serialbench 0.10.0 → 0.12.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: fb37afdfb65ed1d7226788c226fad30d1779b39578ad762bf097783fea638b08
-  data.tar.gz: 8d7a5f9ae73d6b2cd74b84d9861e4fbc9934c371960cc578d3bd0cef2798bc44
+  metadata.gz: fb1d328cda41b77014e07bcca03086cf860a81a9f02c8863918e37d41ab1e201
+  data.tar.gz: 889ba7216bdc343cf5886f2fdaf6257452633161cd291c66e54823caa4cfbb7d
 SHA512:
-  metadata.gz: 4a426f90212feb0b50c4a5b488f82bac91bbdfe30208cefc66390baa864ff217b92d551c38504d666dee1aa6f96b389cbdda65a35b91fac0cb34909ae5a280b9
-  data.tar.gz: 1202d30b4bfc472ea3a30f7b66f5eb0ac2874933411031b0add26ceb3f69dd1fd0d469df71a39665cfc8a53fb84dabf29c7f51811a6e121247d486ea2e235be8
+  metadata.gz: 64d0c977c7efffaa614ac03425eae1cf5c3e83d44f06bc2ba98da8b03671644887ec8813d8c506987e6d0f0ce74424af5bf74bf025ad9a09303f7d57a0c3ffa4
+  data.tar.gz: 4db8775576b87ff4eb268f063e323527f5f639d3c65c55f12f02078d7772ff552dc67578f048179a7c17d0a33a08058527959f96f2b0403595ce349c4bb9c79c

data/.github/workflows/benchmark.yml

@@ -116,27 +116,21 @@ jobs:
       - name: Install gems
         run: bundle install
 
+      - name: Fetch Saxon-HE (XSLT 2.0/3.0 comparison)
+        run: |
+          mkdir -p vendor
+          curl -sL -o vendor/saxonhe.jar "https://repo1.maven.org/maven2/net/sf/saxon/Saxon-HE/12.10/Saxon-HE-12.10.jar"
+          curl -sL -o vendor/xmlresolver.jar "https://repo1.maven.org/maven2/org/xmlresolver/xmlresolver/5.2.2/xmlresolver-5.2.2.jar"
+          echo "SAXON_JAR=$PWD/vendor/saxonhe.jar" >> "$GITHUB_ENV"
+          echo "XMLRESOLVER_JAR=$PWD/vendor/xmlresolver.jar" >> "$GITHUB_ENV"
+        shell: bash
+
       - name: Load canonical fixtures
         run: |
           git clone --depth 1 https://github.com/serialbench/fixtures.git test_data
           (cd test_data && sha256sum -c SHA256SUMS) || (cd test_data && shasum -a 256 -c SHA256SUMS)
         shell: bash
 
-      - name: Point teptris at its vendored DLL (Windows)
-        if: startsWith(matrix.platform, 'windows-')
-        shell: bash
-        run: |
-          # teptris 0.1.0's mingw gem vendors the library as teptris.dll but
-          # its FFI loader searches libteptris.dll; TEPTRIS_LIB_PATH (the
-          # loader's first candidate) bridges until the ext-based gem ships.
-          DLL=$(bundle exec ruby -e "print File.join(Gem::Specification.find_by_name('teptris').gem_dir, 'lib', 'teptris.dll')" 2>/dev/null || true)
-          if [ -n "$DLL" ] && [ -f "$DLL" ]; then
-            echo "TEPTRIS_LIB_PATH=$DLL" >> "$GITHUB_ENV"
-            echo "TEPTRIS_LIB_PATH=$DLL"
-          else
-            echo "no vendored teptris.dll found — teptris stays unavailable on this leg"
-          fi
-
       - name: Create environment config
         run: |
           mkdir -p config/environments

data/Gemfile

@@ -12,12 +12,9 @@ unless Gem.win_platform? && RUBY_PLATFORM.include?('aarch64')
   gem 'libxml-ruby'
 end
 
-gem 'leptris', '1.9.156.2'  # precompiled platform gems bundle libleptris (darwin/linux/mingw)
-gem 'yeptris', '0.2.2.1'    # precompiled darwin/linux; no windows or darwin-intel gems yet
-# teptris stays 0.1.0: 0.2.1's native ext links libruby by absolute build-runner
-# path (loads only under exactly Ruby 3.3.12 at hostedtoolcache) and ships no
-# mingw gems. Bump when the ext links portably and windows variants exist.
-gem 'teptris', '0.1.0'
+gem 'leptris', '1.9.193.4'  # html5 conformance campaign + memchr tokenizer perf
+gem 'yeptris', '0.6.5.4'    # compiled plan walk; dump round-trip fixed (plain keys)
+gem 'teptris', '0.2.36'     # full platform matrix incl. mingw - windows coverage unblocked
 gem 'benchmark'  # Removed from stdlib in Ruby 4.0
 gem 'base64'  # Required for Ruby 3.4+
 gem 'lutaml-model', '~> 0.7'

data/config/benchmarks/full-xml.yml

@@ -23,6 +23,7 @@ operations:
 - xpath
 - xquery
 - xslt
+- xslt30
 - validation
 - memory
 

data/lib/serialbench/benchmark_runner.rb

@@ -142,10 +142,39 @@ _DEFAULT_XSLT_STYLESHEET = <<'XSL'.freeze
 </xsl:stylesheet>
 XSL
 
+_DEFAULT_XSLT30_STYLESHEET = <<'XSL30'.freeze
+<?xml version="1.0" encoding="UTF-8"?>
+<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+                xmlns:xs="http://www.w3.org/2001/XMLSchema"
+                exclude-result-prefixes="xs"
+                version="3.0">
+  <xsl:output method="xml" indent="yes"/>
+  <xsl:template match="/">
+    <report30>
+      <xsl:iterate select="//user | //record | //database | //cache">
+        <xsl:param name="labels" as="xs:string*" select="()"/>
+        <xsl:on-completion>
+          <joined><xsl:value-of select="string-join($labels, ', ')"/></joined>
+          <count><xsl:value-of select="count($labels)"/></count>
+        </xsl:on-completion>
+        <xsl:variable name="label" select="string((name, data/field1, ttl)[1])"/>
+        <xsl:if test="$label != ''">
+          <row kind="{local-name()}" label="{upper-case($label)}"/>
+        </xsl:if>
+        <xsl:next-iteration>
+          <xsl:with-param name="labels" select="if ($label != '') then ($labels, $label) else $labels"/>
+        </xsl:next-iteration>
+      </xsl:iterate>
+    </report30>
+  </xsl:template>
+</xsl:stylesheet>
+XSL30
+
     XPATH_QUERIES = ['//book', "//book[@id='101']", '//book[price > 30]/title'].freeze
     XQUERY_EXPRESSIONS = ['count(//user | //record)', "//record[@id='101']/data/field1", '//user[profile/age > 40]/name'].freeze
     RNG_SCHEMA = File.expand_path('test_data/schema.rng', Dir.pwd).then { |p| File.exist?(p) ? File.read(p) : _DEFAULT_RNG_SCHEMA }
     XSLT_STYLESHEET = File.expand_path('test_data/transform.xsl', Dir.pwd).then { |p| File.exist?(p) ? File.read(p) : _DEFAULT_XSLT_STYLESHEET }
+    XSLT30_STYLESHEET = File.expand_path('test_data/transform30.xsl', Dir.pwd).then { |p| File.exist?(p) ? File.read(p) : _DEFAULT_XSLT30_STYLESHEET }
 
     OPERATIONS = {
       'parsing' => ->(s, data) { s.parse(data) },
@@ -158,7 +187,8 @@ XSL
         doc = s.parse(data)
         XQUERY_EXPRESSIONS.each { |x| s.xquery_eval(doc, x) }
       },
-      'xslt' => ->(s, data) { s.xslt_transform(s.parse(data), XSLT_STYLESHEET) },
+      'xslt' => ->(s, data) { s.xslt_apply(data, XSLT_STYLESHEET) },
+      'xslt30' => ->(s, data) { s.xslt_apply(data, XSLT30_STYLESHEET) },
       'validation' => ->(s, data) { s.validate(s.parse(data), RNG_SCHEMA) },
       'streaming' => ->(s, data) { s.stream_parse(data) { |_event, _data| } },
     }.freeze
@@ -280,6 +310,8 @@ XSL
       serializers = Serializers.for_format(format)
 
       case type_name
+      when 'parsing', 'memory'
+        serializers.select { |s| s.supports?(:parse) }
       when 'generation'
         serializers.select { |s| s.supports?(:generate) }
       when 'streaming'
@@ -292,6 +324,8 @@ XSL
         serializers.select { |s| s.supports?(:xslt) }
       when 'validation'
         serializers.select { |s| s.supports?(:validation) }
+      when 'xslt30'
+        serializers.select { |s| s.supports?(:xslt30) }
       else
         serializers
       end

data/lib/serialbench/models/benchmark_config.rb

@@ -51,7 +51,7 @@ module Serialbench
       attribute :formats, :string, collection: true, values: %w[xml html json yaml toml]
       attribute :iterations, BenchmarkIteration
       attribute :warmup, :integer, default: -> { 1 }
-      attribute :operations, :string, collection: true, values: %w[parsing generation xpath xquery xslt validation streaming memory]
+      attribute :operations, :string, collection: true, values: %w[parsing generation xpath xquery xslt xslt30 validation streaming memory]
 
       key_value do
         map 'name', to: :name

data/lib/serialbench/models/benchmark_result.rb

@@ -73,6 +73,7 @@ module Serialbench
       attribute :xpath, IterationPerformance, collection: true
       attribute :xquery, IterationPerformance, collection: true
       attribute :xslt, IterationPerformance, collection: true
+      attribute :xslt30, IterationPerformance, collection: true
       attribute :validation, IterationPerformance, collection: true
 
       key_value do
@@ -84,6 +85,7 @@ module Serialbench
         map 'xpath', to: :xpath
         map 'xquery', to: :xquery
         map 'xslt', to: :xslt
+        map 'xslt30', to: :xslt30
         map 'validation', to: :validation
       end
     end

data/lib/serialbench/serializers/xml/base_xml_serializer.rb

@@ -27,6 +27,10 @@ module Serialbench
           super | Set.new(%i[namespaces])
         end
 
+        def xslt_apply(source_xml, stylesheet)
+          xslt_transform(parse(source_xml), stylesheet)
+        end
+
         # XML-specific features derive from the capability set
         def features
           {

data/lib/serialbench/serializers/xml/leptris_serializer.rb

@@ -37,7 +37,7 @@ module Serialbench
         end
 
         def capabilities
-          super | Set.new(%i[xpath xquery xslt validation sax stax])
+          super | Set.new(%i[xpath xquery xslt xslt30 validation sax stax])
         end
 
         def xquery_eval(document, expression)

data/lib/serialbench/serializers/xml/saxon_serializer.rb

@@ -0,0 +1,72 @@
+# frozen_string_literal: true
+
+require_relative 'base_xml_serializer'
+require 'open3'
+require 'tmpdir'
+
+module Serialbench
+  module Serializers
+    module Xml
+      # Saxon-HE via its Java CLI: the reference XSLT 2.0/3.0 engine.
+      # Cold-process semantics: every iteration is a full JVM + compile +
+      # transform, so ips is comparable to leptris only end-to-end, never
+      # per-transform — features record cold_process: true.
+      class SaxonSerializer < BaseXmlSerializer
+        def name
+          'saxon-he'
+        end
+
+        def capabilities
+          Set.new(%i[xslt xslt30])
+        end
+
+        def features
+          { 'xslt_version' => '3.0', 'cold_process' => true }
+        end
+
+        def classpath
+          jars = [ENV['SAXON_JAR'], ENV['XMLRESOLVER_JAR']].compact
+          jars.empty? ? nil : jars.join(File::PATH_SEPARATOR)
+        end
+
+        def xslt_apply(source_xml, stylesheet)
+          Dir.mktmpdir('saxonbench') do |dir|
+            src = File.join(dir, 's.xml')
+            xsl = File.join(dir, 's.xsl')
+            File.write(src, source_xml)
+            File.write(xsl, stylesheet)
+            out, _err, _st = Open3.capture3('java', '-cp', classpath, 'net.sf.saxon.Transform',
+                                            "-xsl:#{xsl}", "-s:#{src}")
+            raise Error, "saxon transform failed: #{out}#{_err}"[0, 200] unless _st.success?
+
+            out
+          end
+        end
+
+        def available?
+          return @available if defined?(@available)
+
+          @available = begin
+            cp = classpath
+            raise LoadError, 'SAXON_JAR / XMLRESOLVER_JAR not set' unless cp
+
+            out, _err, st = Open3.capture3('java', '-cp', cp, 'net.sf.saxon.Version')
+            raise LoadError, 'java failed' unless st.success?
+
+            true
+          rescue StandardError, LoadError => e
+            warn "#{name} unavailable: #{e.class}: #{e.message}"
+            false
+          end
+        end
+
+        def version
+          return 'unknown' unless available?
+
+          out, = Open3.capture3('java', '-cp', classpath, 'net.sf.saxon.Version')
+          out[/Saxon-HE (\S+)/, 1] || out.strip[0, 40]
+        end
+      end
+    end
+  end
+end

data/lib/serialbench/serializers.rb

@@ -11,6 +11,7 @@ require_relative 'serializers/xml/nokogiri_serializer'
 require_relative 'serializers/xml/oga_serializer'
 require_relative 'serializers/xml/libxml_serializer'
 require_relative 'serializers/xml/leptris_serializer'
+require_relative 'serializers/xml/saxon_serializer'
 
 # HTML Serializers
 require_relative 'serializers/html/base_html_serializer'
@@ -49,7 +50,8 @@ module Serialbench
         Xml::NokogiriSerializer,
         Xml::OgaSerializer,
         Xml::LibxmlSerializer,
-        Xml::LeptrisSerializer
+        Xml::LeptrisSerializer,
+        Xml::SaxonSerializer
       ],
       json: [
         Json::JsonSerializer,

data/lib/serialbench/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Serialbench
-  VERSION = '0.10.0'
+  VERSION = '0.12.0'
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: serialbench
 version: !ruby/object:Gem::Version
-  version: 0.10.0
+  version: 0.12.0
 platform: ruby
 authors:
 - Ribose
@@ -346,6 +346,7 @@ files:
 - lib/serialbench/serializers/xml/oga_serializer.rb
 - lib/serialbench/serializers/xml/ox_serializer.rb
 - lib/serialbench/serializers/xml/rexml_serializer.rb
+- lib/serialbench/serializers/xml/saxon_serializer.rb
 - lib/serialbench/serializers/yaml/base_yaml_serializer.rb
 - lib/serialbench/serializers/yaml/psych_serializer.rb
 - lib/serialbench/serializers/yaml/syck_serializer.rb