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

[vitalybuka]

Glob will be optimized
Regex we we will keep intact.

Using std::variant to avoid virtual methods,
and allow to switch unique_ptr to move in future.

[llvmbot]

@llvm/pr-subscribers-llvm-support

Author: Vitaly Buka (vitalybuka)

Changes

I am going to optimize the later, but keep the former intact.

---

Full diff: https://github.com/llvm/llvm-project/pull/162303.diff

2 Files Affected:

• (modified) llvm/include/llvm/Support/SpecialCaseList.h (+30-13)
• (modified) llvm/lib/Support/SpecialCaseList.cpp (+57-32)

diff --git a/llvm/include/llvm/Support/SpecialCaseList.h b/llvm/include/llvm/Support/SpecialCaseList.h
index e824cd44f6297..eff36569fcab7 100644
--- a/llvm/include/llvm/Support/SpecialCaseList.h
+++ b/llvm/include/llvm/Support/SpecialCaseList.h
@@ -19,6 +19,7 @@
 #include <memory>
 #include <string>
 #include <utility>
+#include <variant>
 #include <vector>
 
 namespace llvm {
@@ -118,18 +119,20 @@ class SpecialCaseList {
   SpecialCaseList(SpecialCaseList const &) = delete;
   SpecialCaseList &operator=(SpecialCaseList const &) = delete;
 
-  /// Represents a set of globs and their line numbers
-  class Matcher {
+  // Lagacy v1 matcher.
+  class RegexMatcher {
+  public:
+    LLVM_ABI unsigned match(StringRef Query) const;
+    LLVM_ABI Error insert(StringRef Pattern, unsigned LineNumber);
+    std::vector<std::pair<std::unique_ptr<Regex>, unsigned>> RegExes;
+  };
+
+  class GlobMatcher {
   public:
     // Returns the line number in the source file that this query matches to.
     // Returns zero if no match is found.
     LLVM_ABI unsigned match(StringRef Query) const;
-
-  private:
-    friend class SpecialCaseList;
-    LLVM_ABI Error insert(StringRef Pattern, unsigned LineNumber,
-                          bool UseRegex);
-
+    LLVM_ABI Error insert(StringRef Pattern, unsigned LineNumber);
     struct Glob {
       std::string Name;
       unsigned LineNo;
@@ -140,15 +143,29 @@ class SpecialCaseList {
       Glob() = default;
     };
 
-    std::vector<std::unique_ptr<Matcher::Glob>> Globs;
-    std::vector<std::pair<std::unique_ptr<Regex>, unsigned>> RegExes;
+    std::vector<std::unique_ptr<GlobMatcher::Glob>> Globs;
+  };
+
+  /// Represents a set of globs and their line numbers
+  class Matcher {
+  public:
+    LLVM_ABI explicit Matcher(bool UseGlobs);
+    // Returns the line number in the source file that this query matches to.
+    // Returns zero if no match is found.
+    LLVM_ABI unsigned match(StringRef Query) const;
+
+  private:
+    friend class SpecialCaseList;
+    LLVM_ABI Error insert(StringRef Pattern, unsigned LineNumber);
+
+    std::variant<RegexMatcher, GlobMatcher> M;
   };
 
   using SectionEntries = StringMap<StringMap<Matcher>>;
 
   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;
 
@@ -162,7 +179,7 @@ class SpecialCaseList {
 
   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,
diff --git a/llvm/lib/Support/SpecialCaseList.cpp b/llvm/lib/Support/SpecialCaseList.cpp
index 4b038850b62ca..71f7b9aa65796 100644
--- a/llvm/lib/Support/SpecialCaseList.cpp
+++ b/llvm/lib/Support/SpecialCaseList.cpp
@@ -25,36 +25,48 @@
 
 namespace llvm {
 
-Error SpecialCaseList::Matcher::insert(StringRef Pattern, unsigned LineNumber,
-                                       bool UseGlobs) {
-  if (Pattern.empty())
+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");
+  }
+
+  // Replace * with .*
+  auto Regexp = Pattern.str();
+  for (size_t pos = 0; (pos = Regexp.find('*', pos)) != std::string::npos;
+       pos += strlen(".*")) {
+    Regexp.replace(pos, strlen("*"), ".*");
+  }
+
+  Regexp = (Twine("^(") + StringRef(Regexp) + ")$").str();
+
+  // 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();
+  RegExes.emplace_back(
+      std::make_pair(std::make_unique<Regex>(std::move(CheckRE)), LineNumber));
 
-    // 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();
+}
 
-    RegExes.emplace_back(std::make_pair(
-        std::make_unique<Regex>(std::move(CheckRE)), LineNumber));
+unsigned SpecialCaseList::RegexMatcher::match(StringRef Query) const {
+  for (const auto &[Regex, LineNumber] : reverse(RegExes))
+    if (Regex->match(Query))
+      return LineNumber;
+  return 0;
+}
 
-    return Error::success();
+Error SpecialCaseList::GlobMatcher::insert(StringRef Pattern,
+                                           unsigned LineNumber) {
+  if (Pattern.empty()) {
+    return createStringError(errc::invalid_argument, "Supplied glob was blank");
   }
 
-  auto Glob = std::make_unique<Matcher::Glob>();
+  auto Glob = std::make_unique<GlobMatcher::Glob>();
   Glob->Name = Pattern.str();
   Glob->LineNo = LineNumber;
   // We must be sure to use the string in `Glob` rather than the provided
@@ -66,16 +78,28 @@ Error SpecialCaseList::Matcher::insert(StringRef Pattern, unsigned LineNumber,
   return Error::success();
 }
 
-unsigned SpecialCaseList::Matcher::match(StringRef Query) const {
+unsigned SpecialCaseList::GlobMatcher::match(StringRef Query) const {
   for (const auto &Glob : reverse(Globs))
     if (Glob->Pattern.match(Query))
       return Glob->LineNo;
-  for (const auto &[Regex, LineNumber] : reverse(RegExes))
-    if (Regex->match(Query))
-      return LineNumber;
   return 0;
 }
 
+SpecialCaseList::Matcher::Matcher(bool UseGlobs) {
+  if (UseGlobs)
+    M.emplace<GlobMatcher>();
+  else
+    M.emplace<RegexMatcher>();
+}
+
+unsigned SpecialCaseList::Matcher::match(StringRef Query) const {
+  return std::visit([&](auto &V) { return V.match(Query); }, 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<...>
 std::unique_ptr<SpecialCaseList>
 SpecialCaseList::create(const std::vector<std::string> &Paths,
@@ -132,10 +156,10 @@ 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 +
@@ -148,7 +172,7 @@ SpecialCaseList::addSection(StringRef SectionStr, unsigned FileNo,
 bool SpecialCaseList::parse(unsigned FileIdx, const MemoryBuffer *MB,
                             std::string &Error) {
   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;
   }
@@ -194,8 +218,9 @@ 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)))