kreuzberg 4.5.4-aarch64-linux → 4.6.1-aarch64-linux

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: dfd3586f04b1b2b6dcdf06af30712755fd3fd8d76c7bc03cee05a67133134998
-  data.tar.gz: c402149a2765063da19d639eb2aa2b75340db6292b680c9218bc826d5997346c
+  metadata.gz: 7d632985bd042b73c0e51c84e292d9626267f205992bf49c9c62bcef78d9656f
+  data.tar.gz: ded364d7414b8fb4eafbda15af492698338799eb86749f9708f3f595ab7cad6d
 SHA512:
-  metadata.gz: aa8f6ba036c180200182377b90b15f293cf008b440c49d732d5dfa84df203373b173b9928138d58917fd7dddfe41c3b4ea89cab582c69707253ed87163932349
-  data.tar.gz: 211e120a58693b5c385ff0c577391d542a4ee9ca3e1194ca96e2e975a2636d78cdf9526b25852dad70d543142576317c2b344260f83148143d4d9ab6ac6822bd
+  metadata.gz: 5714ba27e6e41a99db90b829cd0c02986eab8c522673f62635ec0c5457d3d7367d290ef1146486fcccdda44fdb55751a99b24758454fdd58c90d93b83b77656e
+  data.tar.gz: ff4e355557401115eb34c227ddc4cd573e6b877f830b5b1a13b749508d202cd691866122e2be4132a883e7c80a4e7bae9f388296c79df7890adbd9aabcb822ca

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=v4.5.4" alt="Go">
+    <img src="https://img.shields.io/github/v/tag/kreuzberg-dev/kreuzberg?label=Go&color=007ec6&filter=v4.6.1" 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#">

data/lib/kreuzberg/config.rb

@@ -926,14 +926,14 @@ module Kreuzberg
     #   )
     #
     class Extraction
-      attr_reader :use_cache, :enable_quality_processing, :force_ocr,
+      attr_reader :use_cache, :enable_quality_processing, :force_ocr, :force_ocr_pages,
                   :include_document_structure,
                   :ocr, :chunking, :language_detection, :pdf_options,
                   :images, :postprocessor,
                   :token_reduction, :keywords, :html_options, :pages,
                   :max_concurrent_extractions, :output_format, :result_format,
                   :security_limits, :layout, :concurrency,
-                  :cache_namespace, :cache_ttl_secs
+                  :cache_namespace, :cache_ttl_secs, :extraction_timeout_secs
 
       # Alias for backward compatibility - image_extraction is the canonical name
       alias image_extraction images
@@ -954,11 +954,11 @@ module Kreuzberg
       #
       # Keys that are allowed in the Extraction config
       ALLOWED_KEYS = %i[
-        use_cache enable_quality_processing force_ocr include_document_structure ocr chunking
+        use_cache enable_quality_processing force_ocr force_ocr_pages include_document_structure ocr chunking
         language_detection pdf_options image_extraction
         postprocessor token_reduction keywords html_options pages
         max_concurrent_extractions output_format result_format
-        security_limits layout concurrency cache_namespace cache_ttl_secs
+        security_limits layout concurrency cache_namespace cache_ttl_secs extraction_timeout_secs
       ].freeze
 
       # Aliases for backward compatibility
@@ -1019,6 +1019,7 @@ module Kreuzberg
                      use_cache: true,
                      enable_quality_processing: true,
                      force_ocr: false,
+                     force_ocr_pages: nil,
                      include_document_structure: false,
                      ocr: nil,
                      chunking: nil,
@@ -1037,10 +1038,12 @@ module Kreuzberg
                      layout: nil,
                      concurrency: nil,
                      cache_namespace: nil,
-                     cache_ttl_secs: nil)
+                     cache_ttl_secs: nil,
+                     extraction_timeout_secs: nil)
         kwargs = {
           use_cache: use_cache, enable_quality_processing: enable_quality_processing,
-          force_ocr: force_ocr, include_document_structure: include_document_structure,
+          force_ocr: force_ocr, force_ocr_pages: force_ocr_pages,
+          include_document_structure: include_document_structure,
           ocr: ocr, chunking: chunking, language_detection: language_detection,
           pdf_options: pdf_options, image_extraction: image_extraction,
           postprocessor: postprocessor,
@@ -1050,7 +1053,8 @@ module Kreuzberg
           security_limits: security_limits, layout: layout,
           concurrency: concurrency,
           cache_namespace: cache_namespace,
-          cache_ttl_secs: cache_ttl_secs
+          cache_ttl_secs: cache_ttl_secs,
+          extraction_timeout_secs: extraction_timeout_secs
         }
         extracted = extract_from_hash(hash, kwargs)
 
@@ -1068,6 +1072,7 @@ module Kreuzberg
         @use_cache = params[:use_cache] ? true : false
         @enable_quality_processing = params[:enable_quality_processing] ? true : false
         @force_ocr = params[:force_ocr] ? true : false
+        @force_ocr_pages = params[:force_ocr_pages]
         @include_document_structure = params[:include_document_structure] ? true : false
         @ocr = normalize_config(params[:ocr], OCR)
         @chunking = normalize_config(params[:chunking], Chunking)
@@ -1086,6 +1091,7 @@ module Kreuzberg
         @result_format = validate_result_format(params[:result_format])
         @cache_namespace = params[:cache_namespace]
         @cache_ttl_secs = params[:cache_ttl_secs]&.to_i
+        @extraction_timeout_secs = params[:extraction_timeout_secs]&.to_i
         @security_limits = params[:security_limits]
       end
 
@@ -1118,12 +1124,14 @@ module Kreuzberg
           use_cache: @use_cache,
           enable_quality_processing: @enable_quality_processing,
           force_ocr: @force_ocr,
+          force_ocr_pages: @force_ocr_pages,
           include_document_structure: @include_document_structure,
           max_concurrent_extractions: @max_concurrent_extractions,
           output_format: @output_format,
           result_format: @result_format,
           cache_namespace: @cache_namespace,
-          cache_ttl_secs: @cache_ttl_secs
+          cache_ttl_secs: @cache_ttl_secs,
+          extraction_timeout_secs: @extraction_timeout_secs
         }
       end
 
@@ -1250,6 +1258,8 @@ module Kreuzberg
           @enable_quality_processing = value ? true : false
         when :force_ocr
           @force_ocr = value ? true : false
+        when :force_ocr_pages
+          @force_ocr_pages = value
         when :include_document_structure
           @include_document_structure = value ? true : false
         when :ocr
@@ -1286,6 +1296,8 @@ module Kreuzberg
           @cache_namespace = value
         when :cache_ttl_secs
           @cache_ttl_secs = value&.to_i
+        when :extraction_timeout_secs
+          @extraction_timeout_secs = value&.to_i
         else
           raise ArgumentError, "Unknown configuration key: #{key}"
         end
@@ -1345,6 +1357,7 @@ module Kreuzberg
         @use_cache = merged.use_cache
         @enable_quality_processing = merged.enable_quality_processing
         @force_ocr = merged.force_ocr
+        @force_ocr_pages = merged.force_ocr_pages
         @include_document_structure = merged.include_document_structure
         @ocr = merged.ocr
         @chunking = merged.chunking
@@ -1369,6 +1382,7 @@ module Kreuzberg
         @result_format = merged.result_format
         @cache_namespace = merged.cache_namespace
         @cache_ttl_secs = merged.cache_ttl_secs
+        @extraction_timeout_secs = merged.extraction_timeout_secs
       end
     end
   end

data/lib/kreuzberg/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kreuzberg
-  VERSION = '4.5.4'
+  VERSION = '4.6.1'
 end

data/sig/kreuzberg.rbs

@@ -481,7 +481,9 @@ module Kreuzberg
       attr_reader enable_quality_processing: bool
       attr_reader cache_namespace: String?
       attr_reader cache_ttl_secs: Integer?
+      attr_reader extraction_timeout_secs: Integer?
       attr_reader force_ocr: bool
+      attr_reader force_ocr_pages: Array[Integer]?
       attr_reader include_document_structure: bool
       attr_reader ocr: OCR?
       attr_reader chunking: Chunking?
@@ -508,6 +510,7 @@ module Kreuzberg
         ?use_cache: bool,
         ?enable_quality_processing: bool,
         ?force_ocr: bool,
+        ?force_ocr_pages: Array[Integer]?,
         ?include_document_structure: bool,
         ?ocr: (OCR | Hash[Symbol, untyped])?,
         ?chunking: (Chunking | Hash[Symbol, untyped])?,
@@ -525,7 +528,8 @@ module Kreuzberg
         ?output_format: String?,
         ?result_format: String?,
         ?cache_namespace: String?,
-        ?cache_ttl_secs: Integer?
+        ?cache_ttl_secs: Integer?,
+        ?extraction_timeout_secs: Integer?
       ) -> void
       def to_h: () -> Hash[Symbol, untyped]
       def to_json: (*untyped) -> String

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: kreuzberg
 version: !ruby/object:Gem::Version
-  version: 4.5.4
+  version: 4.6.1
 platform: aarch64-linux
 authors:
 - Na'aman Hirschfeld
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-03-23 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