kreuzberg 4.2.10 → 4.2.11

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: abf625c4f7eedb0ba24619d640ac572192a112bf29876c25c662c8faf8a7219c
-  data.tar.gz: 460cdf492f802db89332e989340070448c5b60bb44ce0860a1104889814bb9ac
+  metadata.gz: e8a57a9be868cea7df0628ac25c7f0aabea6ced5368b0ec69452abd8ace56cd9
+  data.tar.gz: dde627c6aff4ae060d53e5b44145bdaf6cfc8af870294f7521c200cf2b7e10ba
 SHA512:
-  metadata.gz: 6e9b8b00347a73747e7ab8aad698f2d7a5798609dd1b086fe6df3a723c49bd05c5dff3c8ad0e7c83720cc3944b1a9d66fdec710405c9f1e22e43fe55387cdc92
-  data.tar.gz: dab907905f37a8fbc13d4c3e7e893cf6162fe57c6d735bfe08dfec32ea721706ab683d1baa4cfa6b0db4db8c393e7909ee3cf98195881ea31c8dc5ce0cda0b6a
+  metadata.gz: 1c322dfecd4829e4e3aa13bbdd298f3f06f62877362867cc5795cc0690ef6b632ecbbac515f20c288ae3457ab5a022cb116b4922abc69a11e06151774f6f91f0
+  data.tar.gz: 50e8a2b5489f169afb9f6b60150954463ed31988b049607c86ac501051df8138e21eec3f04d1e7b8a68f00079078416f7b3a5c2e7d93bf6bfb2bb8313c8e8aa3

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    kreuzberg (4.2.10)
+    kreuzberg (4.2.11)
       rb_sys (~> 0.9.119)
 
 GEM
@@ -123,7 +123,7 @@ GEM
       rubocop (~> 1.81)
     ruby-progressbar (1.13.0)
     securerandom (0.4.1)
-    sorbet-runtime (0.6.12914)
+    sorbet-runtime (0.6.12915)
     steep (1.10.0)
       activesupport (>= 5.1)
       concurrent-ruby (>= 1.1.10)
@@ -209,7 +209,7 @@ CHECKSUMS
   i18n (1.14.8) sha256=285778639134865c5e0f6269e0b818256017e8cde89993fdfcbfb64d088824a5
   io-console (0.8.2) sha256=d6e3ae7a7cc7574f4b8893b4fca2162e57a825b223a177b7afa236c5ef9814cc
   json (2.18.1) sha256=fe112755501b8d0466b5ada6cf50c8c3f41e897fa128ac5d263ec09eedc9f986
-  kreuzberg (4.2.10)
+  kreuzberg (4.2.11)
   language_server-protocol (3.17.0.5) sha256=fd1e39a51a28bf3eec959379985a72e296e9f9acfce46f6a79d31ca8760803cc
   lint_roller (1.1.0) sha256=2c0c845b632a7d172cb849cc90c1bce937a28c5c8ccccb50dfd46a485003cc87
   listen (3.10.0) sha256=c6e182db62143aeccc2e1960033bebe7445309c7272061979bb098d03760c9d2
@@ -244,7 +244,7 @@ CHECKSUMS
   rubocop-rspec (3.9.0) sha256=8fa70a3619408237d789aeecfb9beef40576acc855173e60939d63332fdb55e2
   ruby-progressbar (1.13.0) sha256=80fc9c47a9b640d6834e0dc7b3c94c9df37f08cb072b7761e4a71e22cff29b33
   securerandom (0.4.1) sha256=cc5193d414a4341b6e225f0cb4446aceca8e50d5e1888743fac16987638ea0b1
-  sorbet-runtime (0.6.12914) sha256=6d3c985d671dab9ab8ea244b51888b6e8e8e65e881e5bf816d1ac0950479dce6
+  sorbet-runtime (0.6.12915) sha256=21d2866b1edfe57c97d22f36db5bcf2db311f84290e56152e9faf4b4915aa315
   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=v4.2.10" alt="Go">
+    <img src="https://img.shields.io/github/v/tag/kreuzberg-dev/kreuzberg?label=Go&color=007ec6&filter=v4.2.11" 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

@@ -2755,7 +2755,7 @@ dependencies = [
 
 [[package]]
 name = "kreuzberg-rb"
-version = "4.0.0"
+version = "4.2.10"
 dependencies = [
  "async-trait",
  "html-to-markdown-rs",

data/ext/kreuzberg_rb/native/Cargo.toml

@@ -37,7 +37,7 @@ collapsible_if = "allow"
 
 [package]
 name = "kreuzberg-rb"
-version = "4.2.10"
+version = "4.2.11"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]

data/lib/kreuzberg/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kreuzberg
-  VERSION = '4.2.10'
+  VERSION = '4.2.11'
 end

data/vendor/Cargo.toml

@@ -3,7 +3,7 @@ members = ["kreuzberg", "kreuzberg-tesseract", "kreuzberg-ffi"]
 resolver = "2"
 
 [workspace.package]
-version = "4.2.10"
+version = "4.2.11"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]

data/vendor/kreuzberg/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "kreuzberg"
-version = "4.2.10"
+version = "4.2.11"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]
@@ -151,10 +151,10 @@ pdfium-render = { package = "kreuzberg-pdfium-render", version = "0.9.0", featur
     "image_latest",
 ], optional = true }
 lopdf = { version = "0.39.0", optional = true }
-calamine = { version = "0.32.0", features = ["dates"], optional = true }
+calamine = { version = "0.33.0", features = ["dates"], optional = true }
 polars = { version = "0.52.0", default-features = false, features = ["ipc"], optional = true }
 roxmltree = { version = "0.21.1", optional = true }
-zip = { version = "7.2.0", optional = true }
+zip = { version = "7.3.0", optional = true }
 mail-parser = { version = "0.11.1", optional = true }
 msg_parser = { version = "0.1.1", optional = true }
 html-to-markdown-rs = { workspace = true, features = [
@@ -218,7 +218,7 @@ smartcore = { version = "0.4", default-features = false, features = ["serde"] }
 tempfile = { workspace = true }
 filetime = "0.2"
 tar = "0.4.44"
-zip = "7.2.0"
+zip = "7.3.0"
 serial_test = "3.3.1"
 anyhow = { workspace = true }
 tokio-test = "0.4"
@@ -239,7 +239,7 @@ fastembed = { version = "5.8", default-features = false, features = [
     "ort-load-dynamic",
 ], optional = true }
 # Force ureq (transitive dep via hf-hub) to use rustls on non-Windows
-ureq = { version = "3.1", default-features = false, features = ["rustls", "json"] }
+ureq = { version = "3.2", default-features = false, features = ["rustls", "json"] }
 
 # Use native-tls on Windows to avoid aws-lc-sys CMake build issues with MinGW
 [target.'cfg(all(target_os = "windows", not(target_arch = "wasm32")))'.dependencies]
@@ -253,7 +253,7 @@ fastembed = { version = "5.8", default-features = false, features = [
     "ort-load-dynamic",
 ], optional = true }
 # Force ureq (transitive dep via hf-hub) to use native-tls on Windows
-ureq = { version = "3.1", default-features = false, features = ["native-tls", "json"] }
+ureq = { version = "3.2", default-features = false, features = ["native-tls", "json"] }
 
 [target.'cfg(target_arch = "wasm32")'.dependencies]
 wasm-bindgen-rayon = { version = "1.3", optional = true }

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 4.2.10 Release**
+> **🚀 Version 4.2.11 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/core/extractor/batch.rs

@@ -9,6 +9,7 @@ use crate::{KreuzbergError, Result};
 use std::borrow::Cow;
 use std::path::Path;
 use std::sync::Arc;
+use std::time::Instant;
 
 use super::bytes::extract_bytes;
 use super::file::extract_file;
@@ -82,10 +83,18 @@ pub async fn batch_extract_file(
 
         tasks.spawn(async move {
             let _permit = semaphore_clone.acquire().await.unwrap();
-            let result =
+            let start = Instant::now();
+            let mut result =
                 crate::core::batch_mode::with_batch_mode(async { extract_file(&path_buf, None, &config_clone).await })
                     .await;
-            (index, result)
+            let elapsed_ms = start.elapsed().as_millis() as u64;
+
+            // Add extraction timing to result metadata for benchmarking
+            if let Ok(ref mut r) = result {
+                r.metadata.extraction_duration_ms = Some(elapsed_ms);
+            }
+
+            (index, result, elapsed_ms)
         });
     }
 
@@ -93,10 +102,11 @@ pub async fn batch_extract_file(
 
     while let Some(task_result) = tasks.join_next().await {
         match task_result {
-            Ok((index, Ok(result))) => {
+            Ok((index, Ok(result), _elapsed_ms)) => {
+                // Timing already added to result.metadata.extraction_duration_ms
                 results[index] = Some(result);
             }
-            Ok((index, Err(e))) => {
+            Ok((index, Err(e), elapsed_ms)) => {
                 // All errors (including Io) should create error results
                 // instead of causing early return that abandons running tasks
                 let metadata = Metadata {
@@ -104,6 +114,7 @@ pub async fn batch_extract_file(
                         error_type: format!("{:?}", e),
                         message: e.to_string(),
                     }),
+                    extraction_duration_ms: Some(elapsed_ms),
                     ..Default::default()
                 };
 
@@ -196,11 +207,19 @@ pub async fn batch_extract_bytes(
 
         tasks.spawn(async move {
             let _permit = semaphore_clone.acquire().await.unwrap();
-            let result = crate::core::batch_mode::with_batch_mode(async {
+            let start = Instant::now();
+            let mut result = crate::core::batch_mode::with_batch_mode(async {
                 extract_bytes(&bytes, &mime_type, &config_clone).await
             })
             .await;
-            (index, result)
+            let elapsed_ms = start.elapsed().as_millis() as u64;
+
+            // Add extraction timing to result metadata for benchmarking
+            if let Ok(ref mut r) = result {
+                r.metadata.extraction_duration_ms = Some(elapsed_ms);
+            }
+
+            (index, result, elapsed_ms)
         });
     }
 
@@ -208,10 +227,11 @@ pub async fn batch_extract_bytes(
 
     while let Some(task_result) = tasks.join_next().await {
         match task_result {
-            Ok((index, Ok(result))) => {
+            Ok((index, Ok(result), _elapsed_ms)) => {
+                // Timing already added to result.metadata.extraction_duration_ms
                 results[index] = Some(result);
             }
-            Ok((index, Err(e))) => {
+            Ok((index, Err(e), elapsed_ms)) => {
                 // All errors (including Io) should create error results
                 // instead of causing early return that abandons running tasks
                 let metadata = Metadata {
@@ -219,6 +239,7 @@ pub async fn batch_extract_bytes(
                         error_type: format!("{:?}", e),
                         message: e.to_string(),
                     }),
+                    extraction_duration_ms: Some(elapsed_ms),
                     ..Default::default()
                 };
 

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

@@ -253,6 +253,7 @@ mod tests {
             image_preprocessing: None,
             json_schema: None,
             error: None,
+            extraction_duration_ms: None,
             additional: Default::default(),
         }
     }

data/vendor/kreuzberg/src/types/metadata.rs

@@ -132,6 +132,13 @@ pub struct Metadata {
     #[serde(skip_serializing_if = "Option::is_none")]
     pub error: Option<ErrorMetadata>,
 
+    /// Extraction duration in milliseconds (for benchmarking).
+    ///
+    /// This field is populated by batch extraction to provide per-file timing
+    /// information. It's `None` for single-file extraction (which uses external timing).
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub extraction_duration_ms: Option<u64>,
+
     /// Additional custom fields from postprocessors.
     ///
     /// This flattened map allows Python/TypeScript postprocessors to add

data/vendor/kreuzberg/tests/api_embed.rs

@@ -83,6 +83,7 @@ async fn test_embed_empty_texts() {
 
 /// Test embed endpoint with custom embedding configuration.
 #[tokio::test]
+#[cfg_attr(target_arch = "aarch64", ignore = "ONNX Runtime model loading unstable on ARM")]
 async fn test_embed_with_custom_config() {
     let app = create_router(ExtractionConfig::default());
 
@@ -125,6 +126,7 @@ async fn test_embed_with_custom_config() {
 
 /// Test embed endpoint with single text.
 #[tokio::test]
+#[cfg_attr(target_arch = "aarch64", ignore = "ONNX Runtime model loading unstable on ARM")]
 async fn test_embed_single_text() {
     let app = create_router(ExtractionConfig::default());
 
@@ -201,6 +203,7 @@ async fn test_embed_batch() {
 
 /// Test embed endpoint with long text.
 #[tokio::test]
+#[cfg_attr(target_arch = "aarch64", ignore = "ONNX Runtime model loading unstable on ARM")]
 async fn test_embed_long_text() {
     let app = create_router(ExtractionConfig::default());
 
@@ -317,6 +320,7 @@ async fn test_embed_rejects_simple_json_array() {
 
 /// Test embed endpoint preserves embedding vector values across calls.
 #[tokio::test]
+#[cfg_attr(target_arch = "aarch64", ignore = "ONNX Runtime model loading unstable on ARM")]
 async fn test_embed_deterministic() {
     let app = create_router(ExtractionConfig::default());
 
@@ -376,6 +380,7 @@ async fn test_embed_deterministic() {
 
 /// Test embed endpoint with different embedding presets.
 #[tokio::test]
+#[cfg_attr(target_arch = "aarch64", ignore = "ONNX Runtime model loading unstable on ARM")]
 async fn test_embed_different_presets() {
     let app = create_router(ExtractionConfig::default());
 

data/vendor/kreuzberg-tesseract/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "kreuzberg-tesseract"
-version = "4.2.10"
+version = "4.2.11"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]
@@ -26,7 +26,7 @@ image = { workspace = true, features = ["png"] }
 [build-dependencies]
 cc = { version = "^1.2.55", optional = true }
 cmake = { version = "0.1.57", optional = true }
-zip = { version = "7.2.0", optional = true }
+zip = { version = "7.3.0", optional = true }
 
 # Use native-tls on Windows to avoid aws-lc-sys CMake build issues with MinGW
 [target.'cfg(target_os = "windows")'.build-dependencies]

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: kreuzberg
 version: !ruby/object:Gem::Version
-  version: 4.2.10
+  version: 4.2.11
 platform: ruby
 authors:
 - Na'aman Hirschfeld
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-02-05 00:00:00.000000000 Z
+date: 2026-02-06 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rb_sys