kreuzberg 4.5.2-aarch64-linux → 4.6.0-aarch64-linux

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 5d37d9bc188b071a5a193db3a4cf610caaff6e8b9a84ef04bebfbf2abd5f7f16
-  data.tar.gz: 7d568f4f6b5e8e1be95ac814b55974d348a460c1ab4949aaafe3b0de9a3d9dbe
+  metadata.gz: e120139c5153d954f686166a3c23f0a3c64a24dcbbcf2ff1a9d17abeafbdc22d
+  data.tar.gz: a17475f255eceefe8a6c0702866cf00c401cdbc294c8b5eace4cae78b9401161
 SHA512:
-  metadata.gz: 9d5a0a4d9e6917ef0ac38e319410752e0032de7742a167a2a34e5c48356f1af145d1a2f56f0d7d6d28fef439c4f82710655116f03f1077c851f521e5fc4ba09c
-  data.tar.gz: '01898e03048b285ce8347ec975041bea2e633fed829abeaf8d1842fcc5137d8dacecf54f971ddc5a9a61261dc45ac30de3c408f8bab696af4597a648e0683a31'
+  metadata.gz: '01525768eda28284eca9d077377d052d6fb90834282a26a73d8f8b2c3dbfcbad573bb71c51bbf0679329e59f205cd097f23080bdb10a7ad0432d700092aa7501'
+  data.tar.gz: ef96711a4c445cb7487d253731cfd697a0dd179f675efe8d3c9d2a6f6033b2257ada11f4a343cf0cf41a2160e814b3565e4160a0cf3f0a24f08d1765d59e6107

data/README.md

@@ -22,7 +22,7 @@
     <img src="https://img.shields.io/maven-central/v/dev.kreuzberg/kreuzberg?label=Java&color=007ec6" alt="Java">
   </a>
   <a href="https://github.com/kreuzberg-dev/kreuzberg/releases">
-    <img src="https://img.shields.io/github/v/tag/kreuzberg-dev/kreuzberg?label=Go&color=007ec6&filter=v4.5.2" alt="Go">
+    <img src="https://img.shields.io/github/v/tag/kreuzberg-dev/kreuzberg?label=Go&color=007ec6&filter=v4.6.0" alt="Go">
   </a>
   <a href="https://www.nuget.org/packages/Kreuzberg/">
     <img src="https://img.shields.io/nuget/v/Kreuzberg?label=C%23&color=007ec6" alt="C#">
@@ -61,7 +61,7 @@
 </div>
 
 
-Extract text, tables, images, and metadata from 88+ file formats including PDF, Office documents, and images. Ruby bindings with idiomatic Ruby API and native performance.
+Extract text, tables, images, and metadata from 91+ file formats including PDF, Office documents, and images. Ruby bindings with idiomatic Ruby API and native performance.
 
 
 ## Installation
@@ -211,9 +211,9 @@ puts "Processing time: #{result.metadata&.dig('processing_time')}ms"
 
 ## Features
 
-### Supported File Formats (88+)
+### Supported File Formats (91+)
 
-88+ file formats across 8 major categories with intelligent format detection and comprehensive metadata extraction.
+91+ file formats across 8 major categories with intelligent format detection and comprehensive metadata extraction.
 
 #### Office Documents
 

data/lib/kreuzberg/config.rb

@@ -850,19 +850,21 @@ module Kreuzberg
     #   )
     #
     class LayoutDetection
-      attr_reader :preset, :confidence_threshold, :apply_heuristics
+      attr_reader :preset, :confidence_threshold, :apply_heuristics, :table_model
 
-      def initialize(preset: 'fast', confidence_threshold: nil, apply_heuristics: true)
+      def initialize(preset: 'fast', confidence_threshold: nil, apply_heuristics: true, table_model: nil)
         @preset = preset.to_s
         @confidence_threshold = confidence_threshold&.to_f
         @apply_heuristics = apply_heuristics ? true : false
+        @table_model = table_model&.to_s
       end
 
       def to_h
         {
           preset: @preset,
           confidence_threshold: @confidence_threshold,
-          apply_heuristics: @apply_heuristics
+          apply_heuristics: @apply_heuristics,
+          table_model: @table_model
         }.compact
       end
     end

data/lib/kreuzberg/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kreuzberg
-  VERSION = '4.5.2'
+  VERSION = '4.6.0'
 end

data/sig/kreuzberg.rbs

@@ -463,8 +463,9 @@ module Kreuzberg
       attr_reader preset: String
       attr_reader confidence_threshold: Float?
       attr_reader apply_heuristics: bool
+      attr_reader table_model: String?
 
-      def initialize: (?preset: String, ?confidence_threshold: Float?, ?apply_heuristics: bool) -> void
+      def initialize: (?preset: String, ?confidence_threshold: Float?, ?apply_heuristics: bool, ?table_model: String?) -> void
       def to_h: () -> Hash[Symbol, untyped]
     end
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: kreuzberg
 version: !ruby/object:Gem::Version
-  version: 4.5.2
+  version: 4.6.0
 platform: aarch64-linux
 authors:
 - Na'aman Hirschfeld
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-03-21 00:00:00.000000000 Z
+date: 2026-03-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -175,18 +175,7 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- ".gitignore"
-- ".rspec"
-- ".rubocop.yaml"
-- ".rubocop.yml"
-- Gemfile
-- Gemfile.lock
 - README.md
-- Rakefile
-- Steepfile
-- examples/async_patterns.rb
-- extconf.rb
-- kreuzberg.gemspec
 - lib/kreuzberg.rb
 - lib/kreuzberg/api_proxy.rb
 - lib/kreuzberg/cache_api.rb
@@ -232,9 +221,6 @@ files:
 - spec/binding/plugins/postprocessor_spec.rb
 - spec/binding/plugins/validator_spec.rb
 - spec/binding/tables_spec.rb
-- spec/fixtures/config.toml
-- spec/fixtures/config.yaml
-- spec/fixtures/invalid_config.toml
 - spec/serialization_spec.rb
 - spec/smoke/package_spec.rb
 - spec/spec_helper.rb
@@ -254,7 +240,6 @@ files:
 - spec/unit/config/postprocessor_config_spec.rb
 - spec/unit/config/tesseract_config_spec.rb
 - spec/unit/config/token_reduction_config_spec.rb
-- test/metadata_types_test.rb
 homepage: https://github.com/kreuzberg-dev/kreuzberg
 licenses:
 - MIT

data/.gitignore

@@ -1,14 +0,0 @@
-tmp/
-lib/*.bundle
-lib/*.dylib
-lib/*.so
-lib/*.dll
-lib/*.dylib
-lib/*.so
-lib/*.dll
-
-# Vendor directory for local development (symlink to ../../crates/kreuzberg)
-# In CI, this is replaced by the actual vendored crate files
-# Exception: vendor/rb-sys is patched and committed
-!vendor/rb-sys/
-vendor/

data/.rspec

@@ -1,3 +0,0 @@
---require spec_helper
---format documentation
---color

data/.rubocop.yaml

@@ -1 +0,0 @@
-inherit_from: .rubocop.yml

data/.rubocop.yml

@@ -1,543 +0,0 @@
-
-plugins:
-  - rubocop-performance
-  - rubocop-rspec
-
-AllCops:
-  TargetRubyVersion: 3.2
-  NewCops: enable
-  SuggestExtensions: false
-  Exclude:
-    - 'vendor/**/*'
-    - 'tmp/**/*'
-    - 'lib/**/*.bundle'
-    - 'ext/**/*'
-
-Style/StringLiterals:
-  Enabled: true
-  EnforcedStyle: single_quotes
-
-Style/StringLiteralsInInterpolation:
-  Enabled: true
-  EnforcedStyle: single_quotes
-
-Style/FrozenStringLiteralComment:
-  Enabled: true
-  EnforcedStyle: always
-
-Style/Documentation:
-  Enabled: false
-
-Layout/LineLength:
-  Max: 120
-  AllowedPatterns:
-    - '\A\s*#'
-  Exclude:
-    - 'spec/**/*'
-
-Metrics/BlockLength:
-  Enabled: true
-  Max: 350
-  CountComments: false
-  CountAsOne: []
-
-Metrics/MethodLength:
-  Max: 18
-  Exclude:
-    - 'spec/**/*'
-    - 'examples/**/*'
-
-Metrics/AbcSize:
-  Max: 20
-  Exclude:
-    - 'spec/**/*'
-    - 'examples/**/*'
-    - 'lib/kreuzberg/config.rb'
-
-Naming/FileName:
-  Enabled: true
-  Exclude:
-    - 'Gemfile'
-    - 'Rakefile'
-
-RSpec/SpecFilePathFormat:
-  Enabled: false
-RSpec/SpecFilePathSuffix:
-  Enabled: false
-
-RSpec/PendingWithoutReason:
-  Enabled: false
-
-RSpec/LeakyConstantDeclaration:
-  Enabled: false
-
-RSpec/UnspecifiedException:
-  Enabled: false
-
-Lint/ConstantDefinitionInBlock:
-  Enabled: false
-  Exclude:
-    - 'spec/**/*'
-
-RSpec/InstanceVariable:
-  Enabled: false
-
-Gemspec/DevelopmentDependencies:
-  Enabled: false
-
-Metrics/ParameterLists:
-  Max: 14
-  Exclude:
-    - 'lib/kreuzberg/config.rb'
-
-Metrics/CyclomaticComplexity:
-  Max: 7
-  Exclude:
-    - 'lib/kreuzberg/config.rb'
-
-Metrics/PerceivedComplexity:
-  Max: 12
-  Exclude:
-    - 'lib/kreuzberg/config.rb'
-
-Metrics/ClassLength:
-  Exclude:
-    - 'lib/kreuzberg/config.rb'
-
-RSpec/RepeatedExampleGroupBody:
-  Enabled: false
-
-RSpec/DescribeClass:
-  Enabled: false
-
-RSpec/PredicateMatcher:
-  Enabled: false
-
-Gemspec/AddRuntimeDependency:
-  Enabled: true
-
-Gemspec/AttributeAssignment:
-  Enabled: true
-
-Gemspec/DeprecatedAttributeAssignment:
-  Enabled: true
-
-Gemspec/RequireMFA:
-  Enabled: false
-
-Layout/EmptyLinesAfterModuleInclusion:
-  Enabled: true
-
-Layout/LineContinuationLeadingSpace:
-  Enabled: true
-
-Layout/LineContinuationSpacing:
-  Enabled: true
-
-Layout/LineEndStringConcatenationIndentation:
-  Enabled: true
-
-Layout/SpaceBeforeBrackets:
-  Enabled: true
-
-Lint/AmbiguousAssignment:
-  Enabled: true
-
-Lint/AmbiguousOperatorPrecedence:
-  Enabled: true
-
-Lint/AmbiguousRange:
-  Enabled: true
-
-Lint/ArrayLiteralInRegexp:
-  Enabled: true
-
-Lint/ConstantOverwrittenInRescue:
-  Enabled: true
-
-Lint/ConstantReassignment:
-  Enabled: true
-
-Lint/CopDirectiveSyntax:
-  Enabled: true
-
-Lint/DeprecatedConstants:
-  Enabled: true
-
-Lint/DuplicateBranch:
-  Enabled: true
-
-Lint/DuplicateMagicComment:
-  Enabled: true
-
-Lint/DuplicateMatchPattern:
-  Enabled: true
-
-Lint/DuplicateRegexpCharacterClassElement:
-  Enabled: true
-
-Lint/DuplicateSetElement:
-  Enabled: true
-
-Lint/EmptyBlock:
-  Enabled: true
-
-Lint/EmptyClass:
-  Enabled: true
-
-Lint/EmptyInPattern:
-  Enabled: true
-
-Lint/HashNewWithKeywordArgumentsAsDefault:
-  Enabled: true
-
-Lint/IncompatibleIoSelectWithFiberScheduler:
-  Enabled: true
-
-Lint/ItWithoutArgumentsInBlock:
-  Enabled: true
-
-Lint/LambdaWithoutLiteralBlock:
-  Enabled: true
-
-Lint/LiteralAssignmentInCondition:
-  Enabled: true
-
-Lint/MixedCaseRange:
-  Enabled: true
-
-Lint/NoReturnInBeginEndBlocks:
-  Enabled: true
-
-Lint/NonAtomicFileOperation:
-  Enabled: true
-
-Lint/NumberedParameterAssignment:
-  Enabled: true
-
-Lint/NumericOperationWithConstantResult:
-  Enabled: true
-
-Lint/OrAssignmentToConstant:
-  Enabled: true
-
-Lint/RedundantDirGlobSort:
-  Enabled: true
-
-Lint/RedundantRegexpQuantifiers:
-  Enabled: true
-
-Lint/RedundantTypeConversion:
-  Enabled: true
-
-Lint/RefinementImportMethods:
-  Enabled: true
-
-Lint/RequireRangeParentheses:
-  Enabled: true
-
-Lint/RequireRelativeSelfPath:
-  Enabled: true
-
-Lint/SharedMutableDefault:
-  Enabled: true
-
-Lint/SuppressedExceptionInNumberConversion:
-  Enabled: true
-
-Lint/SymbolConversion:
-  Enabled: true
-
-Lint/ToEnumArguments:
-  Enabled: true
-
-Lint/TripleQuotes:
-  Enabled: true
-
-Lint/UnescapedBracketInRegexp:
-  Enabled: true
-
-Lint/UnexpectedBlockArity:
-  Enabled: true
-
-Lint/UnmodifiedReduceAccumulator:
-  Enabled: true
-
-Lint/UselessConstantScoping:
-  Enabled: true
-
-Lint/UselessDefaultValueArgument:
-  Enabled: true
-
-Lint/UselessDefined:
-  Enabled: true
-
-Lint/UselessNumericOperation:
-  Enabled: true
-
-Lint/UselessOr:
-  Enabled: true
-
-Lint/UselessRescue:
-  Enabled: true
-
-Lint/UselessRuby2Keywords:
-  Enabled: true
-
-Metrics/CollectionLiteralLength:
-  Enabled: true
-
-Naming/BlockForwarding:
-  Enabled: true
-
-Naming/PredicateMethod:
-  Enabled: true
-
-Security/CompoundHash:
-  Enabled: true
-
-Security/IoMethods:
-  Enabled: true
-
-Style/AmbiguousEndlessMethodDefinition:
-  Enabled: true
-
-Style/ArgumentsForwarding:
-  Enabled: true
-
-Style/ArrayIntersect:
-  Enabled: true
-
-Style/ArrayIntersectWithSingleElement:
-  Enabled: true
-
-Style/BitwisePredicate:
-  Enabled: true
-
-Style/CollectionCompact:
-  Enabled: true
-
-Style/CollectionQuerying:
-  Enabled: true
-
-Style/CombinableDefined:
-  Enabled: true
-
-Style/ComparableBetween:
-  Enabled: true
-
-Style/ComparableClamp:
-  Enabled: true
-
-Style/ConcatArrayLiterals:
-  Enabled: true
-
-Style/DataInheritance:
-  Enabled: true
-
-Style/DigChain:
-  Enabled: true
-
-Style/DirEmpty:
-  Enabled: true
-
-Style/DocumentDynamicEvalDefinition:
-  Enabled: true
-
-Style/EmptyHeredoc:
-  Enabled: true
-
-Style/EmptyStringInsideInterpolation:
-  Enabled: true
-
-Style/EndlessMethod:
-  Enabled: true
-
-Style/EnvHome:
-  Enabled: true
-
-Style/ExactRegexpMatch:
-  Enabled: true
-
-Style/FetchEnvVar:
-  Enabled: true
-
-Style/FileEmpty:
-  Enabled: true
-
-Style/FileNull:
-  Enabled: true
-
-Style/FileRead:
-  Enabled: true
-
-Style/FileTouch:
-  Enabled: true
-
-Style/FileWrite:
-  Enabled: true
-
-Style/HashConversion:
-  Enabled: true
-
-Style/HashExcept:
-  Enabled: true
-
-Style/HashFetchChain:
-  Enabled: true
-
-Style/HashSlice:
-  Enabled: true
-
-Style/IfWithBooleanLiteralBranches:
-  Enabled: true
-
-Style/InPatternThen:
-  Enabled: true
-
-Style/ItAssignment:
-  Enabled: true
-
-Style/ItBlockParameter:
-  Enabled: true
-
-Style/KeywordArgumentsMerging:
-  Enabled: true
-
-Style/MagicCommentFormat:
-  Enabled: true
-
-Style/MapCompactWithConditionalBlock:
-  Enabled: true
-
-Style/MapIntoArray:
-  Enabled: true
-
-Style/MapToHash:
-  Enabled: true
-
-Style/MapToSet:
-  Enabled: true
-
-Style/MinMaxComparison:
-  Enabled: true
-
-Style/MultilineInPatternThen:
-  Enabled: true
-
-Style/NegatedIfElseCondition:
-  Enabled: true
-
-Style/NestedFileDirname:
-  Enabled: true
-
-Style/NilLambda:
-  Enabled: true
-
-Style/NumberedParameters:
-  Enabled: true
-
-Style/NumberedParametersLimit:
-  Enabled: true
-
-Style/ObjectThen:
-  Enabled: true
-
-Style/OpenStructUse:
-  Enabled: true
-
-Style/OperatorMethodCall:
-  Enabled: true
-
-Style/QuotedSymbols:
-  Enabled: true
-
-Style/RedundantArgument:
-  Enabled: true
-
-Style/RedundantArrayConstructor:
-  Enabled: true
-
-Style/RedundantArrayFlatten:
-  Enabled: true
-
-Style/RedundantConstantBase:
-  Enabled: true
-
-Style/RedundantCurrentDirectoryInPath:
-  Enabled: true
-
-Style/RedundantDoubleSplatHashBraces:
-  Enabled: true
-
-Style/RedundantEach:
-  Enabled: true
-
-Style/RedundantFilterChain:
-  Enabled: true
-
-Style/RedundantFormat:
-  Enabled: true
-
-Style/RedundantHeredocDelimiterQuotes:
-  Enabled: true
-
-Style/RedundantInitialize:
-  Enabled: true
-
-Style/RedundantInterpolationUnfreeze:
-  Enabled: true
-
-Style/RedundantLineContinuation:
-  Enabled: true
-
-Style/RedundantRegexpArgument:
-  Enabled: true
-
-Style/RedundantRegexpConstructor:
-  Enabled: true
-
-Style/RedundantSelfAssignmentBranch:
-  Enabled: true
-
-Style/RedundantStringEscape:
-  Enabled: true
-
-Style/ReturnNilInPredicateMethodDefinition:
-  Enabled: true
-
-Style/SafeNavigationChainLength:
-  Enabled: true
-
-Style/SelectByRegexp:
-  Enabled: true
-
-Style/SendWithLiteralMethodName:
-  Enabled: true
-
-Style/SingleLineDoEndBlock:
-  Enabled: true
-
-Style/StringChars:
-  Enabled: true
-
-Style/SuperArguments:
-  Enabled: true
-
-Style/SuperWithArgsParentheses:
-  Enabled: true
-
-Style/SwapValues:
-  Enabled: true
-
-Style/YAMLFileRead:
-  Enabled: true
-
-RSpec/ExampleLength:
-  Max: 50
-
-RSpec/MultipleExpectations:
-  Max: 25
-
-RSpec/NestedGroups:
-  Max: 6

data/Gemfile

@@ -1,8 +0,0 @@
-# frozen_string_literal: true
-
-source 'https://rubygems.org'
-
-gemspec
-
-gem 'pry', '~> 0.14', group: :development
-gem 'pry-byebug', '~> 3.10', group: :development, platforms: %i[ruby mri]