kreuzberg 4.3.4 → 4.3.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: f3c080ecb465d2860ccaef34aff627bf155689e912b0ff48ae72d99a16f045cf
-  data.tar.gz: 8cd3f4aae7a19c229bfb2451811233f59e04c39a9aa9d6d9d69f9e12d747ec04
+  metadata.gz: 3936788c6812a84428d0467330f573c20c9b569c399eab105cc2815d777b2141
+  data.tar.gz: 1b115f87bc4a40960584de9459725d0afabfefe6244ddea434461a2c6f36a647
 SHA512:
-  metadata.gz: c1948824a088e4d6296bd8f161bd59624a252763b47ef21b2b0763d8be0a472de814f95273e03fd82b6c1f8c66aea2fde8fd70c3315ad296be1eaaf09cf3e64c
-  data.tar.gz: 46340a09386141c8ddc19c4c23f9a3b9e4b74cfa8070d3aeec6ae0e339aec114cf3d83a5d08ef1d9734f293186829b075f1cb79f619fd59133148f9571fa9a4a
+  metadata.gz: 73fb7522dcd091b449d5146e65f008bded545a6c49a35d77468bd9bfa61d30a7413dc7364eef6ce79cca4606c5c710ea3f574509743569a7fb4f8a7bc579f402
+  data.tar.gz: e21fb401768da5005a1edb720b0c00c82aaa9a8ef60b6f2bab3587b3c8c94cd8fdcc220daeb8545bae6523bf8be22dbc1dab7c966ae8f30e888342d96c7e5df2

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    kreuzberg (4.3.4)
+    kreuzberg (4.3.5)
       rb_sys (~> 0.9.119)
 
 GEM
@@ -59,7 +59,7 @@ GEM
       prism (~> 1.5)
     mutex_m (0.3.0)
     parallel (1.27.0)
-    parser (3.3.10.1)
+    parser (3.3.10.2)
       ast (~> 2.4.1)
       racc
     prism (1.9.0)
@@ -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 (4.3.4)
+  kreuzberg (4.3.5)
   language_server-protocol (3.17.0.5) sha256=fd1e39a51a28bf3eec959379985a72e296e9f9acfce46f6a79d31ca8760803cc
   lint_roller (1.1.0) sha256=2c0c845b632a7d172cb849cc90c1bce937a28c5c8ccccb50dfd46a485003cc87
   listen (3.10.0) sha256=c6e182db62143aeccc2e1960033bebe7445309c7272061979bb098d03760c9d2
@@ -219,7 +219,7 @@ CHECKSUMS
   minitest (6.0.1) sha256=7854c74f48e2e975969062833adc4013f249a4b212f5e7b9d5c040bf838d54bb
   mutex_m (0.3.0) sha256=cfcb04ac16b69c4813777022fdceda24e9f798e48092a2b817eb4c0a782b0751
   parallel (1.27.0) sha256=4ac151e1806b755fb4e2dc2332cbf0e54f2e24ba821ff2d3dcf86bf6dc4ae130
-  parser (3.3.10.1) sha256=06f6a725d2cd91e5e7f2b7c32ba143631e1f7c8ae2fb918fc4cebec187e6a688
+  parser (3.3.10.2) sha256=6f60c84aa4bdcedb6d1a2434b738fe8a8136807b6adc8f7f53b97da9bc4e9357
   prism (1.9.0) sha256=7b530c6a9f92c24300014919c9dcbc055bf4cdf51ec30aed099b06cd6674ef85
   pry (0.16.0) sha256=d76c69065698ed1f85e717bd33d7942c38a50868f6b0673c636192b3d1b6054e
   pry-byebug (3.12.0) sha256=594e094ae8a8390a7ad4c7b36ae36e13304ed02664c67417d108dc5f7213d1b7

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=v4.3.4" alt="Go">
+    <img src="https://img.shields.io/github/v/tag/kreuzberg-dev/kreuzberg?label=Go&color=007ec6&filter=v4.3.5" 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.toml

@@ -37,7 +37,7 @@ collapsible_if = "allow"
 
 [package]
 name = "kreuzberg-rb"
-version = "4.3.4"
+version = "4.3.5"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]

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

@@ -100,19 +100,19 @@ pub fn ruby_cache_stats() -> Result<RHash, Error> {
 
 // Validation wrapper functions
 pub fn validate_binarization_method(method: String) -> Result<i32, Error> {
-    unsafe { Ok(kreuzberg_validate_binarization_method(method.as_ptr() as *const i8)) }
+    unsafe { Ok(kreuzberg_validate_binarization_method(method.as_ptr() as *const std::os::raw::c_char)) }
 }
 
 pub fn validate_ocr_backend(backend: String) -> Result<i32, Error> {
-    unsafe { Ok(kreuzberg_validate_ocr_backend(backend.as_ptr() as *const i8)) }
+    unsafe { Ok(kreuzberg_validate_ocr_backend(backend.as_ptr() as *const std::os::raw::c_char)) }
 }
 
 pub fn validate_language_code(code: String) -> Result<i32, Error> {
-    unsafe { Ok(kreuzberg_validate_language_code(code.as_ptr() as *const i8)) }
+    unsafe { Ok(kreuzberg_validate_language_code(code.as_ptr() as *const std::os::raw::c_char)) }
 }
 
 pub fn validate_token_reduction_level(level: String) -> Result<i32, Error> {
-    unsafe { Ok(kreuzberg_validate_token_reduction_level(level.as_ptr() as *const i8)) }
+    unsafe { Ok(kreuzberg_validate_token_reduction_level(level.as_ptr() as *const std::os::raw::c_char)) }
 }
 
 pub fn validate_tesseract_psm(psm: i32) -> Result<i32, Error> {
@@ -124,7 +124,7 @@ pub fn validate_tesseract_oem(oem: i32) -> Result<i32, Error> {
 }
 
 pub fn validate_output_format(format: String) -> Result<i32, Error> {
-    unsafe { Ok(kreuzberg_validate_output_format(format.as_ptr() as *const i8)) }
+    unsafe { Ok(kreuzberg_validate_output_format(format.as_ptr() as *const std::os::raw::c_char)) }
 }
 
 pub fn validate_confidence(confidence: f64) -> Result<i32, Error> {

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

@@ -53,6 +53,16 @@ pub fn extraction_result_to_ruby(ruby: &Ruby, result: RustExtractionResult) -> R
         table_hash.aset("cells", cells_array)?;
         table_hash.aset("markdown", table.markdown)?;
         table_hash.aset("page_number", table.page_number)?;
+        if let Some(bbox) = table.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)?;
+            table_hash.aset("bounding_box", bbox_hash)?;
+        } else {
+            table_hash.aset("bounding_box", ruby.qnil().as_value())?;
+        }
 
         tables_array.push(table_hash)?;
     }
@@ -164,6 +174,16 @@ pub fn extraction_result_to_ruby(ruby: &Ruby, result: RustExtractionResult) -> R
             } else {
                 image_hash.aset("ocr_result", ruby.qnil().as_value())?;
             }
+            if let Some(bbox) = image.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)?;
+                image_hash.aset("bounding_box", bbox_hash)?;
+            } else {
+                image_hash.aset("bounding_box", ruby.qnil().as_value())?;
+            }
             images_array.push(image_hash)?;
         }
         set_hash_entry(ruby, &hash, "images", images_array.into_value_with(ruby))?;
@@ -191,6 +211,16 @@ pub fn extraction_result_to_ruby(ruby: &Ruby, result: RustExtractionResult) -> R
                 table_hash.aset("cells", cells_array)?;
                 table_hash.aset("markdown", table.markdown.clone())?;
                 table_hash.aset("page_number", table.page_number as i64)?;
+                if let Some(ref bbox) = table.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)?;
+                    table_hash.aset("bounding_box", bbox_hash)?;
+                } else {
+                    table_hash.aset("bounding_box", ruby.qnil().as_value())?;
+                }
 
                 tables_array.push(table_hash)?;
             }
@@ -248,6 +278,16 @@ pub fn extraction_result_to_ruby(ruby: &Ruby, result: RustExtractionResult) -> R
                 } else {
                     image_hash.aset("ocr_result", ruby.qnil().as_value())?;
                 }
+                if let Some(ref bbox) = image.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)?;
+                    image_hash.aset("bounding_box", bbox_hash)?;
+                } else {
+                    image_hash.aset("bounding_box", ruby.qnil().as_value())?;
+                }
                 images_array.push(image_hash)?;
             }
             page_hash.aset("images", images_array)?;

data/lib/kreuzberg/result.rb

@@ -22,9 +22,11 @@ module Kreuzberg
     #   @return [String] Markdown representation
     # @!attribute [r] page_number
     #   @return [Integer] Page number where table was found
-    Table = Struct.new(:cells, :markdown, :page_number, keyword_init: true) do
+    # @!attribute [r] bounding_box
+    #   @return [BoundingBox, nil] Bounding box of the table on the page
+    Table = Struct.new(:cells, :markdown, :page_number, :bounding_box, keyword_init: true) do
       def to_h
-        { cells: cells, markdown: markdown, page_number: page_number }
+        { cells: cells, markdown: markdown, page_number: page_number, bounding_box: bounding_box&.to_h }
       end
     end
 
@@ -78,6 +80,7 @@ module Kreuzberg
       :bits_per_component,
       :is_mask,
       :description,
+      :bounding_box,
       :ocr_result,
       keyword_init: true
     ) do
@@ -93,6 +96,7 @@ module Kreuzberg
           bits_per_component: bits_per_component,
           is_mask: is_mask,
           description: description,
+          bounding_box: bounding_box&.to_h,
           ocr_result: ocr_result&.to_h
         }
       end
@@ -486,10 +490,12 @@ module Kreuzberg
       return [] if tables_data.nil? || tables_data.empty?
 
       tables_data.map do |table_hash|
+        bounding_box = parse_bounding_box(table_hash['bounding_box'])
         Table.new(
           cells: table_hash['cells'] || [],
           markdown: table_hash['markdown'] || '',
-          page_number: table_hash['page_number'] || 0
+          page_number: table_hash['page_number'] || 0,
+          bounding_box: bounding_box
         )
       end
     end
@@ -521,23 +527,26 @@ module Kreuzberg
     def parse_images(images_data)
       return nil if images_data.nil?
 
-      images_data.map do |image_hash|
-        data = image_hash['data']
-        data = data.dup.force_encoding(Encoding::BINARY) if data.respond_to?(:force_encoding)
-        Image.new(
-          data: data,
-          format: image_hash['format'],
-          image_index: image_hash['image_index'],
-          page_number: image_hash['page_number'],
-          width: image_hash['width'],
-          height: image_hash['height'],
-          colorspace: image_hash['colorspace'],
-          bits_per_component: image_hash['bits_per_component'],
-          is_mask: image_hash['is_mask'],
-          description: image_hash['description'],
-          ocr_result: image_hash['ocr_result'] ? Result.new(image_hash['ocr_result']) : nil
-        )
-      end
+      images_data.map { |image_hash| parse_single_image(image_hash) }
+    end
+
+    def parse_single_image(image_hash)
+      data = image_hash['data']
+      data = data.dup.force_encoding(Encoding::BINARY) if data.respond_to?(:force_encoding)
+      Image.new(
+        data: data,
+        format: image_hash['format'],
+        image_index: image_hash['image_index'],
+        page_number: image_hash['page_number'],
+        width: image_hash['width'],
+        height: image_hash['height'],
+        colorspace: image_hash['colorspace'],
+        bits_per_component: image_hash['bits_per_component'],
+        is_mask: image_hash['is_mask'],
+        description: image_hash['description'],
+        bounding_box: parse_bounding_box(image_hash['bounding_box']),
+        ocr_result: image_hash['ocr_result'] ? Result.new(image_hash['ocr_result']) : nil
+      )
     end
 
     def parse_pages(pages_data)
@@ -610,6 +619,21 @@ module Kreuzberg
       )
     end
 
+    def parse_bounding_box(bounding_box_data)
+      return nil if bounding_box_data.nil?
+
+      # If it's already a BoundingBox object, return it
+      return bounding_box_data if bounding_box_data.is_a?(BoundingBox)
+
+      # Otherwise parse from hash
+      BoundingBox.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_ocr_elements(ocr_elements_data)
       return nil if ocr_elements_data.nil?
 

data/lib/kreuzberg/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kreuzberg
-  VERSION = '4.3.4'
+  VERSION = '4.3.5'
 end

data/lib/kreuzberg.rb

@@ -21,11 +21,20 @@ module Kreuzberg
   autoload :ValidatorProtocol, 'kreuzberg/validator_protocol'
   autoload :OcrBackendProtocol, 'kreuzberg/ocr_backend_protocol'
 
+  autoload :BoundingBox, 'kreuzberg/types'
+  autoload :ElementMetadata, 'kreuzberg/types'
+  autoload :Element, 'kreuzberg/types'
   autoload :HtmlMetadata, 'kreuzberg/types'
   autoload :HeaderMetadata, 'kreuzberg/types'
   autoload :LinkMetadata, 'kreuzberg/types'
   autoload :ImageMetadata, 'kreuzberg/types'
   autoload :StructuredData, 'kreuzberg/types'
+  autoload :ExtractedKeyword, 'kreuzberg/types'
+  autoload :ProcessingWarning, 'kreuzberg/types'
+  autoload :DocumentBoundingBox, 'kreuzberg/types'
+  autoload :DocumentAnnotation, 'kreuzberg/types'
+  autoload :DocumentNode, 'kreuzberg/types'
+  autoload :DocumentStructure, 'kreuzberg/types'
 
   ExtractionConfig = Config::Extraction
   PageConfig = Config::PageConfig

data/sig/kreuzberg.rbs

@@ -677,7 +677,8 @@ module Kreuzberg
   type table_hash = {
     cells: Array[Array[String]],
     markdown: String,
-    page_number: Integer
+    page_number: Integer,
+    bounding_box: { x0: Float, y0: Float, x1: Float, y1: Float }?
   }
 
   type chunk_hash = {
@@ -703,6 +704,7 @@ module Kreuzberg
     bits_per_component: Integer?,
     is_mask: bool,
     description: String?,
+    bounding_box: { x0: Float, y0: Float, x1: Float, y1: Float }?,
     ocr_result: extraction_result_hash?
   }
 
@@ -746,8 +748,9 @@ module Kreuzberg
       attr_reader cells: Array[Array[String]]
       attr_reader markdown: String
       attr_reader page_number: Integer
+      attr_reader bounding_box: BoundingBox?
 
-      def initialize: (cells: Array[Array[String]], markdown: String, page_number: Integer) -> void
+      def initialize: (cells: Array[Array[String]], markdown: String, page_number: Integer, bounding_box: BoundingBox?) -> void
       def to_h: () -> table_hash
     end
 
@@ -789,6 +792,7 @@ module Kreuzberg
       attr_reader bits_per_component: Integer?
       attr_reader is_mask: bool
       attr_reader description: String?
+      attr_reader bounding_box: BoundingBox?
       attr_reader ocr_result: Result?
 
       def initialize: (
@@ -802,6 +806,7 @@ module Kreuzberg
         bits_per_component: Integer?,
         is_mask: bool,
         description: String?,
+        bounding_box: BoundingBox?,
         ocr_result: Result?
       ) -> void
       def to_h: () -> image_hash

data/vendor/Cargo.toml

@@ -2,7 +2,7 @@
 members = ["kreuzberg", "kreuzberg-ffi", "kreuzberg-tesseract", "kreuzberg-paddle-ocr", "kreuzberg-pdfium-render"]
 
 [workspace.package]
-version = "4.3.4"
+version = "4.3.5"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]
@@ -21,7 +21,7 @@ console_error_panic_hook = "0.1"
 criterion = { version = "0.8", features = ["html_reports"] }
 getrandom = { version = "0.4.1", features = ["wasm_js"] }
 hex = "0.4.3"
-html-to-markdown-rs = { version = "2.25.0", default-features = false }
+html-to-markdown-rs = { version = "2.25.1", default-features = false }
 image = { version = "0.25.9", default-features = false }
 js-sys = "0.3"
 libc = "0.2.182"
@@ -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 = "1.0.1"
+toml = "1.0.2"
 tracing = "0.1"
 wasm-bindgen = { version = "0.2", features = ["enable-interning"] }
 wasm-bindgen-futures = "0.4"

data/vendor/kreuzberg/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "kreuzberg"
-version = "4.3.4"
+version = "4.3.5"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]
@@ -154,7 +154,7 @@ serde = { version = "1.0.228", features = ["derive"] }
 serde_json = { version = "1.0.149" }
 serde_yaml_ng = "0.10.0"
 jotdown = "0.9"
-toml = "1.0.1"
+toml = "1.0.2"
 mime_guess = "2.0"
 rmp-serde = "1.3"
 thiserror = "2.0.18"
@@ -167,11 +167,11 @@ lopdf = { version = "0.39.0", optional = true }
 calamine = { version = "0.33.0", features = ["dates"], optional = true }
 polars = { version = "0.53.0", default-features = false, features = ["ipc"], optional = true }
 roxmltree = { version = "0.21.1", optional = true }
-zip = { version = "8.0.0", optional = true, default-features = false, features = [
+zip = { version = "8.1.0", optional = true, default-features = false, features = [
     "deflate-flate2",
 ] }
 mail-parser = { version = "0.11.2", optional = true }
-html-to-markdown-rs = { version = "2.25.0", default-features = false , features = [
+html-to-markdown-rs = { version = "2.25.1", default-features = false , features = [
     "inline-images", "metadata", ], optional = true }
 cfb = { version = "0.14.0", optional = true }
 quick-xml = { version = "0.39.1", features = ["serialize"], optional = true }
@@ -236,7 +236,7 @@ sha2 = { version = "0.10", optional = true }
 tempfile = "3.25.0"
 filetime = "0.2"
 tar = "0.4.44"
-zip = { version = "8.0.0", default-features = false, features = ["deflate-flate2"] }
+zip = { version = "8.1.0", default-features = false, features = ["deflate-flate2"] }
 serial_test = "3.3.1"
 anyhow = "1.0"
 tokio-test = "0.4"

data/vendor/kreuzberg/README.md

@@ -17,7 +17,7 @@ High-performance document intelligence library for Rust. Extract text, metadata,
 
 This is the core Rust library that powers the Python, TypeScript, and Ruby bindings.
 
-> **🚀 Version 4.3.4 Release**
+> **🚀 Version 4.3.5 Release**
 > This is a pre-release version. We invite you to test the library and [report any issues](https://github.com/kreuzberg-dev/kreuzberg/issues) you encounter.
 >
 > **Note**: The Rust crate is not currently published to crates.io for this RC. Use git dependencies or language bindings (Python, TypeScript, Ruby) instead.

data/vendor/kreuzberg/src/api/router.rs

@@ -99,7 +99,7 @@ pub fn create_router_with_limits(config: ExtractionConfig, limits: ApiSizeLimits
 /// # Examples
 ///
 /// ```no_run
-/// use kreuzberg::{ExtractionConfig, api::create_router_with_limits, core::ServerConfig};
+/// use kreuzberg::{ExtractionConfig, api::{create_router_with_limits_and_server_config, ApiSizeLimits}, core::ServerConfig};
 ///
 /// # #[tokio::main]
 /// # async fn main() -> kreuzberg::Result<()> {
@@ -108,7 +108,7 @@ pub fn create_router_with_limits(config: ExtractionConfig, limits: ApiSizeLimits
 /// server_config.cors_origins = vec!["https://example.com".to_string()];
 /// let router = create_router_with_limits_and_server_config(
 ///     extraction_config,
-///     Default::default(),
+///     ApiSizeLimits::default(),
 ///     server_config
 /// );
 /// # Ok(())

data/vendor/kreuzberg/src/chunking/core.rs

@@ -37,6 +37,8 @@ use super::validation::validate_utf8_boundaries;
 ///     overlap: 50,
 ///     trim: true,
 ///     chunker_type: ChunkerType::Text,
+///     embedding: None,
+///     preset: None,
 /// };
 /// let result = chunk_text("Long text...", &config, None)?;
 /// assert!(!result.chunks.is_empty());

data/vendor/kreuzberg/src/chunking/mod.rs

@@ -27,6 +27,8 @@
 //!     overlap: 50,
 //!     trim: true,
 //!     chunker_type: ChunkerType::Text,
+//!     embedding: None,
+//!     preset: None,
 //! };
 //!
 //! let long_text = "This is a very long document...".repeat(100);

data/vendor/kreuzberg/src/core/pipeline/format.rs

@@ -23,10 +23,12 @@ use std::borrow::Cow;
 /// * `result` - The extraction result to modify
 /// * `output_format` - The desired output format
 pub fn apply_output_format(result: &mut ExtractionResult, output_format: OutputFormat) {
-    // Check if content was already formatted during extraction
-    let already_formatted = match &*result.mime_type {
-        "text/markdown" if output_format == OutputFormat::Markdown => true,
-        "text/djot" if output_format == OutputFormat::Djot => true,
+    // Check if content was already formatted during extraction.
+    // Since extractors now preserve original MIME types, detect by checking
+    // metadata.output_format which is set by extractors that pre-format.
+    let already_formatted = match result.metadata.output_format.as_deref() {
+        Some("markdown") if output_format == OutputFormat::Markdown => true,
+        Some("djot") if output_format == OutputFormat::Djot => true,
         _ => false,
     };
 
@@ -195,7 +197,11 @@ mod tests {
 
         let mut result = ExtractionResult {
             content: "Hello World".to_string(),
-            mime_type: Cow::Borrowed("text/djot"),
+            mime_type: Cow::Borrowed("text/html"),
+            metadata: Metadata {
+                output_format: Some("djot".to_string()),
+                ..Default::default()
+            },
             djot_content: Some(DjotContent {
                 plain_text: "Hello World".to_string(),
                 blocks: vec![FormattedBlock {
@@ -322,6 +328,7 @@ mod tests {
             cells: vec![vec!["A".to_string(), "B".to_string()]],
             markdown: "| A | B |".to_string(),
             page_number: 1,
+            bounding_box: None,
         };
 
         let mut result = ExtractionResult {
@@ -367,7 +374,11 @@ mod tests {
 
         let mut result = ExtractionResult {
             content: "test".to_string(),
-            mime_type: Cow::Borrowed("text/djot"),
+            mime_type: Cow::Borrowed("text/html"),
+            metadata: Metadata {
+                output_format: Some("djot".to_string()),
+                ..Default::default()
+            },
             djot_content: Some(djot_content),
             ..Default::default()
         };

data/vendor/kreuzberg/src/core/pipeline/tests.rs

@@ -40,7 +40,13 @@ async fn test_run_pipeline_basic() {
         Cow::Borrowed(VALIDATION_MARKER_KEY),
         serde_json::json!(ORDER_VALIDATION_MARKER),
     );
-    let config = ExtractionConfig::default();
+    let config = ExtractionConfig {
+        postprocessor: Some(crate::core::config::PostProcessorConfig {
+            enabled: false,
+            ..Default::default()
+        }),
+        ..Default::default()
+    };
 
     let processed = run_pipeline(result, &config).await.unwrap();
     assert_eq!(processed.content, "test");
@@ -98,6 +104,10 @@ async fn test_pipeline_without_quality_processing() {
     };
     let config = ExtractionConfig {
         enable_quality_processing: false,
+        postprocessor: Some(crate::core::config::PostProcessorConfig {
+            enabled: false,
+            ..Default::default()
+        }),
         ..Default::default()
     };
 
@@ -166,6 +176,10 @@ async fn test_pipeline_without_chunking() {
     };
     let config = ExtractionConfig {
         chunking: None,
+        postprocessor: Some(crate::core::config::PostProcessorConfig {
+            enabled: false,
+            ..Default::default()
+        }),
         ..Default::default()
     };
 
@@ -201,7 +215,13 @@ async fn test_pipeline_preserves_metadata() {
         quality_score: None,
         processing_warnings: Vec::new(),
     };
-    let config = ExtractionConfig::default();
+    let config = ExtractionConfig {
+        postprocessor: Some(crate::core::config::PostProcessorConfig {
+            enabled: false,
+            ..Default::default()
+        }),
+        ..Default::default()
+    };
 
     let processed = run_pipeline(result, &config).await.unwrap();
     assert_eq!(
@@ -222,6 +242,7 @@ async fn test_pipeline_preserves_tables() {
         cells: vec![vec!["A".to_string(), "B".to_string()]],
         markdown: "| A | B |".to_string(),
         page_number: 0,
+        bounding_box: None,
     };
 
     let result = ExtractionResult {
@@ -242,7 +263,13 @@ async fn test_pipeline_preserves_tables() {
         quality_score: None,
         processing_warnings: Vec::new(),
     };
-    let config = ExtractionConfig::default();
+    let config = ExtractionConfig {
+        postprocessor: Some(crate::core::config::PostProcessorConfig {
+            enabled: false,
+            ..Default::default()
+        }),
+        ..Default::default()
+    };
 
     let processed = run_pipeline(result, &config).await.unwrap();
     assert_eq!(processed.tables.len(), 1);

data/vendor/kreuzberg/src/extraction/pptx/mod.rs

@@ -157,6 +157,7 @@ fn extract_pptx_from_container<R: std::io::Read + std::io::Seek>(
                     is_mask: false,
                     description: None,
                     ocr_result: None,
+                    bounding_box: None,
                 });
             }
         }

data/vendor/kreuzberg/src/extraction/transform/document_tree.rs

@@ -694,6 +694,7 @@ mod tests {
                 ],
                 markdown: "| Name | Age |\n|---|---|\n| Alice | 30 |".to_string(),
                 page_number: 1,
+                bounding_box: None,
             }],
             ..test_result("Some content")
         };

data/vendor/kreuzberg/src/extraction/transform/mod.rs

@@ -374,6 +374,7 @@ mod tests {
             ],
             markdown: "| Header1 | Header2 |\n| Cell1 | Cell2 |".to_string(),
             page_number: 1,
+            bounding_box: None,
         };
 
         let image = ExtractedImage {
@@ -388,6 +389,7 @@ mod tests {
             is_mask: false,
             description: None,
             ocr_result: None,
+            bounding_box: None,
         };
 
         let result = ExtractionResult {

data/vendor/kreuzberg/src/extractors/csv.rs

@@ -91,6 +91,7 @@ impl DocumentExtractor for CsvExtractor {
             cells: rows.clone(),
             markdown,
             page_number: 1,
+            bounding_box: None,
         };
 
         let row_count = rows.len();

data/vendor/kreuzberg/src/extractors/djot_format/parsing/table_extraction.rs

@@ -56,6 +56,7 @@ pub fn extract_tables_from_events(events: &[Event]) -> Vec<Table> {
                         cells,
                         markdown,
                         page_number: idx + 1,
+                        bounding_box: None,
                     });
                     table_index += 1;
                 }

data/vendor/kreuzberg/src/extractors/docbook.rs

@@ -242,6 +242,7 @@ fn parse_docbook_single_pass(content: &str) -> Result<DocBookParseResult> {
                                 cells: current_table.clone(),
                                 markdown,
                                 page_number: table_index + 1,
+                                bounding_box: None,
                             });
                             table_index += 1;
                             current_table.clear();