kreuzberg 4.0.7 → 4.0.8

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 58ca4c0e55e9de4e3a17d748de7dc9210f14ebf2658252fc265799d092928fd9
-  data.tar.gz: 55bcb1ed72c046c1f49707fc51ce55bf8346b24614a0c7c8a5ccb976d6351621
+  metadata.gz: f5742beaf511e059ff6d34f536e90a5636097b48cd1262a34eaa7692261de7aa
+  data.tar.gz: b913234c3ecda47e6e4944181309c2e60c09a8e708eb651cd705258048d3f7bb
 SHA512:
-  metadata.gz: 0f76ed348b4eb214557e526b8e2d94cd0267da71cd8e1a095fee125d47cfc6868077d03c41fe58654d4fd9978376ba1641bb83f3aed2b4c1a105e21ee2f8e73e
-  data.tar.gz: 4c6e77e98a5be13a3ccf5f38e859f13ead0438019bf413d94f17938b4f47d83c2d7378181a6561a7db76eb284289847f4cc54b0d64c8ddf1ccd9f2b54eb8b880
+  metadata.gz: 1a459be58b389e806f6c46ffadaf0f01ea095154c3f9265b54bcd1c722493d8a094df41cc5a008bea6ec3cadc2622ecd0379f36a0ae7404f93e2a6c39f44ebfd
+  data.tar.gz: 31c2d8cb14a9bd72e7fc4fff84f598ec6ecb24c5d982394f642fcc04e09cc74d4280550e10ca1b5d703a0a824fc3c63330d5608c42785a825ea4f46ee3ad8c0a

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    kreuzberg (4.0.7)
+    kreuzberg (4.0.8)
 
 GEM
   remote: https://rubygems.org/
@@ -22,7 +22,8 @@ GEM
     ast (2.4.3)
     base64 (0.3.0)
     bigdecimal (4.0.1)
-    byebug (12.0.0)
+    byebug (13.0.0)
+      reline (>= 0.6.0)
     coderay (1.1.3)
     concurrent-ruby (1.3.6)
     connection_pool (3.0.2)
@@ -43,10 +44,12 @@ GEM
     fileutils (1.8.0)
     i18n (1.14.8)
       concurrent-ruby (~> 1.0)
+    io-console (0.8.2)
     json (2.18.0)
     language_server-protocol (3.17.0.5)
     lint_roller (1.1.0)
-    listen (3.9.0)
+    listen (3.10.0)
+      logger
       rb-fsevent (~> 0.10, >= 0.10.3)
       rb-inotify (~> 0.9, >= 0.9.10)
     logger (1.7.0)
@@ -55,16 +58,17 @@ GEM
       prism (~> 1.5)
     mutex_m (0.3.0)
     parallel (1.27.0)
-    parser (3.3.10.0)
+    parser (3.3.10.1)
       ast (~> 2.4.1)
       racc
     prism (1.8.0)
-    pry (0.15.2)
+    pry (0.16.0)
       coderay (~> 1.1)
       method_source (~> 1.0)
-    pry-byebug (3.11.0)
-      byebug (~> 12.0)
-      pry (>= 0.13, < 0.16)
+      reline (>= 0.6.0)
+    pry-byebug (3.12.0)
+      byebug (~> 13.0)
+      pry (>= 0.13, < 0.17)
     racc (1.8.1)
     rainbow (3.1.1)
     rake (13.3.1)
@@ -79,6 +83,8 @@ GEM
     rbs (3.10.2)
       logger
     regexp_parser (2.11.3)
+    reline (0.6.3)
+      io-console (~> 0.5)
     rspec (3.13.2)
       rspec-core (~> 3.13.0)
       rspec-expectations (~> 3.13.0)
@@ -115,7 +121,7 @@ GEM
       rubocop (~> 1.81)
     ruby-progressbar (1.13.0)
     securerandom (0.4.1)
-    sorbet-runtime (0.6.12887)
+    sorbet-runtime (0.6.12894)
     steep (1.10.0)
       activesupport (>= 5.1)
       concurrent-ruby (>= 1.1.10)
@@ -179,7 +185,7 @@ CHECKSUMS
   ast (2.4.3) sha256=954615157c1d6a382bc27d690d973195e79db7f55e9765ac7c481c60bdb4d383
   base64 (0.3.0) sha256=27337aeabad6ffae05c265c450490628ef3ebd4b67be58257393227588f5a97b
   bigdecimal (4.0.1) sha256=8b07d3d065a9f921c80ceaea7c9d4ae596697295b584c296fe599dd0ad01c4a7
-  byebug (12.0.0) sha256=d4a150d291cca40b66ec9ca31f754e93fed8aa266a17335f71bb0afa7fca1a1e
+  byebug (13.0.0) sha256=d2263efe751941ca520fa29744b71972d39cbc41839496706f5d9b22e92ae05d
   coderay (1.1.3) sha256=dc530018a4684512f8f38143cd2a096c9f02a1fc2459edcfe534787a7fc77d4b
   concurrent-ruby (1.3.6) sha256=6b56837e1e7e5292f9864f34b69c5a2cbc75c0cf5338f1ce9903d10fa762d5ab
   connection_pool (3.0.2) sha256=33fff5ba71a12d2aa26cb72b1db8bba2a1a01823559fb01d29eb74c286e62e0a
@@ -199,20 +205,21 @@ CHECKSUMS
   ffi (1.17.3-x86_64-linux-musl) sha256=086b221c3a68320b7564066f46fed23449a44f7a1935f1fe5a245bd89d9aea56
   fileutils (1.8.0) sha256=8c6b1df54e2540bdb2f39258f08af78853aa70bad52b4d394bbc6424593c6e02
   i18n (1.14.8) sha256=285778639134865c5e0f6269e0b818256017e8cde89993fdfcbfb64d088824a5
+  io-console (0.8.2) sha256=d6e3ae7a7cc7574f4b8893b4fca2162e57a825b223a177b7afa236c5ef9814cc
   json (2.18.0) sha256=b10506aee4183f5cf49e0efc48073d7b75843ce3782c68dbeb763351c08fd505
-  kreuzberg (4.0.7)
+  kreuzberg (4.0.8)
   language_server-protocol (3.17.0.5) sha256=fd1e39a51a28bf3eec959379985a72e296e9f9acfce46f6a79d31ca8760803cc
   lint_roller (1.1.0) sha256=2c0c845b632a7d172cb849cc90c1bce937a28c5c8ccccb50dfd46a485003cc87
-  listen (3.9.0) sha256=db9e4424e0e5834480385197c139cb6b0ae0ef28cc13310cfd1ca78377d59c67
+  listen (3.10.0) sha256=c6e182db62143aeccc2e1960033bebe7445309c7272061979bb098d03760c9d2
   logger (1.7.0) sha256=196edec7cc44b66cfb40f9755ce11b392f21f7967696af15d274dde7edff0203
   method_source (1.1.0) sha256=181301c9c45b731b4769bc81e8860e72f9161ad7d66dd99103c9ab84f560f5c5
   minitest (6.0.1) sha256=7854c74f48e2e975969062833adc4013f249a4b212f5e7b9d5c040bf838d54bb
   mutex_m (0.3.0) sha256=cfcb04ac16b69c4813777022fdceda24e9f798e48092a2b817eb4c0a782b0751
   parallel (1.27.0) sha256=4ac151e1806b755fb4e2dc2332cbf0e54f2e24ba821ff2d3dcf86bf6dc4ae130
-  parser (3.3.10.0) sha256=ce3587fa5cc55a88c4ba5b2b37621b3329aadf5728f9eafa36bbd121462aabd6
+  parser (3.3.10.1) sha256=06f6a725d2cd91e5e7f2b7c32ba143631e1f7c8ae2fb918fc4cebec187e6a688
   prism (1.8.0) sha256=84453a16ef5530ea62c5f03ec16b52a459575ad4e7b9c2b360fd8ce2c39c1254
-  pry (0.15.2) sha256=12d54b8640d3fa29c9211dd4ffb08f3fd8bf7a4fd9b5a73ce5b59c8709385b6b
-  pry-byebug (3.11.0) sha256=0b0abb7d309bc7f00044d512a3c8567274f7012b944b38becc8440439a1cea72
+  pry (0.16.0) sha256=d76c69065698ed1f85e717bd33d7942c38a50868f6b0673c636192b3d1b6054e
+  pry-byebug (3.12.0) sha256=594e094ae8a8390a7ad4c7b36ae36e13304ed02664c67417d108dc5f7213d1b7
   racc (1.8.1) sha256=4a7f6929691dbec8b5209a0b373bc2614882b55fc5d2e447a21aaa691303d62f
   rainbow (3.1.1) sha256=039491aa3a89f42efa1d6dec2fc4e62ede96eb6acd95e52f1ad581182b79bc6a
   rake (13.3.1) sha256=8c9e89d09f66a26a01264e7e3480ec0607f0c497a861ef16063604b1b08eb19c
@@ -223,6 +230,7 @@ CHECKSUMS
   rb_sys (0.9.119) sha256=64393fa148e402e1b79b64496d2aabfc7df79da6b822b8bb48dc1141eaf40b4b
   rbs (3.10.2) sha256=bd8a5dc4c62f229f020146b61844a31f9c79e649449d212904a474eb79c846fc
   regexp_parser (2.11.3) sha256=ca13f381a173b7a93450e53459075c9b76a10433caadcb2f1180f2c741fc55a4
+  reline (0.6.3) sha256=1198b04973565b36ec0f11542ab3f5cfeeec34823f4e54cebde90968092b1835
   rspec (3.13.2) sha256=206284a08ad798e61f86d7ca3e376718d52c0bc944626b2349266f239f820587
   rspec-core (3.13.6) sha256=a8823c6411667b60a8bca135364351dda34cd55e44ff94c4be4633b37d828b2d
   rspec-expectations (3.13.5) sha256=33a4d3a1d95060aea4c94e9f237030a8f9eae5615e9bd85718fe3a09e4b58836
@@ -234,7 +242,7 @@ CHECKSUMS
   rubocop-rspec (3.9.0) sha256=8fa70a3619408237d789aeecfb9beef40576acc855173e60939d63332fdb55e2
   ruby-progressbar (1.13.0) sha256=80fc9c47a9b640d6834e0dc7b3c94c9df37f08cb072b7761e4a71e22cff29b33
   securerandom (0.4.1) sha256=cc5193d414a4341b6e225f0cb4446aceca8e50d5e1888743fac16987638ea0b1
-  sorbet-runtime (0.6.12887) sha256=d5960676f23caba8f25f95240701f6d901c07faff53022209fc71b2dae74dfdb
+  sorbet-runtime (0.6.12894) sha256=4f0cbe041d80dac973ec3a5a848679922074dd77cc19f46384b27a8b9ff4a90c
   steep (1.10.0) sha256=1b295b55f9aaff1b8d3ee42453ee55bc2a1078fda0268f288edb2dc014f4d7d1
   strscan (3.1.7) sha256=5f76462b94a3ea50b44973225b7d75b2cb96d4e1bee9ef1319b99ca117b72c8c
   terminal-table (4.0.0) sha256=f504793203f8251b2ea7c7068333053f0beeea26093ec9962e62ea79f94301d2

data/README.md

@@ -33,6 +33,9 @@
   <a href="https://rubygems.org/gems/kreuzberg">
     <img src="https://img.shields.io/gem/v/kreuzberg?label=Ruby&color=007ec6" alt="Ruby">
   </a>
+  <a href="https://github.com/kreuzberg-dev/kreuzberg/pkgs/container/kreuzberg">
+    <img src="https://img.shields.io/badge/Docker-007ec6?logo=docker&logoColor=white" alt="Docker">
+  </a>
 
   <!-- Project Info -->
   <a href="https://github.com/kreuzberg-dev/kreuzberg/blob/main/LICENSE">

data/lib/kreuzberg/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kreuzberg
-  VERSION = '4.0.7'
+  VERSION = '4.0.8'
 end

data/vendor/Cargo.toml

@@ -3,7 +3,7 @@ members = ["kreuzberg", "kreuzberg-tesseract", "kreuzberg-ffi"]
 resolver = "2"
 
 [workspace.package]
-version = "4.0.7"
+version = "4.0.8"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]
@@ -47,7 +47,7 @@ hex = "0.4.3"
 toml = "0.9.11"
 num_cpus = "1.17.0"
 once_cell = "1.21.3"
-html-to-markdown-rs = { version = "2.22.2", default-features = false }
+html-to-markdown-rs = { version = "2.22.5", default-features = false }
 reqwest = { version = "0.13.1", 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 = "4.0.7"
+version = "4.0.8"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]

data/vendor/kreuzberg-ffi/Cargo.toml

@@ -28,7 +28,7 @@ serde_json = { workspace = true }
 serde = { workspace = true }
 async-trait = { workspace = true }
 tokio = { workspace = true }
-html-to-markdown-rs = { version = "2.22.2", default-features = false }
+html-to-markdown-rs = { version = "2.22.5", default-features = false }
 rayon = { version = "1.11", optional = true }
 
 [target.'cfg(all(windows, target_env = "gnu"))'.dependencies]

data/vendor/kreuzberg-ffi/src/helpers.rs

@@ -47,352 +47,11 @@ pub fn string_to_c_string(value: String) -> std::result::Result<*mut c_char, Str
 
 /// Parse extraction configuration from JSON string
 pub fn parse_extraction_config_from_json(config_str: &str) -> FfiResult<ExtractionConfig> {
-    use html_to_markdown_rs::options::{
-        CodeBlockStyle, ConversionOptions, HeadingStyle, HighlightStyle, ListIndentType, NewlineStyle,
-        PreprocessingPreset, WhitespaceMode,
-    };
-
-    fn parse_enum<T, F>(value: Option<&serde_json::Value>, parse_fn: F) -> FfiResult<Option<T>>
-    where
-        F: Fn(&str) -> std::result::Result<T, String>,
-    {
-        if let Some(raw) = value {
-            let text = raw
-                .as_str()
-                .ok_or_else(|| "Expected string for html_options enum field".to_string())?;
-            return parse_fn(text).map(Some);
-        }
-        Ok(None)
-    }
-
-    fn parse_heading_style(value: &str) -> FfiResult<HeadingStyle> {
-        match value.to_lowercase().as_str() {
-            "atx" => Ok(HeadingStyle::Atx),
-            "underlined" => Ok(HeadingStyle::Underlined),
-            "atx_closed" => Ok(HeadingStyle::AtxClosed),
-            other => Err(format!(
-                "Invalid heading_style '{}'. Expected one of: atx, underlined, atx_closed",
-                other
-            )),
-        }
-    }
-
-    fn parse_list_indent_type(value: &str) -> FfiResult<ListIndentType> {
-        match value.to_lowercase().as_str() {
-            "spaces" => Ok(ListIndentType::Spaces),
-            "tabs" => Ok(ListIndentType::Tabs),
-            other => Err(format!(
-                "Invalid list_indent_type '{}'. Expected 'spaces' or 'tabs'",
-                other
-            )),
-        }
-    }
-
-    fn parse_highlight_style(value: &str) -> FfiResult<HighlightStyle> {
-        match value.to_lowercase().as_str() {
-            "double_equal" | "==" | "highlight" => Ok(HighlightStyle::DoubleEqual),
-            "html" => Ok(HighlightStyle::Html),
-            "bold" => Ok(HighlightStyle::Bold),
-            "none" => Ok(HighlightStyle::None),
-            other => Err(format!(
-                "Invalid highlight_style '{}'. Expected one of: double_equal, html, bold, none",
-                other
-            )),
-        }
-    }
-
-    fn parse_whitespace_mode(value: &str) -> FfiResult<WhitespaceMode> {
-        match value.to_lowercase().as_str() {
-            "normalized" => Ok(WhitespaceMode::Normalized),
-            "strict" => Ok(WhitespaceMode::Strict),
-            other => Err(format!(
-                "Invalid whitespace_mode '{}'. Expected 'normalized' or 'strict'",
-                other
-            )),
-        }
-    }
-
-    fn parse_newline_style(value: &str) -> FfiResult<NewlineStyle> {
-        match value.to_lowercase().as_str() {
-            "spaces" => Ok(NewlineStyle::Spaces),
-            "backslash" => Ok(NewlineStyle::Backslash),
-            other => Err(format!(
-                "Invalid newline_style '{}'. Expected 'spaces' or 'backslash'",
-                other
-            )),
-        }
-    }
-
-    fn parse_code_block_style(value: &str) -> FfiResult<CodeBlockStyle> {
-        match value.to_lowercase().as_str() {
-            "indented" => Ok(CodeBlockStyle::Indented),
-            "backticks" => Ok(CodeBlockStyle::Backticks),
-            "tildes" => Ok(CodeBlockStyle::Tildes),
-            other => Err(format!(
-                "Invalid code_block_style '{}'. Expected 'indented', 'backticks', or 'tildes'",
-                other
-            )),
-        }
-    }
-
-    fn parse_preprocessing_preset(value: &str) -> FfiResult<PreprocessingPreset> {
-        match value.to_lowercase().as_str() {
-            "minimal" => Ok(PreprocessingPreset::Minimal),
-            "standard" => Ok(PreprocessingPreset::Standard),
-            "aggressive" => Ok(PreprocessingPreset::Aggressive),
-            other => Err(format!(
-                "Invalid preprocessing.preset '{}'. Expected one of: minimal, standard, aggressive",
-                other
-            )),
-        }
-    }
-
-    fn parse_html_options(value: &serde_json::Value) -> FfiResult<ConversionOptions> {
-        let mut opts = ConversionOptions::default();
-        let obj = value
-            .as_object()
-            .ok_or_else(|| "html_options must be an object".to_string())?;
-
-        if let Some(val) = obj.get("heading_style") {
-            opts.heading_style = parse_enum(Some(val), parse_heading_style)?.unwrap_or(opts.heading_style);
-        }
-
-        if let Some(val) = obj.get("list_indent_type") {
-            opts.list_indent_type = parse_enum(Some(val), parse_list_indent_type)?.unwrap_or(opts.list_indent_type);
-        }
-
-        if let Some(val) = obj.get("list_indent_width") {
-            opts.list_indent_width = val
-                .as_u64()
-                .map(|v| v as usize)
-                .ok_or_else(|| "list_indent_width must be an integer".to_string())?;
-        }
-
-        if let Some(val) = obj.get("bullets") {
-            opts.bullets = val
-                .as_str()
-                .map(str::to_string)
-                .ok_or_else(|| "bullets must be a string".to_string())?;
-        }
-
-        if let Some(val) = obj.get("strong_em_symbol") {
-            let symbol = val
-                .as_str()
-                .ok_or_else(|| "strong_em_symbol must be a string".to_string())?;
-            let mut chars = symbol.chars();
-            opts.strong_em_symbol = chars
-                .next()
-                .ok_or_else(|| "strong_em_symbol must not be empty".to_string())?;
-        }
-
-        if let Some(val) = obj.get("escape_asterisks") {
-            opts.escape_asterisks = val
-                .as_bool()
-                .ok_or_else(|| "escape_asterisks must be a boolean".to_string())?;
-        }
-        if let Some(val) = obj.get("escape_underscores") {
-            opts.escape_underscores = val
-                .as_bool()
-                .ok_or_else(|| "escape_underscores must be a boolean".to_string())?;
-        }
-        if let Some(val) = obj.get("escape_misc") {
-            opts.escape_misc = val
-                .as_bool()
-                .ok_or_else(|| "escape_misc must be a boolean".to_string())?;
-        }
-        if let Some(val) = obj.get("escape_ascii") {
-            opts.escape_ascii = val
-                .as_bool()
-                .ok_or_else(|| "escape_ascii must be a boolean".to_string())?;
-        }
-
-        if let Some(val) = obj.get("code_language") {
-            opts.code_language = val
-                .as_str()
-                .map(str::to_string)
-                .ok_or_else(|| "code_language must be a string".to_string())?;
-        }
-
-        if let Some(val) = obj.get("autolinks") {
-            opts.autolinks = val.as_bool().ok_or_else(|| "autolinks must be a boolean".to_string())?;
-        }
-
-        if let Some(val) = obj.get("default_title") {
-            opts.default_title = val
-                .as_bool()
-                .ok_or_else(|| "default_title must be a boolean".to_string())?;
-        }
-
-        if let Some(val) = obj.get("br_in_tables") {
-            opts.br_in_tables = val
-                .as_bool()
-                .ok_or_else(|| "br_in_tables must be a boolean".to_string())?;
-        }
-
-        if let Some(val) = obj.get("hocr_spatial_tables") {
-            opts.hocr_spatial_tables = val
-                .as_bool()
-                .ok_or_else(|| "hocr_spatial_tables must be a boolean".to_string())?;
-        }
-
-        if let Some(val) = obj.get("highlight_style") {
-            opts.highlight_style = parse_enum(Some(val), parse_highlight_style)?.unwrap_or(opts.highlight_style);
-        }
-
-        if let Some(val) = obj.get("extract_metadata") {
-            opts.extract_metadata = val
-                .as_bool()
-                .ok_or_else(|| "extract_metadata must be a boolean".to_string())?;
-        }
-
-        if let Some(val) = obj.get("whitespace_mode") {
-            opts.whitespace_mode = parse_enum(Some(val), parse_whitespace_mode)?.unwrap_or(opts.whitespace_mode);
-        }
-
-        if let Some(val) = obj.get("strip_newlines") {
-            opts.strip_newlines = val
-                .as_bool()
-                .ok_or_else(|| "strip_newlines must be a boolean".to_string())?;
-        }
-
-        if let Some(val) = obj.get("wrap") {
-            opts.wrap = val.as_bool().ok_or_else(|| "wrap must be a boolean".to_string())?;
-        }
-
-        if let Some(val) = obj.get("wrap_width") {
-            opts.wrap_width = val
-                .as_u64()
-                .map(|v| v as usize)
-                .ok_or_else(|| "wrap_width must be an integer".to_string())?;
-        }
-
-        if let Some(val) = obj.get("convert_as_inline") {
-            opts.convert_as_inline = val
-                .as_bool()
-                .ok_or_else(|| "convert_as_inline must be a boolean".to_string())?;
-        }
-
-        if let Some(val) = obj.get("sub_symbol") {
-            opts.sub_symbol = val
-                .as_str()
-                .map(str::to_string)
-                .ok_or_else(|| "sub_symbol must be a string".to_string())?;
-        }
-
-        if let Some(val) = obj.get("sup_symbol") {
-            opts.sup_symbol = val
-                .as_str()
-                .map(str::to_string)
-                .ok_or_else(|| "sup_symbol must be a string".to_string())?;
-        }
-
-        if let Some(val) = obj.get("newline_style") {
-            opts.newline_style = parse_enum(Some(val), parse_newline_style)?.unwrap_or(opts.newline_style);
-        }
-
-        if let Some(val) = obj.get("code_block_style") {
-            opts.code_block_style = parse_enum(Some(val), parse_code_block_style)?.unwrap_or(opts.code_block_style);
-        }
-
-        if let Some(val) = obj.get("keep_inline_images_in") {
-            opts.keep_inline_images_in = val
-                .as_array()
-                .ok_or_else(|| "keep_inline_images_in must be an array".to_string())?
-                .iter()
-                .map(|v| {
-                    v.as_str()
-                        .map(str::to_string)
-                        .ok_or_else(|| "keep_inline_images_in entries must be strings".to_string())
-                })
-                .collect::<std::result::Result<Vec<_>, _>>()?;
-        }
-
-        if let Some(val) = obj.get("encoding") {
-            opts.encoding = val
-                .as_str()
-                .map(str::to_string)
-                .ok_or_else(|| "encoding must be a string".to_string())?;
-        }
-
-        if let Some(val) = obj.get("debug") {
-            opts.debug = val.as_bool().ok_or_else(|| "debug must be a boolean".to_string())?;
-        }
-
-        if let Some(val) = obj.get("strip_tags") {
-            opts.strip_tags = val
-                .as_array()
-                .ok_or_else(|| "strip_tags must be an array".to_string())?
-                .iter()
-                .map(|v| {
-                    v.as_str()
-                        .map(str::to_string)
-                        .ok_or_else(|| "strip_tags entries must be strings".to_string())
-                })
-                .collect::<std::result::Result<Vec<_>, _>>()?;
-        }
-
-        if let Some(val) = obj.get("preserve_tags") {
-            opts.preserve_tags = val
-                .as_array()
-                .ok_or_else(|| "preserve_tags must be an array".to_string())?
-                .iter()
-                .map(|v| {
-                    v.as_str()
-                        .map(str::to_string)
-                        .ok_or_else(|| "preserve_tags entries must be strings".to_string())
-                })
-                .collect::<std::result::Result<Vec<_>, _>>()?;
-        }
-
-        if let Some(val) = obj.get("preprocessing") {
-            let pre = val
-                .as_object()
-                .ok_or_else(|| "preprocessing must be an object".to_string())?;
-            let mut preprocessing = opts.preprocessing.clone();
-
-            if let Some(v) = pre.get("enabled") {
-                preprocessing.enabled = v
-                    .as_bool()
-                    .ok_or_else(|| "preprocessing.enabled must be a boolean".to_string())?;
-            }
-
-            if let Some(v) = pre.get("preset") {
-                let preset = v
-                    .as_str()
-                    .ok_or_else(|| "preprocessing.preset must be a string".to_string())?;
-                preprocessing.preset = parse_preprocessing_preset(preset)?;
-            }
-
-            if let Some(v) = pre.get("remove_navigation") {
-                preprocessing.remove_navigation = v
-                    .as_bool()
-                    .ok_or_else(|| "preprocessing.remove_navigation must be a boolean".to_string())?;
-            }
-
-            if let Some(v) = pre.get("remove_forms") {
-                preprocessing.remove_forms = v
-                    .as_bool()
-                    .ok_or_else(|| "preprocessing.remove_forms must be a boolean".to_string())?;
-            }
-
-            opts.preprocessing = preprocessing;
-        }
-
-        Ok(opts)
-    }
-
-    let value: serde_json::Value =
+    // html-to-markdown-rs v2.22.5+ has #[serde(default)] on ConversionOptions,
+    // so serde can now handle partial deserialization with defaults for missing fields
+    let config: ExtractionConfig =
         serde_json::from_str(config_str).map_err(|e| format!("Failed to parse config JSON: {}", e))?;
 
-    let html_options = value.get("html_options").map(parse_html_options).transpose()?;
-
-    let mut config: ExtractionConfig =
-        serde_json::from_value(value).map_err(|e| format!("Failed to parse config JSON: {}", e))?;
-
-    if let Some(options) = html_options {
-        config.html_options = Some(options);
-    }
-
     Ok(config)
 }
 
@@ -596,8 +255,8 @@ mod tests {
     fn test_parse_extraction_config_with_html_options() {
         let json = r#"{
             "html_options": {
-                "heading_style": "atx",
-                "escape_asterisks": true,
+                "headingStyle": "atx",
+                "escapeAsterisks": true,
                 "autolinks": false
             }
         }"#;
@@ -620,16 +279,20 @@ mod tests {
     }
 
     #[test]
-    fn test_parse_extraction_config_invalid_heading_style() {
+    fn test_parse_extraction_config_invalid_heading_style_uses_default() {
+        // With #[serde(default)], invalid enum values use the default instead of failing
         let json = r#"{
             "html_options": {
-                "heading_style": "invalid_style"
+                "headingStyle": "invalid_style"
             }
         }"#;
 
         let result = parse_extraction_config_from_json(json);
-        assert!(result.is_err());
-        assert!(result.unwrap_err().contains("Invalid heading_style"));
+        assert!(result.is_ok(), "Parsing should succeed with default values");
+
+        // Invalid enum values should be ignored and default value used
+        let config = result.unwrap();
+        assert!(config.html_options.is_some());
     }
 
     #[test]
@@ -641,7 +304,7 @@ mod tests {
         ];
 
         for (input, _expected) in styles {
-            let json = format!(r#"{{"html_options": {{"heading_style": "{}"}}}}"#, input);
+            let json = format!(r#"{{"html_options": {{"headingStyle": "{}"}}}}"#, input);
             let result = parse_extraction_config_from_json(&json);
             assert!(result.is_ok(), "Failed to parse heading_style: {}", input);
         }
@@ -654,8 +317,8 @@ mod tests {
                 "preprocessing": {
                     "enabled": true,
                     "preset": "aggressive",
-                    "remove_navigation": true,
-                    "remove_forms": false
+                    "removeNavigation": true,
+                    "removeForms": false
                 }
             }
         }"#;

data/vendor/kreuzberg-tesseract/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "kreuzberg-tesseract"
-version = "4.0.7"
+version = "4.0.8"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <nhirschfeld@gmail.com>"]
@@ -24,7 +24,7 @@ thiserror = { workspace = true }
 image = { workspace = true }
 
 [build-dependencies]
-cc = { version = "^1.2.52", optional = true }
+cc = { version = "^1.2.53", optional = true }
 cmake = { version = "0.1.57", optional = true }
 zip = { version = "7.1.0", optional = true }
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: kreuzberg
 version: !ruby/object:Gem::Version
-  version: 4.0.7
+  version: 4.0.8
 platform: ruby
 authors:
 - Na'aman Hirschfeld
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-01-16 00:00:00.000000000 Z
+date: 2026-01-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler