kreuzberg 4.2.11 → 4.2.13

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e8a57a9be868cea7df0628ac25c7f0aabea6ced5368b0ec69452abd8ace56cd9
-  data.tar.gz: dde627c6aff4ae060d53e5b44145bdaf6cfc8af870294f7521c200cf2b7e10ba
+  metadata.gz: 3c2053b10256948a215ff0d3552894991e801497ac4b2480eca3c98bb645cc27
+  data.tar.gz: 324b6147e172ecedb2338fab1b14ce2022a8b9c2d6be7fd86ac0f862d81ef7ce
 SHA512:
-  metadata.gz: 1c322dfecd4829e4e3aa13bbdd298f3f06f62877362867cc5795cc0690ef6b632ecbbac515f20c288ae3457ab5a022cb116b4922abc69a11e06151774f6f91f0
-  data.tar.gz: 50e8a2b5489f169afb9f6b60150954463ed31988b049607c86ac501051df8138e21eec3f04d1e7b8a68f00079078416f7b3a5c2e7d93bf6bfb2bb8313c8e8aa3
+  metadata.gz: 84a6636111d240c99eb17546f80c1df31117c700d78282c18a67a79aa613021d33988cbc1b00d5bc62bb2ffeef8c2a8f1759e137329de8f30af7f61b6db1a55b
+  data.tar.gz: 7628ecce3c6fb44c06a9546f2db696ae3486de35e0a05195cbea752bc6f78e573162e6305aec8c8ae0ca0fdbb6709a3e75752b822dbd8aed637eff9577c3e020

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    kreuzberg (4.2.11)
+    kreuzberg (4.2.13)
       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 (0.6.12915)
+    sorbet-runtime (0.6.12925)
     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 (4.2.11)
+  kreuzberg (4.2.13)
   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 (0.6.12915) sha256=21d2866b1edfe57c97d22f36db5bcf2db311f84290e56152e9faf4b4915aa315
+  sorbet-runtime (0.6.12925) sha256=ddd6fb1d8aaf6bc19119ffadbc4b96536f3d6766fa82059112dacb90977c6eca
   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=v4.2.11" alt="Go">
+    <img src="https://img.shields.io/github/v/tag/kreuzberg-dev/kreuzberg?label=Go&color=007ec6&filter=v4.2.13" 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 = "4.2.11"
+version = "4.2.13"
 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 = '4.2.11'
+  VERSION = '4.2.13'
 end

data/vendor/Cargo.toml

@@ -3,7 +3,7 @@ members = ["kreuzberg", "kreuzberg-tesseract", "kreuzberg-ffi"]
 resolver = "2"
 
 [workspace.package]
-version = "4.2.11"
+version = "4.2.13"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]
@@ -49,7 +49,7 @@ toml = "0.9.11"
 num_cpus = "1.17.0"
 once_cell = "1.21.3"
 html-to-markdown-rs = { version = "2.24.5", default-features = false }
-reqwest = { version = "0.13.1", default-features = false, features = ["json", "rustls"] }
+reqwest = { version = "0.13.2", default-features = false, features = ["json", "rustls"] }
 image = { version = "0.25.9", default-features = false }
 lzma-rust2 = { version = "0.15.7" }
 

data/vendor/kreuzberg/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "kreuzberg"
-version = "4.2.11"
+version = "4.2.13"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]
@@ -35,22 +35,21 @@ excel = ["dep:calamine", "dep:polars", "tokio-runtime"]
 office = [
     "dep:roxmltree",
     "dep:zip",
-    "dep:docx-lite",
     "dep:quick-xml",
     "dep:pulldown-cmark",
     "dep:biblatex",
+    "dep:biblib",
     "dep:org",
     "dep:rtf-parser",
     "dep:rst_parser",
     "dep:fb2",
     "dep:typst-syntax",
     "html",
-    "tokio-runtime",
 ]
 email = ["dep:mail-parser", "dep:msg_parser"]
 html = ["dep:html-to-markdown-rs"]
 xml = ["dep:quick-xml", "dep:roxmltree"]
-archives = ["dep:zip", "dep:tar", "dep:sevenz-rust2", "dep:lzma-rust2"]
+archives = ["dep:zip", "dep:tar", "dep:sevenz-rust2", "dep:lzma-rust2", "dep:flate2"]
 
 ocr = [
     "dep:kreuzberg-tesseract",
@@ -59,6 +58,8 @@ ocr = [
     "dep:fast_image_resize",
     "dep:ndarray",
     "dep:kamadak-exif",
+    "dep:hayro-jpeg2000",
+    "dep:hayro-jbig2",
     "html",
 ]
 language-detection = ["dep:whatlang"]
@@ -77,7 +78,7 @@ mcp-http = ["mcp", "api"]
 
 otel = ["dep:opentelemetry", "dep:opentelemetry_sdk", "dep:tracing-opentelemetry"]
 
-wasm-target = ["pdf", "html", "xml", "email", "language-detection", "chunking", "quality"]
+wasm-target = ["pdf", "html", "xml", "email", "language-detection", "chunking", "quality", "office"]
 wasm-threads = ["dep:wasm-bindgen-rayon"]
 
 full = [
@@ -127,7 +128,7 @@ simdutf8 = { version = "0.1", optional = true }
 hex = { workspace = true }
 lazy_static = "1.5.0"
 libc = { workspace = true }
-memchr = "2.7.6"
+memchr = "2.8.0"
 num_cpus = { workspace = true }
 once_cell = { workspace = true }
 parking_lot = { workspace = true }
@@ -154,7 +155,9 @@ lopdf = { version = "0.39.0", 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 = "7.3.0", optional = true }
+zip = { version = "7.4.0", optional = true, default-features = false, features = [
+    "deflate-flate2",
+] }
 mail-parser = { version = "0.11.1", optional = true }
 msg_parser = { version = "0.1.1", optional = true }
 html-to-markdown-rs = { workspace = true, features = [
@@ -165,10 +168,16 @@ quick-xml = { version = "0.39.0", features = ["serialize"], optional = true }
 tar = { version = "0.4.44", optional = true }
 sevenz-rust2 = { version = "0.20.1", optional = true }
 lzma-rust2 = { workspace = true, optional = true }
-docx-lite = { version = "0.2.0", optional = true }
+flate2 = { version = "1.0", optional = true }
 
 pulldown-cmark = { version = "0.13", optional = true }
 biblatex = { version = "0.11", optional = true }
+biblib = { version = "0.3", default-features = false, features = [
+    "ris",
+    "pubmed",
+    "xml",
+    "regex",
+], optional = true }
 org = { version = "0.3", optional = true }
 rtf-parser = { version = "0.4", optional = true }
 rst_parser = { version = "0.4", optional = true }
@@ -183,12 +192,18 @@ image = { workspace = true, default-features = false, features = [
     "bmp",
     "tiff",
     "gif",
+    "pnm",
     "rayon",
 ], optional = true }
 tiff = { version = "0.11", optional = true }
 fast_image_resize = { version = "6.0.0", optional = true }
 ndarray = { version = "0.17.2", optional = true }
 kamadak-exif = { version = "0.6.1", optional = true }
+hayro-jpeg2000 = { version = "0.3", default-features = false, features = [
+    "std",
+    "simd",
+], optional = true }
+hayro-jbig2 = { version = "0.1", default-features = false, features = ["std"], optional = true }
 whatlang = { version = "0.18.0", optional = true }
 text-splitter = { version = "0.29.3", features = ["markdown"], optional = true }
 unicode-normalization = { version = "0.1.25", optional = true }
@@ -218,7 +233,7 @@ smartcore = { version = "0.4", default-features = false, features = ["serde"] }
 tempfile = { workspace = true }
 filetime = "0.2"
 tar = "0.4.44"
-zip = "7.3.0"
+zip = { version = "7.4.0", default-features = false, features = ["deflate-flate2"] }
 serial_test = "3.3.1"
 anyhow = { workspace = true }
 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 4.2.11 Release**
+> **🚀 Version 4.2.13 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/config/extraction/core.rs

@@ -102,6 +102,15 @@ pub struct ExtractionConfig {
     #[serde(default)]
     pub result_format: crate::types::OutputFormat,
 
+    /// Security limits for archive extraction.
+    ///
+    /// Controls maximum archive size, compression ratio, file count, and other
+    /// security thresholds to prevent decompression bomb attacks.
+    /// When `None`, default limits are used (500MB archive, 100:1 ratio, 10K files).
+    #[cfg(feature = "archives")]
+    #[serde(default)]
+    pub security_limits: Option<crate::extractors::security::SecurityLimits>,
+
     /// Content text format (default: Plain).
     ///
     /// Controls the format of the extracted content:
@@ -137,6 +146,8 @@ impl Default for ExtractionConfig {
             #[cfg(feature = "html")]
             html_options: None,
             max_concurrent_extractions: None,
+            #[cfg(feature = "archives")]
+            security_limits: None,
             result_format: crate::types::OutputFormat::Unified,
             output_format: OutputFormat::Plain,
         }

data/vendor/kreuzberg/src/core/extractor/bytes.rs

@@ -5,16 +5,16 @@
 //! - Legacy format conversion (DOC, PPT)
 //! - Extraction pipeline orchestration
 
-#[cfg(not(feature = "office"))]
+#[cfg(not(all(feature = "office", not(target_arch = "wasm32"))))]
 use crate::KreuzbergError;
 use crate::Result;
 use crate::core::config::ExtractionConfig;
 use crate::core::mime::{LEGACY_POWERPOINT_MIME_TYPE, LEGACY_WORD_MIME_TYPE};
-#[cfg(feature = "office")]
+#[cfg(all(feature = "office", not(target_arch = "wasm32")))]
 use crate::extraction::libreoffice::{convert_doc_to_docx, convert_ppt_to_pptx};
 use crate::types::ExtractionResult;
 
-#[cfg(feature = "office")]
+#[cfg(all(feature = "office", not(target_arch = "wasm32")))]
 use super::file::apply_libreoffice_metadata;
 use super::file::extract_bytes_with_extractor;
 #[cfg(feature = "otel")]
@@ -72,7 +72,7 @@ pub async fn extract_bytes(content: &[u8], mime_type: &str, config: &ExtractionC
         let validated_mime = mime::validate_mime_type(mime_type)?;
 
         match validated_mime.as_str() {
-            #[cfg(feature = "office")]
+            #[cfg(all(feature = "office", not(target_arch = "wasm32")))]
             LEGACY_WORD_MIME_TYPE => {
                 let conversion = convert_doc_to_docx(content).await?;
                 let mut result =
@@ -80,13 +80,13 @@ pub async fn extract_bytes(content: &[u8], mime_type: &str, config: &ExtractionC
                 apply_libreoffice_metadata(&mut result, LEGACY_WORD_MIME_TYPE, &conversion);
                 return Ok(result);
             }
-            #[cfg(not(feature = "office"))]
+            #[cfg(not(all(feature = "office", not(target_arch = "wasm32"))))]
             LEGACY_WORD_MIME_TYPE => {
                 return Err(KreuzbergError::UnsupportedFormat(
                     "Legacy Word conversion requires the `office` feature or LibreOffice support".to_string(),
                 ));
             }
-            #[cfg(feature = "office")]
+            #[cfg(all(feature = "office", not(target_arch = "wasm32")))]
             LEGACY_POWERPOINT_MIME_TYPE => {
                 let conversion = convert_ppt_to_pptx(content).await?;
                 let mut result =
@@ -94,7 +94,7 @@ pub async fn extract_bytes(content: &[u8], mime_type: &str, config: &ExtractionC
                 apply_libreoffice_metadata(&mut result, LEGACY_POWERPOINT_MIME_TYPE, &conversion);
                 return Ok(result);
             }
-            #[cfg(not(feature = "office"))]
+            #[cfg(not(all(feature = "office", not(target_arch = "wasm32"))))]
             LEGACY_POWERPOINT_MIME_TYPE => {
                 return Err(KreuzbergError::UnsupportedFormat(
                     "Legacy PowerPoint conversion requires the `office` feature or LibreOffice support".to_string(),

data/vendor/kreuzberg/src/core/extractor/file.rs

@@ -6,23 +6,23 @@
 //! - File validation and reading
 //! - Extraction pipeline orchestration
 
-#[cfg(any(feature = "otel", not(feature = "office")))]
+#[cfg(any(feature = "otel", not(all(feature = "office", not(target_arch = "wasm32")))))]
 use crate::KreuzbergError;
 use crate::Result;
 use crate::core::config::ExtractionConfig;
 use crate::core::mime::{LEGACY_POWERPOINT_MIME_TYPE, LEGACY_WORD_MIME_TYPE};
-#[cfg(feature = "office")]
+#[cfg(all(feature = "office", not(target_arch = "wasm32")))]
 use crate::extraction::libreoffice::{convert_doc_to_docx, convert_ppt_to_pptx};
 use crate::types::ExtractionResult;
-#[cfg(feature = "office")]
+#[cfg(all(feature = "office", not(target_arch = "wasm32")))]
 use crate::types::LibreOfficeConversionResult;
-#[cfg(feature = "office")]
+#[cfg(all(feature = "office", not(target_arch = "wasm32")))]
 use serde_json::json;
-#[cfg(feature = "office")]
+#[cfg(all(feature = "office", not(target_arch = "wasm32")))]
 use std::borrow::Cow;
 use std::path::Path;
 
-#[cfg(feature = "office")]
+#[cfg(all(feature = "office", not(target_arch = "wasm32")))]
 use super::helpers::pool_mime_type;
 
 use super::helpers::get_extractor;
@@ -151,7 +151,7 @@ pub async fn extract_file(
         let detected_mime = mime::detect_or_validate(Some(path), mime_type)?;
 
         match detected_mime.as_str() {
-            #[cfg(feature = "office")]
+            #[cfg(all(feature = "office", not(target_arch = "wasm32")))]
             LEGACY_WORD_MIME_TYPE => {
                 let original_bytes = tokio::fs::read(path).await?;
                 let conversion = convert_doc_to_docx(&original_bytes).await?;
@@ -160,13 +160,13 @@ pub async fn extract_file(
                 apply_libreoffice_metadata(&mut result, LEGACY_WORD_MIME_TYPE, &conversion);
                 return Ok(result);
             }
-            #[cfg(not(feature = "office"))]
+            #[cfg(not(all(feature = "office", not(target_arch = "wasm32"))))]
             LEGACY_WORD_MIME_TYPE => {
                 return Err(KreuzbergError::UnsupportedFormat(
                     "Legacy Word conversion requires the `office` feature or LibreOffice support".to_string(),
                 ));
             }
-            #[cfg(feature = "office")]
+            #[cfg(all(feature = "office", not(target_arch = "wasm32")))]
             LEGACY_POWERPOINT_MIME_TYPE => {
                 let original_bytes = tokio::fs::read(path).await?;
                 let conversion = convert_ppt_to_pptx(&original_bytes).await?;
@@ -175,7 +175,7 @@ pub async fn extract_file(
                 apply_libreoffice_metadata(&mut result, LEGACY_POWERPOINT_MIME_TYPE, &conversion);
                 return Ok(result);
             }
-            #[cfg(not(feature = "office"))]
+            #[cfg(not(all(feature = "office", not(target_arch = "wasm32"))))]
             LEGACY_POWERPOINT_MIME_TYPE => {
                 return Err(KreuzbergError::UnsupportedFormat(
                     "Legacy PowerPoint conversion requires the `office` feature or LibreOffice support".to_string(),
@@ -222,7 +222,7 @@ pub(in crate::core::extractor) async fn extract_bytes_with_extractor(
     Ok(result)
 }
 
-#[cfg(feature = "office")]
+#[cfg(all(feature = "office", not(target_arch = "wasm32")))]
 pub(in crate::core::extractor) fn apply_libreoffice_metadata(
     result: &mut ExtractionResult,
     legacy_mime: &str,

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

@@ -80,6 +80,10 @@ static EXT_TO_MIME: Lazy<HashMap<&'static str, &'static str>> = Lazy::new(|| {
     m.insert("jpx", "image/jpx");
     m.insert("jpm", "image/jpm");
     m.insert("mj2", "image/mj2");
+    m.insert("j2k", "image/jp2");
+    m.insert("j2c", "image/jp2");
+    m.insert("jbig2", "image/x-jbig2");
+    m.insert("jb2", "image/x-jbig2");
     m.insert("pnm", "image/x-portable-anymap");
     m.insert("pbm", "image/x-portable-bitmap");
     m.insert("pgm", "image/x-portable-graymap");
@@ -108,10 +112,18 @@ static EXT_TO_MIME: Lazy<HashMap<&'static str, &'static str>> = Lazy::new(|| {
     m.insert("epub", "application/epub+zip");
     m.insert("rtf", "application/rtf");
     m.insert("bib", "application/x-bibtex");
+    m.insert("ris", "application/x-research-info-systems");
+    m.insert("nbib", "application/x-pubmed");
+    m.insert("enw", "application/x-endnote+xml");
+    m.insert("fb2", "application/x-fictionbook+xml");
+    m.insert("opml", "application/xml+opml");
+    m.insert("dbk", "application/docbook+xml");
     m.insert("ipynb", "application/x-ipynb+json");
     m.insert("tex", "application/x-latex");
     m.insert("latex", "application/x-latex");
     m.insert("typst", "application/x-typst");
+    m.insert("typ", "application/x-typst");
+    m.insert("djot", "text/x-djot");
     m.insert("commonmark", "text/x-commonmark");
 
     m
@@ -137,6 +149,7 @@ static SUPPORTED_MIME_TYPES: Lazy<HashSet<&'static str>> = Lazy::new(|| {
     set.insert("image/tiff");
     set.insert("image/webp");
     set.insert("image/x-bmp");
+    set.insert("image/x-jbig2");
     set.insert("image/x-ms-bmp");
     set.insert("image/x-portable-anymap");
     set.insert("image/x-portable-bitmap");
@@ -146,20 +159,25 @@ static SUPPORTED_MIME_TYPES: Lazy<HashSet<&'static str>> = Lazy::new(|| {
 
     set.insert("application/csl+json");
     set.insert("application/docbook+xml");
+    set.insert("text/docbook");
     set.insert("application/epub+zip");
     set.insert("application/rtf");
     set.insert("application/vnd.oasis.opendocument.text");
     set.insert(DOCX_MIME_TYPE);
     set.insert("application/x-biblatex");
     set.insert("application/x-bibtex");
+    set.insert("text/x-bibtex");
     set.insert("application/x-endnote+xml");
     set.insert("application/x-fictionbook+xml");
+    set.insert("application/x-fictionbook");
+    set.insert("text/x-fictionbook");
     set.insert("application/x-ipynb+json");
     set.insert("application/x-jats+xml");
     set.insert("application/x-latex");
     set.insert("application/xml+opml");
     set.insert("application/x-opml+xml");
     set.insert("application/x-research-info-systems");
+    set.insert("application/x-pubmed");
     set.insert("application/x-typst");
     set.insert("text/csv");
     set.insert("text/tab-separated-values");
@@ -210,8 +228,26 @@ static SUPPORTED_MIME_TYPES: Lazy<HashSet<&'static str>> = Lazy::new(|| {
     set.insert("application/tar");
     set.insert("application/x-gtar");
     set.insert("application/x-ustar");
+    set.insert("application/gzip");
+    set.insert("application/x-gzip");
     set.insert("application/x-7z-compressed");
 
+    set.insert("text/djot");
+    set.insert("text/x-djot");
+
+    // Additional extractor-supported MIME types that must stay in sync
+    set.insert("text/jats");
+    set.insert("application/x-epub+zip");
+    set.insert("application/vnd.epub+zip");
+    set.insert("text/rtf");
+    set.insert("text/prs.fallenstein.rst");
+    set.insert("text/x-tex");
+    set.insert("text/org");
+    set.insert("application/x-org");
+    set.insert("application/xhtml+xml");
+    set.insert("text/x-typst");
+    set.insert("image/jpg");
+
     set
 });
 
@@ -291,6 +327,15 @@ pub fn validate_mime_type(mime_type: &str) -> Result<String> {
         return Ok(mime_type.to_string());
     }
 
+    // Case-insensitive fallback: MIME types are case-insensitive per RFC 2045.
+    // This handles common mismatches like "macroEnabled" vs "macroenabled".
+    let lower = mime_type.to_ascii_lowercase();
+    for supported in SUPPORTED_MIME_TYPES.iter() {
+        if supported.to_ascii_lowercase() == lower {
+            return Ok(supported.to_string());
+        }
+    }
+
     Err(KreuzbergError::UnsupportedFormat(mime_type.to_string()))
 }
 
@@ -621,8 +666,8 @@ mod tests {
         let file_path = dir.path().join("testfile");
         File::create(&file_path).unwrap();
 
-        let result = detect_mime_type(&file_path, true);
-        assert!(result.is_err() || result.is_ok());
+        let _result = detect_mime_type(&file_path, true);
+        // Files without extensions may or may not be detected via mime_guess fallback
     }
 
     #[test]

data/vendor/kreuzberg/src/extraction/archive/gzip.rs

@@ -0,0 +1,129 @@
+//! Gzip decompression and extraction.
+//!
+//! Provides functions for decompressing gzip files and extracting
+//! metadata and text content from the compressed data.
+
+use super::{ArchiveEntry, ArchiveMetadata};
+use crate::error::{KreuzbergError, Result};
+use crate::extractors::security::SecurityLimits;
+use flate2::read::GzDecoder;
+use std::collections::HashMap;
+use std::io::Read;
+
+/// Decompress gzip bytes with a size limit to prevent decompression bombs.
+fn decompress_gzip_limited(bytes: &[u8], max_size: u64) -> Result<Vec<u8>> {
+    let decoder = GzDecoder::new(bytes);
+    let mut limited = decoder.take(max_size + 1);
+    let mut decompressed = Vec::new();
+    limited
+        .read_to_end(&mut decompressed)
+        .map_err(|e| KreuzbergError::parsing(format!("Failed to decompress gzip: {}", e)))?;
+
+    if decompressed.len() as u64 > max_size {
+        return Err(KreuzbergError::validation(format!(
+            "Gzip decompressed size exceeds {} byte limit",
+            max_size
+        )));
+    }
+
+    Ok(decompressed)
+}
+
+/// Decompress gzip bytes, returning the raw decompressed data.
+pub fn decompress_gzip(bytes: &[u8], limits: &SecurityLimits) -> Result<Vec<u8>> {
+    decompress_gzip_limited(bytes, limits.max_archive_size as u64)
+}
+
+/// Extract both metadata and text content from gzip in a single decompression pass.
+///
+/// This avoids the overhead of decompressing the data multiple times when both
+/// metadata and text content are needed.
+pub fn extract_gzip(bytes: &[u8], limits: &SecurityLimits) -> Result<(ArchiveMetadata, HashMap<String, String>)> {
+    let decompressed = decompress_gzip_limited(bytes, limits.max_archive_size as u64)?;
+
+    // Re-read header for filename (lightweight - no decompression)
+    let mut decoder = GzDecoder::new(bytes);
+    let mut _discard = [0u8; 1];
+    let _ = decoder.read(&mut _discard); // trigger header read
+    let filename = decoder
+        .header()
+        .and_then(|h| h.filename())
+        .and_then(|f| std::str::from_utf8(f).ok())
+        .unwrap_or("compressed_content")
+        .to_string();
+
+    let size = decompressed.len() as u64;
+
+    let metadata = ArchiveMetadata {
+        format: "GZIP".to_string(),
+        file_list: vec![ArchiveEntry {
+            path: filename.clone(),
+            size,
+            is_dir: false,
+        }],
+        file_count: 1,
+        total_size: size,
+    };
+
+    let mut contents = HashMap::new();
+    if let Ok(text) = String::from_utf8(decompressed) {
+        contents.insert(filename, text);
+    }
+
+    Ok((metadata, contents))
+}
+
+/// Extract metadata from a gzip-compressed file.
+///
+/// Gzip wraps a single stream, so the metadata contains one entry
+/// with the original filename (from gzip header) and decompressed size.
+pub fn extract_gzip_metadata(bytes: &[u8], limits: &SecurityLimits) -> Result<ArchiveMetadata> {
+    let decompressed = decompress_gzip_limited(bytes, limits.max_archive_size as u64)?;
+
+    let mut decoder = GzDecoder::new(bytes);
+    let mut _discard = [0u8; 1];
+    let _ = decoder.read(&mut _discard);
+    let filename = decoder
+        .header()
+        .and_then(|h| h.filename())
+        .and_then(|f| std::str::from_utf8(f).ok())
+        .unwrap_or("compressed_content")
+        .to_string();
+
+    let size = decompressed.len() as u64;
+
+    Ok(ArchiveMetadata {
+        format: "GZIP".to_string(),
+        file_list: vec![ArchiveEntry {
+            path: filename,
+            size,
+            is_dir: false,
+        }],
+        file_count: 1,
+        total_size: size,
+    })
+}
+
+/// Extract text content from a gzip-compressed file.
+///
+/// Decompresses and attempts to read the result as UTF-8 text.
+pub fn extract_gzip_text_content(bytes: &[u8], limits: &SecurityLimits) -> Result<HashMap<String, String>> {
+    let decompressed = decompress_gzip_limited(bytes, limits.max_archive_size as u64)?;
+
+    let mut decoder = GzDecoder::new(bytes);
+    let mut _discard = [0u8; 1];
+    let _ = decoder.read(&mut _discard);
+    let filename = decoder
+        .header()
+        .and_then(|h| h.filename())
+        .and_then(|f| std::str::from_utf8(f).ok())
+        .unwrap_or("compressed_content")
+        .to_string();
+
+    let mut contents = HashMap::new();
+    if let Ok(text) = String::from_utf8(decompressed) {
+        contents.insert(filename, text);
+    }
+
+    Ok(contents)
+}