perf: Replace unordered_map with bitset for vocabulary lookups (#2040)

Signed-off-by: Syed Azeez <syedazeez337@gmail.com>

Author: syedazeez337

src/core/jsonschema/CMakeLists.txt

@@ -4,9 +4,9 @@ include(./official_resolver.cmake)
 
 sourcemeta_library(NAMESPACE sourcemeta PROJECT core NAME jsonschema
   PRIVATE_HEADERS bundle.h walker.h frame.h error.h
-                  types.h transform.h
-  SOURCES jsonschema.cc official_walker.cc frame.cc
-          walker.cc bundle.cc transformer.cc format.cc
+                  types.h transform.h vocabularies.h
+  SOURCES jsonschema.cc vocabularies.cc official_walker.cc
+          frame.cc walker.cc bundle.cc transformer.cc format.cc
     "${CMAKE_CURRENT_BINARY_DIR}/official_resolver.cc")
 
 if(SOURCEMETA_CORE_INSTALL)

src/core/jsonschema/include/sourcemeta/core/jsonschema_types.h

@@ -3,21 +3,17 @@
 
 #include <sourcemeta/core/json.h>
 #include <sourcemeta/core/jsonpointer.h>
+#include <sourcemeta/core/jsonschema_vocabularies.h>
 
-#include <cstdint>       // std::uint8_t
-#include <functional>    // std::function, std::reference_wrapper
-#include <optional>      // std::optional
-#include <set>           // std::set
-#include <string>        // std::string
-#include <string_view>   // std::string_view
-#include <unordered_map> // std::unordered_map
+#include <cstdint>     // std::uint8_t
+#include <functional>  // std::function, std::reference_wrapper
+#include <optional>    // std::optional
+#include <set>         // std::set
+#include <string>      // std::string
+#include <string_view> // std::string_view
 
 namespace sourcemeta::core {
 
-/// @ingroup jsonschema
-/// A set of vocabularies
-using Vocabularies = std::unordered_map<JSON::String, bool>;
-
 // Take a URI and get back a schema
 /// @ingroup jsonschema
 ///

src/core/jsonschema/include/sourcemeta/core/jsonschema_vocabularies.h

@@ -0,0 +1,125 @@
+#ifndef SOURCEMETA_CORE_JSONSCHEMA_VOCABULARIES_H_
+#define SOURCEMETA_CORE_JSONSCHEMA_VOCABULARIES_H_
+
+#ifndef SOURCEMETA_CORE_JSONSCHEMA_EXPORT
+#include <sourcemeta/core/jsonschema_export.h>
+#endif
+
+#include <sourcemeta/core/json.h>
+
+#include <bitset>        // std::bitset
+#include <cassert>       // assert
+#include <cstdint>       // std::uint32_t, std::size_t
+#include <optional>      // std::optional
+#include <stdexcept>     // std::out_of_range
+#include <string>        // std::string
+#include <string_view>   // std::string_view
+#include <unordered_map> // std::unordered_map
+#include <utility>       // std::pair
+#include <vector>        // std::vector
+
+namespace sourcemeta::core {
+
+/// @ingroup jsonschema
+/// Optimized vocabulary set using bitflags for known vocabularies
+/// and a fallback `std::unordered_map` for custom vocabularies.
+///
+/// TODO: To maximize performance gains, convert string-based vocabulary checks
+/// throughout the codebase to use enum-based methods.
+struct SOURCEMETA_CORE_JSONSCHEMA_EXPORT Vocabularies {
+  enum class Known : std::uint8_t {
+    // Pre-vocabulary dialects (treated as vocabularies)
+    JSON_Schema_Draft_0 = 0,
+    JSON_Schema_Draft_0_Hyper = 1,
+    JSON_Schema_Draft_1 = 2,
+    JSON_Schema_Draft_1_Hyper = 3,
+    JSON_Schema_Draft_2 = 4,
+    JSON_Schema_Draft_2_Hyper = 5,
+    JSON_Schema_Draft_3 = 6,
+    JSON_Schema_Draft_3_Hyper = 7,
+    JSON_Schema_Draft_4 = 8,
+    JSON_Schema_Draft_4_Hyper = 9,
+    JSON_Schema_Draft_6 = 10,
+    JSON_Schema_Draft_6_Hyper = 11,
+    JSON_Schema_Draft_7 = 12,
+    JSON_Schema_Draft_7_Hyper = 13,
+    // 2019-09 vocabularies
+    JSON_Schema_2019_09_Core = 14,
+    JSON_Schema_2019_09_Applicator = 15,
+    JSON_Schema_2019_09_Validation = 16,
+    JSON_Schema_2019_09_Meta_Data = 17,
+    JSON_Schema_2019_09_Format = 18,
+    JSON_Schema_2019_09_Content = 19,
+    JSON_Schema_2019_09_Hyper_Schema = 20,
+    // 2020-12 vocabularies
+    JSON_Schema_2020_12_Core = 21,
+    JSON_Schema_2020_12_Applicator = 22,
+    JSON_Schema_2020_12_Unevaluated = 23,
+    JSON_Schema_2020_12_Validation = 24,
+    JSON_Schema_2020_12_Meta_Data = 25,
+    JSON_Schema_2020_12_Format_Annotation = 26,
+    JSON_Schema_2020_12_Format_Assertion = 27,
+    JSON_Schema_2020_12_Content = 28
+  };
+
+  // NOTE: Must be kept in sync with the Known enum above
+  static constexpr std::size_t KNOWN_VOCABULARY_COUNT = 29;
+
+public:
+  Vocabularies() = default;
+  Vocabularies(const Vocabularies &) = default;
+  Vocabularies(Vocabularies &&) noexcept = default;
+  auto operator=(const Vocabularies &) -> Vocabularies & = default;
+  auto operator=(Vocabularies &&) noexcept -> Vocabularies & = default;
+  ~Vocabularies() = default;
+
+  /// Construct from initializer list
+  Vocabularies(std::initializer_list<std::pair<JSON::String, bool>> init);
+
+  /// Construct from initializer list using known vocabulary enums
+  Vocabularies(std::initializer_list<std::pair<Known, bool>> init);
+
+  /// Check if a vocabulary is enabled
+  [[nodiscard]] auto contains(const JSON::String &uri) const noexcept -> bool;
+
+  /// Check if a known vocabulary is enabled
+  [[nodiscard]] auto contains(Known vocabulary) const noexcept -> bool;
+
+  /// Insert a vocabulary with its required/optional status
+  auto insert(const JSON::String &uri, bool required) noexcept -> void;
+
+  /// Insert a known vocabulary with its required/optional status
+  auto insert(Known vocabulary, bool required) noexcept -> void;
+
+  /// Get vocabulary status by URI
+  [[nodiscard]] auto get(const JSON::String &uri) const noexcept
+      -> std::optional<bool>;
+
+  /// Get known vocabulary status
+  [[nodiscard]] auto get(Known vocabulary) const noexcept
+      -> std::optional<bool>;
+
+  /// Get the number of vocabularies (required + optional + custom)
+  [[nodiscard]] auto size() const noexcept -> std::size_t;
+
+  /// Check if there are no vocabularies
+  [[nodiscard]] auto empty() const noexcept -> bool;
+
+private:
+  // Invariant: required_known and optional_known must be mutually exclusive
+  // A vocabulary can be either required (true) OR optional (false), never both
+#ifdef _MSC_VER
+#pragma warning(push)
+#pragma warning(disable : 4251)
+#endif
+  std::bitset<KNOWN_VOCABULARY_COUNT> required_known{};
+  std::bitset<KNOWN_VOCABULARY_COUNT> optional_known{};
+  std::unordered_map<JSON::String, bool> custom;
+#ifdef _MSC_VER
+#pragma warning(pop)
+#endif
+};
+
+} // namespace sourcemeta::core
+
+#endif

src/core/jsonschema/jsonschema.cc

@@ -259,18 +259,67 @@ auto sourcemeta::core::base_dialect(
 }
 
 namespace {
-auto core_vocabulary(std::string_view base_dialect) -> std::string {
+auto core_vocabulary_known(std::string_view base_dialect)
+    -> sourcemeta::core::Vocabularies::Known {
   if (base_dialect == "https://json-schema.org/draft/2020-12/schema" ||
       base_dialect == "https://json-schema.org/draft/2020-12/hyper-schema") {
-    return "https://json-schema.org/draft/2020-12/vocab/core";
+    return sourcemeta::core::Vocabularies::Known::JSON_Schema_2020_12_Core;
   } else if (base_dialect == "https://json-schema.org/draft/2019-09/schema" ||
              base_dialect ==
                  "https://json-schema.org/draft/2019-09/hyper-schema") {
-    return "https://json-schema.org/draft/2019-09/vocab/core";
+    return sourcemeta::core::Vocabularies::Known::JSON_Schema_2019_09_Core;
   } else {
     throw sourcemeta::core::SchemaBaseDialectError(std::string{base_dialect});
   }
 }
+
+auto dialect_to_known(std::string_view dialect)
+    -> std::optional<sourcemeta::core::Vocabularies::Known> {
+  using sourcemeta::core::Vocabularies;
+  if (dialect == "http://json-schema.org/draft-07/schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_7;
+  }
+  if (dialect == "http://json-schema.org/draft-07/hyper-schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_7_Hyper;
+  }
+  if (dialect == "http://json-schema.org/draft-06/schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_6;
+  }
+  if (dialect == "http://json-schema.org/draft-06/hyper-schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_6_Hyper;
+  }
+  if (dialect == "http://json-schema.org/draft-04/schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_4;
+  }
+  if (dialect == "http://json-schema.org/draft-04/hyper-schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_4_Hyper;
+  }
+  if (dialect == "http://json-schema.org/draft-03/schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_3;
+  }
+  if (dialect == "http://json-schema.org/draft-03/hyper-schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_3_Hyper;
+  }
+  if (dialect == "http://json-schema.org/draft-02/schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_2;
+  }
+  if (dialect == "http://json-schema.org/draft-02/hyper-schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_2_Hyper;
+  }
+  if (dialect == "http://json-schema.org/draft-01/schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_1;
+  }
+  if (dialect == "http://json-schema.org/draft-01/hyper-schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_1_Hyper;
+  }
+  if (dialect == "http://json-schema.org/draft-00/schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_0;
+  }
+  if (dialect == "http://json-schema.org/draft-00/hyper-schema#") {
+    return Vocabularies::Known::JSON_Schema_Draft_0_Hyper;
+  }
+  return std::nullopt;
+}
 } // namespace
 
 auto sourcemeta::core::vocabularies(
@@ -304,21 +353,22 @@ auto sourcemeta::core::vocabularies(const SchemaResolver &resolver,
   // As a performance optimization shortcut
   if (base_dialect == dialect) {
     if (dialect == "https://json-schema.org/draft/2020-12/schema") {
-      return {{"https://json-schema.org/draft/2020-12/vocab/core", true},
-              {"https://json-schema.org/draft/2020-12/vocab/applicator", true},
-              {"https://json-schema.org/draft/2020-12/vocab/unevaluated", true},
-              {"https://json-schema.org/draft/2020-12/vocab/validation", true},
-              {"https://json-schema.org/draft/2020-12/vocab/meta-data", true},
-              {"https://json-schema.org/draft/2020-12/vocab/format-annotation",
-               true},
-              {"https://json-schema.org/draft/2020-12/vocab/content", true}};
+      return Vocabularies{
+          {Vocabularies::Known::JSON_Schema_2020_12_Core, true},
+          {Vocabularies::Known::JSON_Schema_2020_12_Applicator, true},
+          {Vocabularies::Known::JSON_Schema_2020_12_Unevaluated, true},
+          {Vocabularies::Known::JSON_Schema_2020_12_Validation, true},
+          {Vocabularies::Known::JSON_Schema_2020_12_Meta_Data, true},
+          {Vocabularies::Known::JSON_Schema_2020_12_Format_Annotation, true},
+          {Vocabularies::Known::JSON_Schema_2020_12_Content, true}};
     } else if (dialect == "https://json-schema.org/draft/2019-09/schema") {
-      return {{"https://json-schema.org/draft/2019-09/vocab/core", true},
-              {"https://json-schema.org/draft/2019-09/vocab/applicator", true},
-              {"https://json-schema.org/draft/2019-09/vocab/validation", true},
-              {"https://json-schema.org/draft/2019-09/vocab/meta-data", true},
-              {"https://json-schema.org/draft/2019-09/vocab/format", false},
-              {"https://json-schema.org/draft/2019-09/vocab/content", true}};
+      return Vocabularies{
+          {Vocabularies::Known::JSON_Schema_2019_09_Core, true},
+          {Vocabularies::Known::JSON_Schema_2019_09_Applicator, true},
+          {Vocabularies::Known::JSON_Schema_2019_09_Validation, true},
+          {Vocabularies::Known::JSON_Schema_2019_09_Meta_Data, true},
+          {Vocabularies::Known::JSON_Schema_2019_09_Format, false},
+          {Vocabularies::Known::JSON_Schema_2019_09_Content, true}};
     }
   }
 
@@ -336,7 +386,11 @@ auto sourcemeta::core::vocabularies(const SchemaResolver &resolver,
       dialect == "http://json-schema.org/draft-02/schema#" ||
       dialect == "http://json-schema.org/draft-01/schema#" ||
       dialect == "http://json-schema.org/draft-00/schema#") {
-    return {{dialect, true}};
+    const auto known = dialect_to_known(dialect);
+    if (known.has_value()) {
+      return Vocabularies{{known.value(), true}};
+    }
+    return Vocabularies{{dialect, true}};
   }
 
   /*
@@ -356,7 +410,11 @@ auto sourcemeta::core::vocabularies(const SchemaResolver &resolver,
       base_dialect == "http://json-schema.org/draft-02/hyper-schema#" ||
       base_dialect == "http://json-schema.org/draft-01/hyper-schema#" ||
       base_dialect == "http://json-schema.org/draft-00/hyper-schema#") {
-    return {{base_dialect, true}};
+    const auto known = dialect_to_known(base_dialect);
+    if (known.has_value()) {
+      return Vocabularies{{known.value(), true}};
+    }
+    return Vocabularies{{base_dialect, true}};
   }
 
   /*
@@ -384,25 +442,28 @@ auto sourcemeta::core::vocabularies(const SchemaResolver &resolver,
    */
 
   Vocabularies result;
-  const std::string core{core_vocabulary(base_dialect)};
+  const auto core{core_vocabulary_known(base_dialect)};
   if (schema_dialect.defines("$vocabulary")) {
     const sourcemeta::core::JSON &vocabularies{
         schema_dialect.at("$vocabulary")};
     assert(vocabularies.is_object());
     for (const auto &entry : vocabularies.as_object()) {
-      result.insert({entry.first, entry.second.to_boolean()});
+      result.insert(entry.first, entry.second.to_boolean());
     }
   } else {
-    result.insert({core, true});
+    result.insert(core, true);
   }
 
   // The specification recommends these checks
   if (!result.contains(core)) {
     throw sourcemeta::core::SchemaError(
         "The core vocabulary must always be present");
-  } else if (!result.at(core)) {
-    throw sourcemeta::core::SchemaError(
-        "The core vocabulary must always be required");
+  } else {
+    const auto core_status{result.get(core)};
+    if (core_status.has_value() && !core_status.value()) {
+      throw sourcemeta::core::SchemaError(
+          "The core vocabulary must always be required");
+    }
   }
 
   return result;