kreuzberg 4.3.6 → 4.3.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 1b3eb519a94cf2a82e9d9b649ce98e122fec14d01568ced3edc925f1cb49f4ad
-  data.tar.gz: cdfa987af6f7bc8b0a6defb76b3426a616ed9d9451bd3aa27680eae7eba5325c
+  metadata.gz: 1c9f9cd70dd541fd9c193c1ab60782eb04e287019f707eeb0d6eb853f64fc039
+  data.tar.gz: ce8345437d8e47062a21799b0352ceb2d5917a0546421b39a928fe1b659dc28d
 SHA512:
-  metadata.gz: 1c6170355aa3f4443b68aed401e2d7b8f20a4b792c18e130b5f59be9ea3102527bb9098f77f30d1383287d404846ca99a397807827284e7b0c774df58d85cd51
-  data.tar.gz: 5362d11257dd57715e8f9e0743a3f5e8fba64c9177f4858b587fd7230dbae2c54fac194d0aa64da7db1f970ad8eaf605683d6547cdd39c81b55d0bebe43502d5
+  metadata.gz: 012fac8575af2561d8b114649ef370ee281e90629a3ac95abad398dc3371853109e76e19b4aca9aeeba315853cae72f2ff9788681bd7b1bcfa3cb8245dbaa939
+  data.tar.gz: fce6aa4f32a4651821d20ba61c77d6b61fc1038492e4b7d960c585ce2c131dd228430835fcb35f945998c64ae3faa71655ab5748b6243e2c70453e9505a442a0

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    kreuzberg (4.3.6)
+    kreuzberg (4.3.7)
       rb_sys (~> 0.9.119)
       sorbet-runtime (~> 0.5)
 
@@ -124,7 +124,7 @@ GEM
       rubocop (~> 1.81)
     ruby-progressbar (1.13.0)
     securerandom (0.4.1)
-    sorbet-runtime (0.6.12942)
+    sorbet-runtime (0.6.12956)
     steep (1.10.0)
       activesupport (>= 5.1)
       concurrent-ruby (>= 1.1.10)
@@ -210,7 +210,7 @@ CHECKSUMS
   i18n (1.14.8) sha256=285778639134865c5e0f6269e0b818256017e8cde89993fdfcbfb64d088824a5
   io-console (0.8.2) sha256=d6e3ae7a7cc7574f4b8893b4fca2162e57a825b223a177b7afa236c5ef9814cc
   json (2.18.1) sha256=fe112755501b8d0466b5ada6cf50c8c3f41e897fa128ac5d263ec09eedc9f986
-  kreuzberg (4.3.6)
+  kreuzberg (4.3.7)
   language_server-protocol (3.17.0.5) sha256=fd1e39a51a28bf3eec959379985a72e296e9f9acfce46f6a79d31ca8760803cc
   lint_roller (1.1.0) sha256=2c0c845b632a7d172cb849cc90c1bce937a28c5c8ccccb50dfd46a485003cc87
   listen (3.10.0) sha256=c6e182db62143aeccc2e1960033bebe7445309c7272061979bb098d03760c9d2
@@ -245,7 +245,7 @@ CHECKSUMS
   rubocop-rspec (3.9.0) sha256=8fa70a3619408237d789aeecfb9beef40576acc855173e60939d63332fdb55e2
   ruby-progressbar (1.13.0) sha256=80fc9c47a9b640d6834e0dc7b3c94c9df37f08cb072b7761e4a71e22cff29b33
   securerandom (0.4.1) sha256=cc5193d414a4341b6e225f0cb4446aceca8e50d5e1888743fac16987638ea0b1
-  sorbet-runtime (0.6.12942) sha256=967bda04814d234e4239c4f883c1d0ee6de3e47bf8bafd2c0cc30d18df2ddd3a
+  sorbet-runtime (0.6.12956) sha256=fee716a62d0b1d94ebc8e6ba23e76a7654eeac66c1f5cc1e1bef78b8e9ff87c7
   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=v4.3.6" alt="Go">
+    <img src="https://img.shields.io/github/v/tag/kreuzberg-dev/kreuzberg?label=Go&color=007ec6&filter=v4.3.7" 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/ext/kreuzberg_rb/native/Cargo.lock

@@ -16,7 +16,7 @@ checksum = "b169f7a6d4742236a0a00c541b845991d0ac43e546831af1249753ab4c3aa3a0"
 dependencies = [
  "cfg-if",
  "cipher",
- "cpufeatures",
+ "cpufeatures 0.2.17",
 ]
 
 [[package]]
@@ -795,6 +795,17 @@ version = "0.2.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "613afe47fcd5fac7ccf1db93babcb082c5994d996f20b8b159f2ad1658eb5724"
 
+[[package]]
+name = "chacha20"
+version = "0.10.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6f8d983286843e49675a4b7a2d174efe136dc93a18d69130dd18198a6c167601"
+dependencies = [
+ "cfg-if",
+ "cpufeatures 0.3.0",
+ "rand_core 0.10.0",
+]
+
 [[package]]
 name = "chardetng"
 version = "0.1.17"
@@ -1099,6 +1110,15 @@ dependencies = [
  "libc",
 ]
 
+[[package]]
+name = "cpufeatures"
+version = "0.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8b2a41393f66f16b0823bb79094d54ac5fbd34ab292ddafb9a0456ac9f87d201"
+dependencies = [
+ "libc",
+]
+
 [[package]]
 name = "crc"
 version = "3.3.0"
@@ -1632,9 +1652,9 @@ dependencies = [
 
 [[package]]
 name = "fastembed"
-version = "5.9.0"
+version = "5.11.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f6f8b48452e32595e2fcb89d0c582490681ef4febca0ae037fc75a122525953a"
+checksum = "b4339d45a80579ab8305616a501eacdbf18fb0f7def7fa6e4c0b75941416d5b0"
 dependencies = [
  "anyhow",
  "hf-hub",
@@ -1961,6 +1981,7 @@ dependencies = [
  "js-sys",
  "libc",
  "r-efi",
+ "rand_core 0.10.0",
  "wasip2",
  "wasip3",
  "wasm-bindgen",
@@ -2258,9 +2279,9 @@ dependencies = [
 
 [[package]]
 name = "html-to-markdown-rs"
-version = "2.25.0"
+version = "2.25.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bb31d75f2fdbc8d889d78a912e10c22c30451afb44ee3310f5bfcabf79a31a17"
+checksum = "c05335c6bf406653110ad8447c84461c6d0cda5e0aff9d3d3518f87502d30abe"
 dependencies = [
  "ahash",
  "astral-tl",
@@ -2880,7 +2901,7 @@ dependencies = [
 
 [[package]]
 name = "kreuzberg"
-version = "4.3.4"
+version = "4.3.6"
 dependencies = [
  "ahash",
  "async-trait",
@@ -2955,7 +2976,7 @@ dependencies = [
  "thiserror 2.0.18",
  "tiff 0.11.0",
  "tokio",
- "toml 1.0.1+spec-1.1.0",
+ "toml 1.0.3+spec-1.1.0",
  "tower",
  "tower-http",
  "tracing",
@@ -2967,12 +2988,12 @@ dependencies = [
  "uuid",
  "whatlang",
  "yake-rust",
- "zip 8.0.0",
+ "zip 8.1.0",
 ]
 
 [[package]]
 name = "kreuzberg-ffi"
-version = "4.3.4"
+version = "4.3.6"
 dependencies = [
  "ahash",
  "async-trait",
@@ -2988,7 +3009,7 @@ dependencies = [
 
 [[package]]
 name = "kreuzberg-paddle-ocr"
-version = "4.3.4"
+version = "4.3.6"
 dependencies = [
  "geo-clipper",
  "geo-types",
@@ -3003,7 +3024,7 @@ dependencies = [
 
 [[package]]
 name = "kreuzberg-pdfium-render"
-version = "4.3.4"
+version = "4.3.6"
 dependencies = [
  "bitflags",
  "bytemuck",
@@ -3027,7 +3048,7 @@ dependencies = [
 
 [[package]]
 name = "kreuzberg-rb"
-version = "4.3.4"
+version = "4.3.6"
 dependencies = [
  "async-trait",
  "html-to-markdown-rs",
@@ -3044,14 +3065,14 @@ dependencies = [
 
 [[package]]
 name = "kreuzberg-tesseract"
-version = "4.3.4"
+version = "4.3.6"
 dependencies = [
  "cc",
  "cmake",
  "libc",
  "reqwest 0.13.2",
  "thiserror 2.0.18",
- "zip 8.0.0",
+ "zip 8.1.0",
 ]
 
 [[package]]
@@ -4985,6 +5006,17 @@ dependencies = [
  "rand_core 0.9.3",
 ]
 
+[[package]]
+name = "rand"
+version = "0.10.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "bc266eb313df6c5c09c1c7b1fbe2510961e5bcd3add930c1e31f7ed9da0feff8"
+dependencies = [
+ "chacha20",
+ "getrandom 0.4.1",
+ "rand_core 0.10.0",
+]
+
 [[package]]
 name = "rand_chacha"
 version = "0.9.0"
@@ -5010,6 +5042,12 @@ dependencies = [
  "getrandom 0.3.4",
 ]
 
+[[package]]
+name = "rand_core"
+version = "0.10.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0c8d0fd677905edcbeedbf2edb6494d676f0e98d54d5cf9bda0b061cb8fb8aba"
+
 [[package]]
 name = "rand_distr"
 version = "0.5.1"
@@ -5362,12 +5400,11 @@ dependencies = [
 
 [[package]]
 name = "rmcp"
-version = "0.15.0"
+version = "0.16.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1bef41ebc9ebed2c1b1d90203e9d1756091e8a00bbc3107676151f39868ca0ee"
+checksum = "cc4c9c94680f75470ee8083a0667988b5d7b5beb70b9f998a8e51de7c682ce60"
 dependencies = [
  "async-trait",
- "axum",
  "base64 0.22.1",
  "bytes",
  "chrono",
@@ -5377,7 +5414,7 @@ dependencies = [
  "http-body-util",
  "pastey 0.2.1",
  "pin-project-lite",
- "rand 0.9.2",
+ "rand 0.10.0",
  "rmcp-macros",
  "schemars",
  "serde",
@@ -5394,9 +5431,9 @@ dependencies = [
 
 [[package]]
 name = "rmcp-macros"
-version = "0.15.0"
+version = "0.16.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0e88ad84b8b6237a934534a62b379a5be6388915663c0cc598ceb9b3292bbbfe"
+checksum = "90c23c8f26cae4da838fbc3eadfaecf2d549d97c04b558e7bd90526a9c28b42a"
 dependencies = [
  "darling 0.23.0",
  "proc-macro2",
@@ -5861,7 +5898,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "e3bf829a2d51ab4a5ddf1352d8470c140cadc8301b2ae1789db023f01cedd6ba"
 dependencies = [
  "cfg-if",
- "cpufeatures",
+ "cpufeatures 0.2.17",
  "digest",
 ]
 
@@ -5872,7 +5909,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "a7507d819769d01a365ab707794a4084392c824f54a7a6a7862f8c3d0892b283"
 dependencies = [
  "cfg-if",
- "cpufeatures",
+ "cpufeatures 0.2.17",
  "digest",
 ]
 
@@ -6584,9 +6621,9 @@ dependencies = [
 
 [[package]]
 name = "toml"
-version = "1.0.1+spec-1.1.0"
+version = "1.0.3+spec-1.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bbe30f93627849fa362d4a602212d41bb237dc2bd0f8ba0b2ce785012e124220"
+checksum = "c7614eaf19ad818347db24addfa201729cf2a9b6fdfd9eb0ab870fcacc606c0c"
 dependencies = [
  "indexmap",
  "serde_core",
@@ -6640,9 +6677,9 @@ dependencies = [
 
 [[package]]
 name = "toml_parser"
-version = "1.0.8+spec-1.1.0"
+version = "1.0.9+spec-1.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0742ff5ff03ea7e67c8ae6c93cac239e0d9784833362da3f9a9c1da8dfefcbdc"
+checksum = "702d4415e08923e7e1ef96cd5727c0dfed80b4d2fa25db9647fe5eb6f7c5a4c4"
 dependencies = [
  "winnow",
 ]
@@ -8072,9 +8109,9 @@ dependencies = [
 
 [[package]]
 name = "zip"
-version = "8.0.0"
+version = "8.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "79b32dd4ad3aca14ae109f8cce0495ac1c57f6f4f00ad459a40e582f89440d97"
+checksum = "6e499faf5c6b97a0d086f4a8733de6d47aee2252b8127962439d8d4311a73f72"
 dependencies = [
  "aes",
  "bzip2",

data/ext/kreuzberg_rb/native/Cargo.toml

@@ -37,7 +37,7 @@ collapsible_if = "allow"
 
 [package]
 name = "kreuzberg-rb"
-version = "4.3.6"
+version = "4.3.7"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]

data/ext/kreuzberg_rb/native/src/config/types.rs

@@ -240,11 +240,40 @@ pub fn parse_pdf_config(ruby: &Ruby, hash: RHash) -> Result<PdfConfig, Error> {
         None
     };
 
+    let extract_annotations = if let Some(val) = get_kw(ruby, hash, "extract_annotations") {
+        bool::try_convert(val)?
+    } else {
+        false
+    };
+
+    let top_margin_fraction = if let Some(val) = get_kw(ruby, hash, "top_margin_fraction") {
+        if !val.is_nil() {
+            Some(f32::try_convert(val)?)
+        } else {
+            None
+        }
+    } else {
+        None
+    };
+
+    let bottom_margin_fraction = if let Some(val) = get_kw(ruby, hash, "bottom_margin_fraction") {
+        if !val.is_nil() {
+            Some(f32::try_convert(val)?)
+        } else {
+            None
+        }
+    } else {
+        None
+    };
+
     let config = PdfConfig {
         extract_images,
         passwords,
         extract_metadata,
         hierarchy,
+        extract_annotations,
+        top_margin_fraction,
+        bottom_margin_fraction,
     };
 
     Ok(config)

data/ext/kreuzberg_rb/native/src/plugins/ocr_backend.rs

@@ -104,6 +104,7 @@ impl OcrBackend for RubyOcrBackend {
                 extracted_keywords: None,
                 quality_score: None,
                 processing_warnings: vec![],
+                annotations: None,
             })
         })
     }

data/ext/kreuzberg_rb/native/src/result.rs

@@ -640,5 +640,38 @@ pub fn extraction_result_to_ruby(ruby: &Ruby, result: RustExtractionResult) -> R
     }
     set_hash_entry(ruby, &hash, "processing_warnings", warnings_array.into_value_with(ruby))?;
 
+    // Convert annotations
+    if let Some(annotations) = result.annotations {
+        let annotations_array = ruby.ary_new();
+        for annot in annotations {
+            let annot_hash = ruby.hash_new();
+            let type_str = serde_json::to_value(&annot.annotation_type)
+                .ok()
+                .and_then(|v| v.as_str().map(String::from))
+                .unwrap_or_default();
+            annot_hash.aset("annotation_type", type_str.as_str())?;
+            if let Some(content) = annot.content {
+                annot_hash.aset("content", content.as_str())?;
+            } else {
+                annot_hash.aset("content", ruby.qnil().as_value())?;
+            }
+            annot_hash.aset("page_number", annot.page_number as i64)?;
+            if let Some(bbox) = annot.bounding_box {
+                let bbox_hash = ruby.hash_new();
+                bbox_hash.aset("x0", bbox.x0)?;
+                bbox_hash.aset("y0", bbox.y0)?;
+                bbox_hash.aset("x1", bbox.x1)?;
+                bbox_hash.aset("y1", bbox.y1)?;
+                annot_hash.aset("bounding_box", bbox_hash)?;
+            } else {
+                annot_hash.aset("bounding_box", ruby.qnil().as_value())?;
+            }
+            annotations_array.push(annot_hash)?;
+        }
+        set_hash_entry(ruby, &hash, "annotations", annotations_array.into_value_with(ruby))?;
+    } else {
+        set_hash_entry(ruby, &hash, "annotations", ruby.qnil().as_value())?;
+    }
+
     Ok(hash)
 }

data/lib/kreuzberg/config.rb

@@ -391,14 +391,18 @@ module Kreuzberg
     #   pdf = PDF.new(extract_images: true, hierarchy: hierarchy)
     #
     class PDF
-      attr_reader :extract_images, :passwords, :extract_metadata, :font_config, :hierarchy
+      attr_reader :extract_images, :passwords, :extract_metadata, :font_config, :hierarchy,
+                  :extract_annotations, :top_margin_fraction, :bottom_margin_fraction
 
       def initialize(
         extract_images: false,
         passwords: nil,
         extract_metadata: true,
         font_config: nil,
-        hierarchy: nil
+        hierarchy: nil,
+        extract_annotations: false,
+        top_margin_fraction: nil,
+        bottom_margin_fraction: nil
       )
         @extract_images = extract_images ? true : false
         @passwords = if passwords.is_a?(Array)
@@ -409,6 +413,9 @@ module Kreuzberg
         @extract_metadata = extract_metadata ? true : false
         @font_config = normalize_font_config(font_config)
         @hierarchy = normalize_hierarchy(hierarchy)
+        @extract_annotations = extract_annotations ? true : false
+        @top_margin_fraction = top_margin_fraction&.to_f
+        @bottom_margin_fraction = bottom_margin_fraction&.to_f
       end
 
       def to_h
@@ -417,7 +424,10 @@ module Kreuzberg
           passwords: @passwords,
           extract_metadata: @extract_metadata,
           font_config: @font_config&.to_h,
-          hierarchy: @hierarchy&.to_h
+          hierarchy: @hierarchy&.to_h,
+          extract_annotations: @extract_annotations,
+          top_margin_fraction: @top_margin_fraction,
+          bottom_margin_fraction: @bottom_margin_fraction
         }.compact
       end
 

data/lib/kreuzberg/result.rb

@@ -14,7 +14,7 @@ module Kreuzberg
   class Result
     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
+                :document, :extracted_keywords, :quality_score, :processing_warnings, :annotations
 
     # @!attribute [r] cells
     #   @return [Array<Array<String>>] Table cells (2D array)
@@ -339,6 +339,7 @@ module Kreuzberg
       @extracted_keywords = parse_extracted_keywords(get_value(hash, 'extracted_keywords'))
       @quality_score = get_value(hash, 'quality_score')
       @processing_warnings = parse_processing_warnings(get_value(hash, 'processing_warnings'))
+      @annotations = parse_annotations(get_value(hash, 'annotations'))
     end
     # rubocop:enable Metrics/AbcSize
 
@@ -346,6 +347,7 @@ module Kreuzberg
     #
     # @return [Hash] Hash representation
     #
+    # rubocop:disable Metrics/CyclomaticComplexity
     def to_h
       {
         content: @content,
@@ -362,9 +364,11 @@ module Kreuzberg
         document: @document&.to_h,
         extracted_keywords: @extracted_keywords&.map(&:to_h),
         quality_score: @quality_score,
-        processing_warnings: @processing_warnings.map(&:to_h)
+        processing_warnings: @processing_warnings.map(&:to_h),
+        annotations: @annotations&.map(&:to_h)
       }
     end
+    # rubocop:enable Metrics/CyclomaticComplexity
 
     # Convert to JSON
     #
@@ -707,6 +711,32 @@ module Kreuzberg
         )
       end
     end
+
+    def parse_annotations(annotations_data)
+      return nil if annotations_data.nil?
+
+      annotations_data.map { |a_hash| build_annotation(a_hash) }
+    end
+
+    def build_annotation(a_hash)
+      PdfAnnotation.new(
+        annotation_type: a_hash['annotation_type'] || '',
+        content: a_hash['content'],
+        page_number: a_hash['page_number']&.to_i,
+        bounding_box: build_annotation_bbox(a_hash['bounding_box'])
+      )
+    end
+
+    def build_annotation_bbox(bbox_data)
+      return nil if bbox_data.nil?
+
+      PdfAnnotationBoundingBox.new(
+        left: bbox_data['left']&.to_f,
+        top: bbox_data['top']&.to_f,
+        right: bbox_data['right']&.to_f,
+        bottom: bbox_data['bottom']&.to_f
+      )
+    end
   end
   # rubocop:enable Metrics/ClassLength
 end

data/lib/kreuzberg/types.rb

@@ -411,4 +411,24 @@ module Kreuzberg
 
     const :nodes, T::Array[DocumentNode]
   end
+
+  # Bounding box for a PDF annotation.
+  class PdfAnnotationBoundingBox < T::Struct
+    extend T::Sig
+
+    const :left, T.nilable(Float)
+    const :top, T.nilable(Float)
+    const :right, T.nilable(Float)
+    const :bottom, T.nilable(Float)
+  end
+
+  # A PDF annotation extracted from a document page.
+  class PdfAnnotation < T::Struct
+    extend T::Sig
+
+    const :annotation_type, String
+    const :content, T.nilable(String)
+    const :page_number, T.nilable(Integer)
+    const :bounding_box, T.nilable(PdfAnnotationBoundingBox)
+  end
 end

data/lib/kreuzberg/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kreuzberg
-  VERSION = '4.3.6'
+  VERSION = '4.3.7'
 end

data/sig/kreuzberg.rbs

@@ -323,8 +323,11 @@ module Kreuzberg
       attr_reader extract_metadata: bool
       attr_reader font_config: FontConfig?
       attr_reader hierarchy: Hierarchy?
+      attr_reader extract_annotations: bool
+      attr_reader top_margin_fraction: Float?
+      attr_reader bottom_margin_fraction: Float?
 
-      def initialize: (?extract_images: bool, ?passwords: (Array[String] | String)?, ?extract_metadata: bool, ?font_config: (FontConfig | Hash[Symbol, untyped])?, ?hierarchy: (Hierarchy | Hash[Symbol, untyped])?) -> 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?) -> void
       def to_h: () -> Hash[Symbol, untyped]
     end
 
@@ -525,6 +528,15 @@ module Kreuzberg
   end
 
   # Extraction result type
+  type pdf_annotation_type = 'text' | 'highlight' | 'link' | 'stamp' | 'underline' | 'strike_out' | 'other'
+
+  type pdf_annotation_hash = {
+    annotation_type: String,
+    content: String?,
+    page_number: Integer,
+    bounding_box: { x0: Float, y0: Float, x1: Float, y1: Float }?
+  }
+
   type extraction_result_hash = {
     content: String,
     mime_type: String,
@@ -541,7 +553,8 @@ module Kreuzberg
     document: document_structure_hash?,
     extracted_keywords: Array[extracted_keyword_hash]?,
     quality_score: Float?,
-    processing_warnings: Array[processing_warning_hash]?
+    processing_warnings: Array[processing_warning_hash]?,
+    annotations: Array[pdf_annotation_hash]?
   }
 
   type extracted_keyword_hash = {
@@ -1076,6 +1089,18 @@ module Kreuzberg
     attr_reader extracted_keywords: Array[ExtractedKeyword]?
     attr_reader quality_score: Float?
     attr_reader processing_warnings: Array[ProcessingWarning]?
+    attr_reader annotations: Array[PdfAnnotation]?
+
+    # PDF annotation extracted from a document page (Struct from result.rb)
+    class PdfAnnotation
+      attr_reader annotation_type: String
+      attr_reader content: String?
+      attr_reader page_number: Integer
+      attr_reader bounding_box: BoundingBox?
+
+      def initialize: (annotation_type: String, content: String?, page_number: Integer, bounding_box: BoundingBox?) -> void
+      def to_h: () -> pdf_annotation_hash
+    end
 
     def initialize: (extraction_result_hash hash) -> void
     def to_h: () -> Hash[Symbol, untyped]
@@ -1113,6 +1138,7 @@ module Kreuzberg
     def parse_ocr_geometry: (Hash[String, untyped]? data) -> OcrBoundingGeometry?
     def parse_ocr_confidence: (Hash[String, untyped]? data) -> OcrConfidence?
     def parse_ocr_rotation: (Hash[String, untyped]? data) -> OcrRotation?
+    def parse_annotations: (Array[pdf_annotation_hash]? annotations_data) -> Array[PdfAnnotation]?
   end
 
   # Module methods (extraction API)

data/vendor/Cargo.toml

@@ -2,7 +2,7 @@
 members = ["kreuzberg", "kreuzberg-ffi", "kreuzberg-tesseract", "kreuzberg-paddle-ocr", "kreuzberg-pdfium-render"]
 
 [workspace.package]
-version = "4.3.6"
+version = "4.3.7"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]
@@ -37,7 +37,7 @@ serde_json = { version = "1.0.149" }
 tempfile = "3.25.0"
 thiserror = "2.0.18"
 tokio = { version = "1.49.0", features = ["rt", "rt-multi-thread", "macros", "sync", "process", "fs", "time", "io-util"] }
-toml = "1.0.2"
+toml = "1.0.3"
 tracing = "0.1"
 wasm-bindgen = { version = "0.2", features = ["enable-interning"] }
 wasm-bindgen-futures = "0.4"