kreuzberg 4.8.4-aarch64-linux → 4.9.0-aarch64-linux

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: b72926d04c538087ca728eaaede75af8f20c630e1b77109a78345434081393ef
-  data.tar.gz: 6ca5e1e5acf2084b88ef624792ec64a43f026e68abbd04b6ba49d9443e407ea0
+  metadata.gz: c7a0c15c0ef0421fe0b2fe83a5feba779f2aa18a1be817b5c7b2f52323601531
+  data.tar.gz: 8553c7bcee1466b95ead2688b219d8953ddd941590e55a589f9c7b98317a45a5
 SHA512:
-  metadata.gz: fc3ee80cf5129eed83000213a057b785cedf04627b9f8761bba70d5971da18a5e1868410296982522aa0e431dcdcce1d8528f5d1c615c0bef845ab4af7dfbab6
-  data.tar.gz: '06538a19fcfbccedd23548f49f42c5fbf1f08e256f579a8b0b8b5c1bc9b2411a489fdb7e653424ca3ced2e6de7008a3701c5ac64047ca0a78d8b7821478bcaf6'
+  metadata.gz: 5b3e598e8cdecc91dbe6d6e68772f9fc3ceac7688d87d50b70e5120fb8957dafd4abe47dec6961b3d2046741b3a1bf0d8d618e0f8a36c255e687bbb290d2bd20
+  data.tar.gz: 8e82431165e64e89141d3077d2068a8435a08daf9d4708499121936493995dd874ba561a6c4b791f14be746bb91bd4d15ef444d6163a754ab5b555bc87c1f2eb

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=v4.8.4" alt="Go">
+    <img src="https://img.shields.io/github/v/tag/kreuzberg-dev/kreuzberg?label=Go&color=007ec6&filter=v4.9.0" 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#">
@@ -39,10 +39,13 @@
   <a href="https://github.com/kreuzberg-dev/kreuzberg/pkgs/container/kreuzberg">
     <img src="https://img.shields.io/badge/Docker-007ec6?logo=docker&logoColor=white" alt="Docker">
   </a>
+  <a href="https://artifacthub.io/packages/search?repo=kreuzberg">
+    <img src="https://img.shields.io/endpoint?url=https://artifacthub.io/badge/repository/kreuzberg" alt="Artifact Hub">
+  </a>
 
   <!-- Project Info -->
   <a href="https://github.com/kreuzberg-dev/kreuzberg/blob/main/LICENSE">
-    <img src="https://img.shields.io/badge/License-Elastic--2.0-blue.svg" alt="License">
+    <img src="https://img.shields.io/badge/License-MIT-007ec6" alt="License">
   </a>
   <a href="https://docs.kreuzberg.dev">
     <img src="https://img.shields.io/badge/docs-kreuzberg.dev-007ec6" alt="Documentation">
@@ -419,7 +422,7 @@ Contributions are welcome! See [Contributing Guide](https://github.com/kreuzberg
 
 ## License
 
-Elastic License 2.0 (ELv2) - see [LICENSE](../../LICENSE) for details.
+MIT License - see LICENSE file for details.
 
 ## Support
 

data/lib/kreuzberg/errors.rb

@@ -116,5 +116,8 @@ module Kreuzberg
 
     # Raised when embedding fails
     class EmbeddingError < Error; end
+
+    # Raised when an extraction is cancelled via a cancellation token
+    class CancelledError < Error; end
   end
 end

data/lib/kreuzberg/result.rb

@@ -15,7 +15,7 @@ module Kreuzberg
     attr_reader :content, :mime_type, :metadata, :metadata_json, :tables,
                 :detected_languages, :chunks, :images, :pages, :elements, :ocr_elements, :djot_content,
                 :document, :extracted_keywords, :quality_score, :processing_warnings, :annotations,
-                :uris, :children
+                :uris, :children, :structured_output
 
     # @!attribute [r] cells
     #   @return [Array<Array<String>>] Table cells (2D array)
@@ -145,7 +145,7 @@ module Kreuzberg
     #   @return [Array<Image>] Images on this page
     # @!attribute [r] hierarchy
     #   @return [PageHierarchy, nil] Hierarchy information for the page
-    PageContent = Struct.new(:page_number, :content, :tables, :images, :hierarchy, :is_blank) do
+    PageContent = Struct.new(:page_number, :content, :tables, :images, :hierarchy, :is_blank, :layout_regions) do
       def to_h
         {
           page_number: page_number,
@@ -153,7 +153,27 @@ module Kreuzberg
           tables: tables.map(&:to_h),
           images: images.map(&:to_h),
           hierarchy: hierarchy&.to_h,
-          is_blank: is_blank
+          is_blank: is_blank,
+          layout_regions: layout_regions&.map(&:to_h)
+        }
+      end
+    end
+
+    # @!attribute [r] class_name
+    #   @return [String] Layout class name (e.g. "picture", "table", "text")
+    # @!attribute [r] confidence
+    #   @return [Float] Detection confidence score (0.0 to 1.0)
+    # @!attribute [r] bounding_box
+    #   @return [ElementBoundingBox] Bounding box in document coordinate space
+    # @!attribute [r] area_fraction
+    #   @return [Float] Fraction of page area covered (0.0 to 1.0)
+    LayoutRegion = Struct.new(:class_name, :confidence, :bounding_box, :area_fraction) do
+      def to_h
+        {
+          class: class_name,
+          confidence: confidence,
+          bounding_box: bounding_box&.to_h,
+          area_fraction: area_fraction
         }
       end
     end
@@ -342,6 +362,7 @@ module Kreuzberg
       @annotations = parse_annotations(get_value(hash, 'annotations'))
       @uris = parse_uris(get_value(hash, 'uris'))
       @children = parse_children(get_value(hash, 'children'))
+      @structured_output = get_value(hash, 'structured_output')
     end
     # rubocop:enable Metrics/AbcSize, Metrics/MethodLength
 
@@ -369,7 +390,8 @@ module Kreuzberg
         processing_warnings: @processing_warnings.map(&:to_h),
         annotations: @annotations&.map(&:to_h),
         uris: @uris&.map(&:to_h),
-        children: @children&.map(&:to_h)
+        children: @children&.map(&:to_h),
+        structured_output: @structured_output
       }
     end
     # rubocop:enable Metrics/CyclomaticComplexity, Metrics/MethodLength
@@ -568,11 +590,36 @@ module Kreuzberg
           tables: parse_tables(page_hash['tables']),
           images: parse_images(page_hash['images']),
           hierarchy: parse_page_hierarchy(page_hash['hierarchy']),
-          is_blank: page_hash['is_blank']
+          is_blank: page_hash['is_blank'],
+          layout_regions: parse_layout_regions(page_hash['layout_regions'])
         )
       end
     end
 
+    def parse_layout_regions(regions_data)
+      return nil if regions_data.nil?
+
+      regions_data.map do |region_hash|
+        LayoutRegion.new(
+          class_name: region_hash['class'],
+          confidence: region_hash['confidence']&.to_f,
+          bounding_box: parse_element_bounding_box(region_hash['bounding_box']),
+          area_fraction: region_hash['area_fraction']&.to_f
+        )
+      end
+    end
+
+    def parse_element_bounding_box(bounding_box_data)
+      return nil if bounding_box_data.nil?
+
+      ElementBoundingBox.new(
+        x0: bounding_box_data['x0'].to_f,
+        y0: bounding_box_data['y0'].to_f,
+        x1: bounding_box_data['x1'].to_f,
+        y1: bounding_box_data['y1'].to_f
+      )
+    end
+
     def parse_page_hierarchy(hierarchy_data)
       return nil if hierarchy_data.nil?
 

data/lib/kreuzberg/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kreuzberg
-  VERSION = '4.8.4'
+  VERSION = '4.9.0'
 end

data/sig/kreuzberg.rbs

@@ -18,7 +18,8 @@ module Kreuzberg
   type element_type = 'title' | 'narrative_text' | 'heading' | 'list_item' | 'table' | 'image' | 'page_break' | 'code_block' | 'block_quote' | 'footer' | 'header'
 
   # Bounding box coordinates for element positioning (T::Struct from types.rb)
-  class BoundingBox     attr_reader x0: Float
+  class BoundingBox
+    attr_reader x0: Float
     attr_reader y0: Float
     attr_reader x1: Float
     attr_reader y1: Float
@@ -28,7 +29,8 @@ module Kreuzberg
   end
 
   # Metadata for a semantic element (T::Struct from types.rb)
-  class ElementMetadata     attr_reader page_number: Integer?
+  class ElementMetadata
+    attr_reader page_number: Integer?
     attr_reader filename: String?
     attr_reader coordinates: BoundingBox?
     attr_reader element_index: Integer?
@@ -39,7 +41,8 @@ module Kreuzberg
   end
 
   # Semantic element extracted from document (T::Struct from types.rb)
-  class Element     attr_reader element_id: String
+  class Element
+    attr_reader element_id: String
     attr_reader element_type: String
     attr_reader text: String
     attr_reader metadata: ElementMetadata
@@ -49,7 +52,8 @@ module Kreuzberg
   end
 
   # Header/Heading metadata (T::Struct from types.rb)
-  class HeaderMetadata     attr_reader level: Integer
+  class HeaderMetadata
+    attr_reader level: Integer
     attr_reader text: String
     attr_reader id: String?
     attr_reader depth: Integer
@@ -60,7 +64,8 @@ module Kreuzberg
   end
 
   # Link metadata (T::Struct from types.rb)
-  class LinkMetadata     attr_reader href: String
+  class LinkMetadata
+    attr_reader href: String
     attr_reader text: String
     attr_reader title: String?
     attr_reader link_type: String
@@ -72,7 +77,8 @@ module Kreuzberg
   end
 
   # Image metadata (T::Struct from types.rb)
-  class ImageMetadata     attr_reader src: String
+  class ImageMetadata
+    attr_reader src: String
     attr_reader alt: String?
     attr_reader title: String?
     attr_reader dimensions: Array[Integer]?
@@ -84,7 +90,8 @@ module Kreuzberg
   end
 
   # Structured data metadata (T::Struct from types.rb)
-  class StructuredData     attr_reader data_type: String
+  class StructuredData
+    attr_reader data_type: String
     attr_reader raw_json: String
     attr_reader schema_type: String?
 
@@ -210,7 +217,8 @@ module Kreuzberg
   end
 
   # HTML metadata (T::Struct from types.rb)
-  class HtmlMetadata     attr_reader title: String?
+  class HtmlMetadata
+    attr_reader title: String?
     attr_reader description: String?
     attr_reader author: String?
     attr_reader copyright: String?
@@ -261,7 +269,8 @@ module Kreuzberg
   end
 
   # Extracted keyword with relevance metadata (T::Struct from types.rb)
-  class ExtractedKeyword     attr_reader text: String
+  class ExtractedKeyword
+    attr_reader text: String
     attr_reader score: Float
     attr_reader algorithm: String
     attr_reader positions: Array[Integer]?
@@ -271,15 +280,39 @@ module Kreuzberg
   end
 
   # Processing warning from a pipeline stage (T::Struct from types.rb)
-  class ProcessingWarning     attr_reader source: String
+  class ProcessingWarning
+    attr_reader source: String
     attr_reader message: String
 
     def initialize: (source: String, message: String) -> void
     def serialize: () -> Hash[Symbol, untyped]
   end
 
+  # LLM token usage from an LLM-assisted extraction step (T::Struct from types.rb)
+  class LlmUsage
+    attr_reader model: String
+    attr_reader source: String
+    attr_reader input_tokens: Integer?
+    attr_reader output_tokens: Integer?
+    attr_reader total_tokens: Integer?
+    attr_reader estimated_cost: Float?
+    attr_reader finish_reason: String?
+
+    def initialize: (
+      model: String,
+      source: String,
+      ?input_tokens: Integer?,
+      ?output_tokens: Integer?,
+      ?total_tokens: Integer?,
+      ?estimated_cost: Float?,
+      ?finish_reason: String?
+    ) -> void
+    def serialize: () -> Hash[Symbol, untyped]
+  end
+
   # Bounding box for document node positioning (T::Struct from types.rb)
-  class DocumentBoundingBox     attr_reader x0: Float
+  class DocumentBoundingBox
+    attr_reader x0: Float
     attr_reader y0: Float
     attr_reader x1: Float
     attr_reader y1: Float
@@ -289,7 +322,8 @@ module Kreuzberg
   end
 
   # Annotation for a document node (T::Struct from types.rb)
-  class DocumentAnnotation     attr_reader key: String
+  class DocumentAnnotation
+    attr_reader key: String
     attr_reader value: String
 
     def initialize: (key: String, value: String) -> void
@@ -297,7 +331,8 @@ module Kreuzberg
   end
 
   # Single node in the document structure tree (T::Struct from types.rb)
-  class DocumentNode     attr_reader id: String
+  class DocumentNode
+    attr_reader id: String
     attr_reader content: String
     attr_reader parent: Integer?
     attr_reader children: Array[Integer]
@@ -322,7 +357,8 @@ module Kreuzberg
   end
 
   # Structured document representation (T::Struct from types.rb)
-  class DocumentStructure     attr_reader nodes: Array[DocumentNode]
+  class DocumentStructure
+    attr_reader nodes: Array[DocumentNode]
 
     def initialize: (nodes: Array[DocumentNode]) -> void
     def serialize: () -> Hash[Symbol, untyped]
@@ -560,13 +596,15 @@ module Kreuzberg
       attr_reader batch_size: Integer?
       attr_reader show_download_progress: bool?
       attr_reader cache_dir: String?
+      attr_reader acceleration: Acceleration?
 
       def initialize: (
         ?model: Hash[Symbol, untyped],
         ?normalize: bool?,
         ?batch_size: Integer?,
         ?show_download_progress: bool?,
-        ?cache_dir: String?
+        ?cache_dir: String?,
+        ?acceleration: (Acceleration | Hash[Symbol, untyped])?
       ) -> void
       def to_h: () -> Hash[Symbol, untyped]
     end
@@ -763,8 +801,9 @@ module Kreuzberg
       attr_reader confidence_threshold: Float?
       attr_reader apply_heuristics: bool
       attr_reader table_model: String?
+      attr_reader acceleration: Acceleration?
 
-      def initialize: (?preset: String, ?confidence_threshold: Float?, ?apply_heuristics: bool, ?table_model: String?) -> void
+      def initialize: (?preset: String, ?confidence_threshold: Float?, ?apply_heuristics: bool, ?table_model: String?, ?acceleration: (Acceleration | Hash[Symbol, untyped])?) -> void
       def to_h: () -> Hash[Symbol, untyped]
     end
 
@@ -927,7 +966,10 @@ module Kreuzberg
     extracted_keywords: Array[extracted_keyword_hash]?,
     quality_score: Float?,
     processing_warnings: Array[processing_warning_hash]?,
-    annotations: Array[pdf_annotation_hash]?
+    annotations: Array[pdf_annotation_hash]?,
+    uris: Array[uri_hash]?,
+    children: Array[archive_entry_hash]?,
+    llm_usage: Array[llm_usage_hash]?
   }
 
   type extracted_keyword_hash = {
@@ -942,12 +984,36 @@ module Kreuzberg
     message: String
   }
 
+  type llm_usage_hash = {
+    model: String,
+    source: String,
+    input_tokens: Integer?,
+    output_tokens: Integer?,
+    total_tokens: Integer?,
+    estimated_cost: Float?,
+    finish_reason: String?
+  }
+
+  type uri_hash = {
+    url: String,
+    label: String?,
+    page: Integer?,
+    kind: String
+  }
+
+  type archive_entry_hash = {
+    path: String,
+    mime_type: String,
+    result: extraction_result_hash?
+  }
+
   type page_content_hash = {
     page_number: Integer,
     content: String,
     tables: Array[table_hash],
     images: Array[image_hash],
-    is_blank: bool?
+    is_blank: bool?,
+    layout_regions: Array[untyped]?
   }
 
   type djot_content_hash = {
@@ -1218,6 +1284,17 @@ module Kreuzberg
       def to_h: () -> image_hash
     end
 
+    # Layout detection region on a page (Struct from result.rb)
+    class LayoutRegion
+      attr_reader class_name: String
+      attr_reader confidence: Float
+      attr_reader bounding_box: ElementBoundingBox?
+      attr_reader area_fraction: Float
+
+      def initialize: (class_name: String, confidence: Float, bounding_box: ElementBoundingBox?, area_fraction: Float) -> void
+      def to_h: () -> Hash[Symbol, untyped]
+    end
+
     # Page content with text and extracted elements (Struct from result.rb)
     class PageContent
       attr_reader page_number: Integer
@@ -1226,8 +1303,9 @@ module Kreuzberg
       attr_reader images: Array[Image]?
       attr_reader hierarchy: PageHierarchy?
       attr_reader is_blank: bool?
+      attr_reader layout_regions: Array[LayoutRegion]?
 
-      def initialize: (page_number: Integer, content: String, tables: Array[Table], images: Array[Image]?, hierarchy: PageHierarchy?, is_blank: bool?) -> void
+      def initialize: (page_number: Integer, content: String, tables: Array[Table], images: Array[Image]?, hierarchy: PageHierarchy?, is_blank: bool?, layout_regions: Array[LayoutRegion]?) -> void
       def to_h: () -> Hash[Symbol, untyped]
     end
 
@@ -1483,6 +1561,10 @@ module Kreuzberg
     attr_reader quality_score: Float?
     attr_reader processing_warnings: Array[ProcessingWarning]?
     attr_reader annotations: Array[PdfAnnotation]?
+    attr_reader uris: Array[uri_hash]?
+    attr_reader children: Array[archive_entry_hash]?
+    attr_reader llm_usage: Array[LlmUsage]?
+    attr_reader structured_output: (Hash[String, untyped] | Array[untyped] | Integer | Float | String | bool | nil)
 
     # PDF annotation extracted from a document page (Struct from result.rb)
     class PdfAnnotation
@@ -1516,11 +1598,18 @@ module Kreuzberg
     def parse_element: (Hash[String, untyped] element_hash) -> ElementStruct
     def parse_element_coordinates: (Hash[String, untyped]? coordinates_data) -> ElementBoundingBox?
     def parse_ocr_elements: (Array[ocr_element_hash]? ocr_elements_data) -> Array[OcrElement]?
+    def parse_layout_regions: (Array[untyped]? regions_data) -> Array[LayoutRegion]?
+    def parse_element_bounding_box: (Hash[String, untyped]? bounding_box_data) -> ElementBoundingBox?
     def parse_page_hierarchy: (Hash[String, untyped]? hierarchy_data) -> PageHierarchy?
     def parse_djot_content: (Hash[String, untyped]? djot_data) -> DjotContent?
     def parse_document_structure: (Hash[String, untyped]? document_data) -> DocumentStructure?
     def parse_extracted_keywords: (Array[extracted_keyword_hash]? keywords_data) -> Array[ExtractedKeyword]?
     def parse_processing_warnings: (Array[processing_warning_hash]? warnings_data) -> Array[ProcessingWarning]
+    def parse_uris: (Array[uri_hash]? uris_data) -> Array[uri_hash]?
+    def build_uri: (Hash[String, untyped] u_hash) -> uri_hash
+    def parse_children: (Array[untyped]? children_data) -> Array[archive_entry_hash]?
+    def build_archive_entry: (Hash[String, untyped] c_hash) -> archive_entry_hash
+    def parse_llm_usage: (Array[llm_usage_hash]? usage_data) -> Array[LlmUsage]?
     def get_value: (Hash[String | Symbol, untyped] hash, String key, ?untyped default) -> untyped
     def serialize_tables: () -> Array[table_hash]
     def serialize_chunks: () -> Array[chunk_hash]?
@@ -1742,6 +1831,9 @@ module Kreuzberg
 
     class EmbeddingError < Error
     end
+
+    class CancelledError < Error
+    end
   end
 
   # Internal modules (prepended to Kreuzberg singleton)

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: kreuzberg
 version: !ruby/object:Gem::Version
-  version: 4.8.4
+  version: 4.9.0
 platform: aarch64-linux
 authors:
 - Na'aman Hirschfeld
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-04-13 00:00:00.000000000 Z
+date: 2026-04-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -86,14 +86,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '3.0'
+        version: '4.0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '3.0'
+        version: '4.0'
 - !ruby/object:Gem::Dependency
   name: rubocop
   requirement: !ruby/object:Gem::Requirement
@@ -142,14 +142,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.8'
+        version: '2.0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.8'
+        version: '2.0'
 - !ruby/object:Gem::Dependency
   name: yard
   requirement: !ruby/object:Gem::Requirement