex &&Rg)
-          : Name(Name), LineNo(LineNo), Rg(std::move(Rg)) {}
-      std::string Name;
-      unsigned LineNo;
-      Regex Rg;
-      Reg(Reg &&) = delete;
-      Reg() = default;
-    };
+    std::vector<std::unique_ptr<Glob>> Globs;
+  };
 
-    std::vector<std::unique_ptr<Matcher::Glob>> Globs;
-    std::vector<std::unique_ptr<Reg>> RegExes;
+  /// Represents a set of patterns and their line numbers
+  class Matcher {
+  public:
+    LLVM_ABI explicit Matcher(bool UseGlobs);
+
+    LLVM_ABI void
+    match(StringRef Query,
+          llvm::function_ref<void(StringRef Rule, unsigned LineNo)> Cb) const;
+
+    LLVM_ABI bool matchAny(StringRef Query) const {
+      bool R = false;
+      match(Query, [&](StringRef, unsigned) { R = true; });
+      return R;
+    }
+
+    LLVM_ABI Error insert(StringRef Pattern, unsigned LineNumber);
+
+    std::variant<RegexMatcher, GlobMatcher> M;
   };
 
   using SectionEntries = StringMap<StringMap<Matcher>>;
 
+protected:
   struct Section {
-    Section(StringRef Str, unsigned FileIdx)
-        : SectionStr(Str), FileIdx(FileIdx) {};
+    Section(StringRef Str, unsigned FileIdx, bool UseGlobs)
+        : SectionMatcher(UseGlobs), SectionStr(Str), FileIdx(FileIdx) {}
 
     Section(Section &&) = default;
 
@@ -186,11 +212,14 @@ class SpecialCaseList {
     findMatcher(StringRef Prefix, StringRef Category) const;
   };
 
+  ArrayRef<const Section> sections() const { return Sections; }
+
+private:
   std::vector<Section> Sections;
 
   LLVM_ABI Expected<Section *> addSection(StringRef SectionStr,
                                           unsigned FileIdx, unsigned LineNo,
-                                          bool UseGlobs = true);
+                                          bool UseGlobs);
 
   /// Parses just-constructed SpecialCaseList entries from a memory buffer.
   LLVM_ABI bool parse(unsigned FileIdx, const MemoryBuffer *MB,

llvm/lib/Support/SpecialCaseList.cpp

@@ -29,55 +29,79 @@
 
 namespace llvm {
 
-Error SpecialCaseList::Matcher::insert(StringRef Pattern, unsigned LineNumber,
-                                       bool UseGlobs) {
+Error SpecialCaseList::RegexMatcher::insert(StringRef Pattern,
+                                            unsigned LineNumber) {
   if (Pattern.empty())
     return createStringError(errc::invalid_argument,
-                             Twine("Supplied ") +
-                                 (UseGlobs ? "glob" : "regex") + " was blank");
-
-  if (!UseGlobs) {
-    // Replace * with .*
-    auto Regexp = Pattern.str();
-    for (size_t pos = 0; (pos = Regexp.find('*', pos)) != std::string::npos;
-         pos += strlen(".*")) {
-      Regexp.replace(pos, strlen("*"), ".*");
-    }
+                             "Supplied regex was blank");
 
-    Regexp = (Twine("^(") + StringRef(Regexp) + ")$").str();
+  // Replace * with .*
+  auto Regexp = Pattern.str();
+  for (size_t pos = 0; (pos = Regexp.find('*', pos)) != std::string::npos;
+       pos += strlen(".*")) {
+    Regexp.replace(pos, strlen("*"), ".*");
+  }
 
-    // Check that the regexp is valid.
-    Regex CheckRE(Regexp);
-    std::string REError;
-    if (!CheckRE.isValid(REError))
-      return createStringError(errc::invalid_argument, REError);
+  Regexp = (Twine("^(") + StringRef(Regexp) + ")$").str();
 
-    auto Rg =
-        std::make_unique<Matcher::Reg>(Pattern, LineNumber, std::move(CheckRE));
-    RegExes.emplace_back(std::move(Rg));
+  // Check that the regexp is valid.
+  Regex CheckRE(Regexp);
+  std::string REError;
+  if (!CheckRE.isValid(REError))
+    return createStringError(errc::invalid_argument, REError);
 
-    return Error::success();
-  }
+  auto Rg = std::make_unique<Reg>(Pattern, LineNumber, std::move(CheckRE));
+  RegExes.emplace_back(std::move(Rg));
 
-  auto Glob = std::make_unique<Matcher::Glob>(Pattern, LineNumber);
+  return Error::success();
+}
+
+void SpecialCaseList::RegexMatcher::match(
+    StringRef Query,
+    llvm::function_ref<void(StringRef Rule, unsigned LineNo)> Cb) const {
+  for (const auto &Regex : reverse(RegExes))
+    if (Regex->Rg.match(Query))
+      Cb(Regex->Name, Regex->LineNo);
+}
+
+Error SpecialCaseList::GlobMatcher::insert(StringRef Pattern,
+                                           unsigned LineNumber) {
+  if (Pattern.empty())
+    return createStringError(errc::invalid_argument, "Supplied glob was blank");
+
+  auto G = std::make_unique<Glob>(Pattern, LineNumber);
   // We must be sure to use the string in `Glob` rather than the provided
   // reference which could be destroyed before match() is called
-  if (auto Err = GlobPattern::create(Glob->Name, /*MaxSubPatterns=*/1024)
-                     .moveInto(Glob->Pattern))
+  if (auto Err = GlobPattern::create(G->Name, /*MaxSubPatterns=*/1024)
+                     .moveInto(G->Pattern))
     return Err;
-  Globs.push_back(std::move(Glob));
+  Globs.emplace_back(std::move(G));
   return Error::success();
 }
 
-void SpecialCaseList::Matcher::match(
+void SpecialCaseList::GlobMatcher::match(
     StringRef Query,
     llvm::function_ref<void(StringRef Rule, unsigned LineNo)> Cb) const {
   for (const auto &Glob : reverse(Globs))
     if (Glob->Pattern.match(Query))
       Cb(Glob->Name, Glob->LineNo);
-  for (const auto &Regex : reverse(RegExes))
-    if (Regex->Rg.match(Query))
-      Cb(Regex->Name, Regex->LineNo);
+}
+
+SpecialCaseList::Matcher::Matcher(bool UseGlobs) {
+  if (UseGlobs)
+    M.emplace<GlobMatcher>();
+  else
+    M.emplace<RegexMatcher>();
+}
+
+void SpecialCaseList::Matcher::match(
+    StringRef Query,
+    llvm::function_ref<void(StringRef Rule, unsigned LineNo)> Cb) const {
+  return std::visit([&](auto &V) { return V.match(Query, Cb); }, M);
+}
+
+Error SpecialCaseList::Matcher::insert(StringRef Pattern, unsigned LineNumber) {
+  return std::visit([&](auto &V) { return V.insert(Pattern, LineNumber); }, M);
 }
 
 // TODO: Refactor this to return Expected<...>
@@ -136,10 +160,10 @@ bool SpecialCaseList::createInternal(const MemoryBuffer *MB,
 Expected<SpecialCaseList::Section *>
 SpecialCaseList::addSection(StringRef SectionStr, unsigned FileNo,
                             unsigned LineNo, bool UseGlobs) {
-  Sections.emplace_back(SectionStr, FileNo);
+  Sections.emplace_back(SectionStr, FileNo, UseGlobs);
   auto &Section = Sections.back();
 
-  if (auto Err = Section.SectionMatcher.insert(SectionStr, LineNo, UseGlobs)) {
+  if (auto Err = Section.SectionMatcher.insert(SectionStr, LineNo)) {
     return createStringError(errc::invalid_argument,
                              "malformed section at line " + Twine(LineNo) +
                                  ": '" + SectionStr +
@@ -165,7 +189,7 @@ bool SpecialCaseList::parse(unsigned FileIdx, const MemoryBuffer *MB,
   bool UseGlobs = Version > 1;
 
   Section *CurrentSection;
-  if (auto Err = addSection("*", FileIdx, 1).moveInto(CurrentSection)) {
+  if (auto Err = addSection("*", FileIdx, 1, true).moveInto(CurrentSection)) {
     Error = toString(std::move(Err));
     return false;
   }
@@ -204,8 +228,9 @@ bool SpecialCaseList::parse(unsigned FileIdx, const MemoryBuffer *MB,
     }
 
     auto [Pattern, Category] = Postfix.split("=");
-    auto &Entry = CurrentSection->Entries[Prefix][Category];
-    if (auto Err = Entry.insert(Pattern, LineNo, UseGlobs)) {
+    auto [It, _] =
+        CurrentSection->Entries[Prefix].try_emplace(Category, UseGlobs);
+    if (auto Err = It->second.insert(Pattern, LineNo)) {
       Error =
           (Twine("malformed ") + (UseGlobs ? "glob" : "regex") + " in line " +
            Twine(LineNo) + ": '" + Pattern + "': " + toString(std::move(Err)))
@@ -262,4 +287,17 @@ unsigned SpecialCaseList::Section::getLastMatch(StringRef Prefix,
   return LastLine;
 }
 
+StringRef SpecialCaseList::Section::getLongestMatch(StringRef Prefix,
+                                                    StringRef Query,
+                                                    StringRef Category) const {
+  StringRef LongestRule;
+  if (const Matcher *M = findMatcher(Prefix, Category)) {
+    M->match(Query, [&](StringRef Rule, unsigned) {
+      if (LongestRule.size() < Rule.size())
+        LongestRule = Rule;
+    });
+  }
+  return LongestRule;
+}
+
 } // namespace llvm