kreuzberg 4.4.6-x86_64-linux → 4.5.1-x86_64-linux

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8c3ecd15e895a368716d16c76ae0ab5ea8ae73161763fcbc79072024bbe7b74a
-  data.tar.gz: 0bc54f1d02426e0d2fdb716c97108ed89db7e8c44b6859ff8749e4cea55cc2d6
+  metadata.gz: a41906b6bf000dbec2fa20fd9080fca7ec759d514664aa4f3703054935888756
+  data.tar.gz: 22fb7ae544516195203c19bc133f90ae5a80d67b60a6654a17b3d8f622e6a6aa
 SHA512:
-  metadata.gz: 9f84210a644ffddb2b1438bb6f5d53f05b8b4ea9fa356708901a0c994614358c27a98fcbdab6f66abb26b762337c7c7895f6fc49ad0eaf4671b09acee0061690
-  data.tar.gz: 25c32ef228a645ee4e73be69072f7a7f5b8aea48b4df7f8bd561d325aaa6141f8001f778bf9fee6061a1069ef5e16b33ec34e10c5fdeec74fe71cb14a9a4903f
+  metadata.gz: 6380bb10531bcf895663281168a9c1dba1cd0c0185710de458e9c7b448063fcacae0fc613959f863f8eb43177965900a9d242bff6c745adcf6df6958393de769
+  data.tar.gz: b3d594c9009d729b57531b59726d6e7e3cdb064a63c9635d6642d96dcda8f842b4d970afb86931a8c096dcd9ecb1b4f89e8fdadefe292b2c6feb9deb72126f11

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    kreuzberg (4.4.6)
+    kreuzberg (4.5.1)
       rb_sys (~> 0.9.119)
       sorbet-runtime (~> 0.5)
 
@@ -49,7 +49,7 @@ GEM
     i18n (1.14.8)
       concurrent-ruby (~> 1.0)
     io-console (0.8.2)
-    json (2.19.1)
+    json (2.19.2)
     json-schema (6.2.0)
       addressable (~> 2.8)
       bigdecimal (>= 3.1, < 5)
@@ -60,7 +60,7 @@ GEM
       rb-fsevent (~> 0.10, >= 0.10.3)
       rb-inotify (~> 0.9, >= 0.9.10)
     logger (1.7.0)
-    mcp (0.8.0)
+    mcp (0.9.0)
       json-schema (>= 4.1)
     method_source (1.1.0)
     minitest (6.0.2)
@@ -134,7 +134,7 @@ GEM
       rubocop (~> 1.81)
     ruby-progressbar (1.13.0)
     securerandom (0.4.1)
-    sorbet-runtime (0.6.13023)
+    sorbet-runtime (0.6.13051)
     steep (1.10.0)
       activesupport (>= 5.1)
       concurrent-ruby (>= 1.1.10)
@@ -220,14 +220,14 @@ CHECKSUMS
   fileutils (1.8.0) sha256=8c6b1df54e2540bdb2f39258f08af78853aa70bad52b4d394bbc6424593c6e02
   i18n (1.14.8) sha256=285778639134865c5e0f6269e0b818256017e8cde89993fdfcbfb64d088824a5
   io-console (0.8.2) sha256=d6e3ae7a7cc7574f4b8893b4fca2162e57a825b223a177b7afa236c5ef9814cc
-  json (2.19.1) sha256=dd94fdc59e48bff85913829a32350b3148156bc4fd2a95a2568a78b11344082d
+  json (2.19.2) sha256=e7e1bd318b2c37c4ceee2444841c86539bc462e81f40d134cf97826cb14e83cf
   json-schema (6.2.0) sha256=e8bff46ed845a22c1ab2bd0d7eccf831c01fe23bb3920caa4c74db4306813666
-  kreuzberg (4.4.6)
+  kreuzberg (4.5.1)
   language_server-protocol (3.17.0.5) sha256=fd1e39a51a28bf3eec959379985a72e296e9f9acfce46f6a79d31ca8760803cc
   lint_roller (1.1.0) sha256=2c0c845b632a7d172cb849cc90c1bce937a28c5c8ccccb50dfd46a485003cc87
   listen (3.10.0) sha256=c6e182db62143aeccc2e1960033bebe7445309c7272061979bb098d03760c9d2
   logger (1.7.0) sha256=196edec7cc44b66cfb40f9755ce11b392f21f7967696af15d274dde7edff0203
-  mcp (0.8.0) sha256=ae8bd146bb8e168852866fd26f805f52744f6326afb3211e073f78a95e0c34fb
+  mcp (0.9.0) sha256=a0a3737b0ac9df0772f4ef7e2b013c260ddbcf217a5d50a66bff0baeddf03e47
   method_source (1.1.0) sha256=181301c9c45b731b4769bc81e8860e72f9161ad7d66dd99103c9ab84f560f5c5
   minitest (6.0.2) sha256=db6e57956f6ecc6134683b4c87467d6dd792323c7f0eea7b93f66bd284adbc3d
   mutex_m (0.3.0) sha256=cfcb04ac16b69c4813777022fdceda24e9f798e48092a2b817eb4c0a782b0751
@@ -259,7 +259,7 @@ CHECKSUMS
   rubocop-rspec (3.9.0) sha256=8fa70a3619408237d789aeecfb9beef40576acc855173e60939d63332fdb55e2
   ruby-progressbar (1.13.0) sha256=80fc9c47a9b640d6834e0dc7b3c94c9df37f08cb072b7761e4a71e22cff29b33
   securerandom (0.4.1) sha256=cc5193d414a4341b6e225f0cb4446aceca8e50d5e1888743fac16987638ea0b1
-  sorbet-runtime (0.6.13023) sha256=c00d11cc54951efbc0aece994dd6b20b1d1cb2a2606100c24d4ae7f840383073
+  sorbet-runtime (0.6.13051) sha256=ae5495bf229c5e3e5e3a2e17ac4853798d993437f5b5f12b763d462183852452
   steep (1.10.0) sha256=1b295b55f9aaff1b8d3ee42453ee55bc2a1078fda0268f288edb2dc014f4d7d1
   strscan (3.1.7) sha256=5f76462b94a3ea50b44973225b7d75b2cb96d4e1bee9ef1319b99ca117b72c8c
   terminal-table (4.0.0) sha256=f504793203f8251b2ea7c7068333053f0beeea26093ec9962e62ea79f94301d2

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=v4.4.6" alt="Go">
+    <img src="https://img.shields.io/github/v/tag/kreuzberg-dev/kreuzberg?label=Go&color=007ec6&filter=v4.5.1" 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#">
@@ -47,6 +47,9 @@
   <a href="https://docs.kreuzberg.dev">
     <img src="https://img.shields.io/badge/docs-kreuzberg.dev-blue" alt="Documentation">
   </a>
+  <a href="https://huggingface.co/Kreuzberg">
+    <img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Models-yellow" alt="Hugging Face">
+  </a>
 </div>
 
 <img width="1128" height="191" alt="Banner2" src="https://github.com/user-attachments/assets/419fc06c-8313-4324-b159-4b4d3cfce5c0" />

data/lib/kreuzberg/config.rb

@@ -392,7 +392,8 @@ module Kreuzberg
     #
     class PDF
       attr_reader :extract_images, :passwords, :extract_metadata, :font_config, :hierarchy,
-                  :extract_annotations, :top_margin_fraction, :bottom_margin_fraction
+                  :extract_annotations, :top_margin_fraction, :bottom_margin_fraction,
+                  :allow_single_column_tables
 
       def initialize(
         extract_images: false,
@@ -402,7 +403,8 @@ module Kreuzberg
         hierarchy: nil,
         extract_annotations: false,
         top_margin_fraction: nil,
-        bottom_margin_fraction: nil
+        bottom_margin_fraction: nil,
+        allow_single_column_tables: false
       )
         @extract_images = extract_images ? true : false
         @passwords = if passwords.is_a?(Array)
@@ -416,6 +418,7 @@ module Kreuzberg
         @extract_annotations = extract_annotations ? true : false
         @top_margin_fraction = top_margin_fraction&.to_f
         @bottom_margin_fraction = bottom_margin_fraction&.to_f
+        @allow_single_column_tables = allow_single_column_tables ? true : false
       end
 
       def to_h
@@ -427,7 +430,8 @@ module Kreuzberg
           hierarchy: @hierarchy&.to_h,
           extract_annotations: @extract_annotations,
           top_margin_fraction: @top_margin_fraction,
-          bottom_margin_fraction: @bottom_margin_fraction
+          bottom_margin_fraction: @bottom_margin_fraction,
+          allow_single_column_tables: @allow_single_column_tables
         }.compact
       end
 
@@ -803,6 +807,85 @@ module Kreuzberg
       end
     end
 
+    # Hardware acceleration configuration for ONNX Runtime
+    #
+    # Controls which execution provider (CPU, CoreML, CUDA, TensorRT) is used
+    # for inference in layout detection and embedding generation.
+    #
+    # @example Auto-select provider (CoreML on macOS, CUDA on Linux, CPU elsewhere)
+    #   acceleration = Acceleration.new
+    #
+    # @example Force CPU only
+    #   acceleration = Acceleration.new(provider: 'cpu')
+    #
+    # @example Use CUDA with specific device
+    #   acceleration = Acceleration.new(provider: 'cuda', device_id: 0)
+    #
+    class Acceleration
+      attr_reader :provider, :device_id
+
+      def initialize(provider: 'auto', device_id: 0)
+        @provider = provider.to_s
+        @device_id = device_id.to_i
+      end
+
+      def to_h
+        {
+          provider: @provider,
+          device_id: @device_id
+        }
+      end
+    end
+
+    # Layout detection configuration
+    #
+    # @example Basic usage with fast preset
+    #   layout = LayoutDetection.new(preset: "fast")
+    #
+    # @example Accurate preset with custom threshold
+    #   layout = LayoutDetection.new(
+    #     preset: "accurate",
+    #     confidence_threshold: 0.5,
+    #     apply_heuristics: true
+    #   )
+    #
+    class LayoutDetection
+      attr_reader :preset, :confidence_threshold, :apply_heuristics
+
+      def initialize(preset: 'fast', confidence_threshold: nil, apply_heuristics: true)
+        @preset = preset.to_s
+        @confidence_threshold = confidence_threshold&.to_f
+        @apply_heuristics = apply_heuristics ? true : false
+      end
+
+      def to_h
+        {
+          preset: @preset,
+          confidence_threshold: @confidence_threshold,
+          apply_heuristics: @apply_heuristics
+        }.compact
+      end
+    end
+
+    # Concurrency configuration for thread pool management
+    #
+    # @example Limit max threads
+    #   concurrency = Concurrency.new(max_threads: 4)
+    #
+    class Concurrency
+      attr_reader :max_threads
+
+      def initialize(max_threads: nil)
+        @max_threads = max_threads&.to_i
+      end
+
+      def to_h
+        h = {}
+        h[:max_threads] = @max_threads unless @max_threads.nil?
+        h
+      end
+    end
+
     # Main extraction configuration
     #
     # @example Basic usage
@@ -847,7 +930,7 @@ module Kreuzberg
                   :images, :postprocessor,
                   :token_reduction, :keywords, :html_options, :pages,
                   :max_concurrent_extractions, :output_format, :result_format,
-                  :security_limits
+                  :security_limits, :layout, :concurrency
 
       # Alias for backward compatibility - image_extraction is the canonical name
       alias image_extraction images
@@ -872,7 +955,7 @@ module Kreuzberg
         language_detection pdf_options image_extraction
         postprocessor token_reduction keywords html_options pages
         max_concurrent_extractions output_format result_format
-        security_limits
+        security_limits layout concurrency
       ].freeze
 
       # Aliases for backward compatibility
@@ -947,7 +1030,9 @@ module Kreuzberg
                      max_concurrent_extractions: nil,
                      output_format: nil,
                      result_format: nil,
-                     security_limits: nil)
+                     security_limits: nil,
+                     layout: nil,
+                     concurrency: nil)
         kwargs = {
           use_cache: use_cache, enable_quality_processing: enable_quality_processing,
           force_ocr: force_ocr, include_document_structure: include_document_structure,
@@ -957,7 +1042,8 @@ module Kreuzberg
           token_reduction: token_reduction, keywords: keywords, html_options: html_options,
           pages: pages, max_concurrent_extractions: max_concurrent_extractions,
           output_format: output_format, result_format: result_format,
-          security_limits: security_limits
+          security_limits: security_limits, layout: layout,
+          concurrency: concurrency
         }
         extracted = extract_from_hash(hash, kwargs)
 
@@ -971,7 +1057,7 @@ module Kreuzberg
         defaults.merge(hash.slice(*defaults.keys))
       end
 
-      def assign_attributes(params)
+      def assign_attributes(params) # rubocop:disable Metrics/MethodLength
         @use_cache = params[:use_cache] ? true : false
         @enable_quality_processing = params[:enable_quality_processing] ? true : false
         @force_ocr = params[:force_ocr] ? true : false
@@ -986,6 +1072,8 @@ module Kreuzberg
         @keywords = normalize_config(params[:keywords], Keywords)
         @html_options = normalize_config(params[:html_options], HtmlOptions)
         @pages = normalize_config(params[:pages], PageConfig)
+        @layout = normalize_config(params[:layout], LayoutDetection)
+        @concurrency = normalize_config(params[:concurrency], Concurrency)
         @max_concurrent_extractions = params[:max_concurrent_extractions]&.to_i
         @output_format = validate_output_format(params[:output_format])
         @result_format = validate_result_format(params[:result_format])
@@ -1034,7 +1122,8 @@ module Kreuzberg
           language_detection: @language_detection&.to_h, pdf_options: @pdf_options&.to_h,
           image_extraction: @images&.to_h, postprocessor: @postprocessor&.to_h,
           token_reduction: @token_reduction&.to_h, keywords: @keywords&.to_h,
-          html_options: @html_options&.to_h, pages: @pages&.to_h
+          html_options: @html_options&.to_h, pages: @pages&.to_h,
+          layout: @layout&.to_h, concurrency: @concurrency&.to_h
         }
       end
 
@@ -1172,6 +1261,10 @@ module Kreuzberg
           @html_options = normalize_config(value, HtmlOptions)
         when :pages
           @pages = normalize_config(value, PageConfig)
+        when :layout
+          @layout = normalize_config(value, LayoutDetection)
+        when :concurrency
+          @concurrency = normalize_config(value, Concurrency)
         when :max_concurrent_extractions
           @max_concurrent_extractions = value&.to_i
         when :output_format
@@ -1228,6 +1321,12 @@ module Kreuzberg
       end
 
       def update_from_merged(merged)
+        update_core_options(merged)
+        update_processing_options(merged)
+        update_output_options(merged)
+      end
+
+      def update_core_options(merged)
         @use_cache = merged.use_cache
         @enable_quality_processing = merged.enable_quality_processing
         @force_ocr = merged.force_ocr
@@ -1235,6 +1334,9 @@ module Kreuzberg
         @ocr = merged.ocr
         @chunking = merged.chunking
         @language_detection = merged.language_detection
+      end
+
+      def update_processing_options(merged)
         @pdf_options = merged.pdf_options
         @images = merged.image_extraction
         @postprocessor = merged.postprocessor
@@ -1242,6 +1344,11 @@ module Kreuzberg
         @keywords = merged.keywords
         @html_options = merged.html_options
         @pages = merged.pages
+        @layout = merged.layout
+      end
+
+      def update_output_options(merged)
+        @concurrency = merged.concurrency
         @max_concurrent_extractions = merged.max_concurrent_extractions
         @output_format = merged.output_format
         @result_format = merged.result_format

data/lib/kreuzberg/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kreuzberg
-  VERSION = '4.4.6'
+  VERSION = '4.5.1'
 end

data/sig/kreuzberg.rbs

@@ -240,7 +240,9 @@ module Kreuzberg
       attr_reader det_db_unclip_ratio: Float?
       attr_reader det_limit_side_len: Integer?
       attr_reader rec_batch_num: Integer?
-      def initialize: (?language: String?, ?cache_dir: String?, ?use_angle_cls: bool?, ?enable_table_detection: bool?, ?det_db_thresh: Float?, ?det_db_box_thresh: Float?, ?det_db_unclip_ratio: Float?, ?det_limit_side_len: Integer?, ?rec_batch_num: Integer?) -> void
+      attr_reader padding: Integer?
+      attr_reader model_tier: String?
+      def initialize: (?language: String?, ?cache_dir: String?, ?use_angle_cls: bool?, ?enable_table_detection: bool?, ?det_db_thresh: Float?, ?det_db_box_thresh: Float?, ?det_db_unclip_ratio: Float?, ?det_limit_side_len: Integer?, ?rec_batch_num: Integer?, ?padding: Integer?, ?model_tier: String?) -> void
       def to_h: () -> Hash[Symbol, untyped]
     end
 
@@ -332,8 +334,9 @@ module Kreuzberg
       attr_reader extract_annotations: bool
       attr_reader top_margin_fraction: Float?
       attr_reader bottom_margin_fraction: Float?
+      attr_reader allow_single_column_tables: bool
 
-      def initialize: (?extract_images: bool, ?passwords: (Array[String] | String)?, ?extract_metadata: bool, ?font_config: (FontConfig | Hash[Symbol, untyped])?, ?hierarchy: (Hierarchy | Hash[Symbol, untyped])?, ?extract_annotations: bool, ?top_margin_fraction: Float?, ?bottom_margin_fraction: Float?) -> void
+      def initialize: (?extract_images: bool, ?passwords: (Array[String] | String)?, ?extract_metadata: bool, ?font_config: (FontConfig | Hash[Symbol, untyped])?, ?hierarchy: (Hierarchy | Hash[Symbol, untyped])?, ?extract_annotations: bool, ?top_margin_fraction: Float?, ?bottom_margin_fraction: Float?, ?allow_single_column_tables: bool) -> void
       def to_h: () -> Hash[Symbol, untyped]
     end
 
@@ -456,6 +459,22 @@ module Kreuzberg
       def to_h: () -> Hash[Symbol, untyped]
     end
 
+    class LayoutDetection
+      attr_reader preset: String
+      attr_reader confidence_threshold: Float?
+      attr_reader apply_heuristics: bool
+
+      def initialize: (?preset: String, ?confidence_threshold: Float?, ?apply_heuristics: bool) -> void
+      def to_h: () -> Hash[Symbol, untyped]
+    end
+
+    class Concurrency
+      attr_reader max_threads: Integer?
+
+      def initialize: (?max_threads: Integer?) -> void
+      def to_h: () -> Hash[Symbol, untyped]
+    end
+
     class Extraction
       attr_reader use_cache: bool
       attr_reader enable_quality_processing: bool
@@ -471,6 +490,8 @@ module Kreuzberg
       attr_reader keywords: Keywords?
       attr_reader html_options: HtmlOptions?
       attr_reader pages: PageConfig?
+      attr_reader layout: LayoutDetection?
+      attr_reader concurrency: Concurrency?
       attr_reader max_concurrent_extractions: Integer?
       attr_reader output_format: String?
       attr_reader result_format: String?
@@ -495,6 +516,8 @@ module Kreuzberg
         ?keywords: (Keywords | Hash[Symbol, untyped])?,
         ?html_options: (HtmlOptions | Hash[Symbol, untyped])?,
         ?pages: (PageConfig | Hash[Symbol, untyped])?,
+        ?layout: (LayoutDetection | Hash[Symbol, untyped])?,
+        ?concurrency: (Concurrency | Hash[Symbol, untyped])?,
         ?max_concurrent_extractions: Integer?,
         ?output_format: String?,
         ?result_format: String?
@@ -755,6 +778,7 @@ module Kreuzberg
 
   type config_hash = Hash[Symbol, untyped]
   type config_input = config_hash | _ToH
+  type file_config_input = Hash[Symbol, untyped]?
 
   interface _ToH
     def to_h: () -> config_hash
@@ -1179,13 +1203,15 @@ module Kreuzberg
 
   def self.batch_extract_files_sync: (
     paths: Array[String | Pathname],
-    ?config: config_input?
+    ?config: config_input?,
+    ?file_configs: Array[file_config_input]?
   ) -> Array[Result]
 
   def self.batch_extract_bytes_sync: (
     data_array: Array[String],
     mime_types: Array[String],
-    ?config: config_input?
+    ?config: config_input?,
+    ?file_configs: Array[file_config_input]?
   ) -> Array[Result]
 
   def self.extract_file: (
@@ -1202,13 +1228,15 @@ module Kreuzberg
 
   def self.batch_extract_files: (
     paths: Array[String | Pathname],
-    ?config: config_input?
+    ?config: config_input?,
+    ?file_configs: Array[file_config_input]?
   ) -> Array[Result]
 
   def self.batch_extract_bytes: (
     data_array: Array[String],
     mime_types: Array[String],
-    ?config: config_input?
+    ?config: config_input?,
+    ?file_configs: Array[file_config_input]?
   ) -> Array[Result]
 
   # Cache API

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: kreuzberg
 version: !ruby/object:Gem::Version
-  version: 4.4.6
+  version: 4.5.1
 platform: x86_64-linux
 authors:
 - Na'aman Hirschfeld
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-03-13 00:00:00.000000000 Z
+date: 2026-03-21 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler