xberg 1.0.0.pre.rc.42 → 1.0.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 843c35b446da636fad246d5b41d7e0e0b1e19ddb858bb91ab88709713a7426b7
-  data.tar.gz: 270431cbbb6679ff55d279751fc3b7d994022ce35bd5fe50d10508f1ce7eb581
+  metadata.gz: 3e42f004ec632c1b79662972b8326a86ad630232b3959f79cfec2fb12f5baf5a
+  data.tar.gz: 63cf4f30361a3ad1b88cbe8710bedb4afef421134173ab065457f4f129ddedf2
 SHA512:
-  metadata.gz: edb429989127405f6f241a900dac9f702cf568a465d78883eff72988c96ec514d5f4ef0f07f5da60b9ba1333c0ea7205c2c9e57f29312ef1064512754563d324
-  data.tar.gz: ab0000882144aa9c555f12ea83b1a68945ee93bd238c99c82440f3e2031bb7c781c505693500e1f937abc279cec19aeda79c663622addc9912a5e11632c02b4f
+  metadata.gz: cdd62fee277dcad1ade38a4bd08127e9b1bbe5131dc05b4556ce16e04d4e04904bc612db1a30c40e427bf9d35babff96c86ca2d670ad7b108dfd6517e9ed0f38
+  data.tar.gz: 34d121faec6d409d71b402d8454d09054cf990390460320b404f04bcc13b9f6238f3172d50f5fb8ae65ecbdf2034413dc81de417a500df4a264e93004b7f70ec

data/README.md

@@ -80,7 +80,7 @@
   </a>
 </div>
 
-Extract text, tables, images, metadata, and code intelligence from 97 file formats and 306 programming languages including PDF, Office documents, images, and audio/video transcripts where native transcription is available. Ruby bindings with idiomatic Ruby API and native performance.
+Extract text, tables, images, metadata, and code intelligence from 98 file formats and 306 programming languages including PDF, Office documents, images, and audio/video transcripts where native transcription is available. Ruby bindings with idiomatic Ruby API and native performance.
 
 ## What This Package Provides
 

data/ext/xberg_rb/native/Cargo.lock

@@ -827,9 +827,9 @@ dependencies = [
 
 [[package]]
 name = "calamine"
-version = "0.36.0"
+version = "0.36.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6975084f43060e56343ffba7f9731fa52a7dcf2e1cd8e2459fd4c6bf4a1bff59"
+checksum = "5fa68281b1a76b54a62156474adb06bb380a67e07dd60656e3217152b42183f3"
 dependencies = [
  "atoi_simd",
  "byteorder",
@@ -1346,9 +1346,9 @@ dependencies = [
 
 [[package]]
 name = "crawlberg"
-version = "1.0.9"
+version = "1.0.10"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b2841d7bdd19a54e025f61e852c54e0af8f5d492e6ed862dbb0b1363e4f9a29f"
+checksum = "cab5814c0d5db694be37a31c5a322dc85027650cb0b09f88ad0b6188ebdfe181"
 dependencies = [
  "ahash 0.8.12",
  "aho-corasick",
@@ -3406,9 +3406,9 @@ checksum = "47f142fe24a9c9944451e8349de0a56af5f3e7226dc46f3ed4d4ecc0b85af75e"
 
 [[package]]
 name = "jiff"
-version = "0.2.34"
+version = "0.2.35"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e184d09547b80eb7e20d141ba2fb1fbac843ca53f4cf1b31210adc4c1adc6e16"
+checksum = "668b7183bd07af9a4885f5c35b0cc5c83c4607a913c16b7e17291832910d2dcc"
 dependencies = [
  "defmt",
  "jiff-core",
@@ -3430,9 +3430,9 @@ dependencies = [
 
 [[package]]
 name = "jiff-static"
-version = "0.2.34"
+version = "0.2.35"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "323da076b7a6faf914dc677cb05a4b907742ff7375c8322c9e7f5061e5e0e9de"
+checksum = "3a69dcb3a21cfb32ce1cd056169337ca284af0766dd766e7878819b251a49204"
 dependencies = [
  "jiff-core",
  "proc-macro2",
@@ -3739,6 +3739,18 @@ dependencies = [
  "libc",
 ]
 
+[[package]]
+name = "libz-sys"
+version = "1.1.29"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "85bc9657773828b90eeb625adff10eeac83cc21bbfd8e23a03eaa8a33c9e28d9"
+dependencies = [
+ "cc",
+ "libc",
+ "pkg-config",
+ "vcpkg",
+]
+
 [[package]]
 name = "link-section"
 version = "0.19.1"
@@ -3765,9 +3777,9 @@ checksum = "92daf443525c4cce67b150400bc2316076100ce0b3686209eb8cf3c31612e6f0"
 
 [[package]]
 name = "liter-llm"
-version = "1.11.2"
+version = "1.11.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6b9cadc4e7d91a51d237279b777d58e2ca25223159963f0a66bf17f4157e0274"
+checksum = "e9c2f37e8337fa18108051c50b49360effb2d22d797b94d792f99861da092eee"
 dependencies = [
  "ahash 0.8.12",
  "arc-swap",
@@ -4833,9 +4845,9 @@ dependencies = [
 
 [[package]]
 name = "pdf_oxide"
-version = "0.3.75"
+version = "0.3.76"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2e00ca5f52780bb764da7fc1783c0c6f549fcf960d69734c8d7cf7996bdbb49d"
+checksum = "a5ec03a138df06a1b063e5053bd957aab153ccfb5a59a326e285b57b3e1bd600"
 dependencies = [
  "aes 0.9.1",
  "base64 0.22.1",
@@ -7837,9 +7849,9 @@ checksum = "009994f150cc0cd50ff54917d5bc8bffe8cad10ca10d81c34da2ec421ae61782"
 
 [[package]]
 name = "tree-sitter-language-pack"
-version = "1.13.4"
+version = "1.13.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "57d7a66a4a143bc10fa060c2f28f81758ec20b4a335fdb8bdf95292709af409b"
+checksum = "445e69b4cdfd2758eeed0badaacd9eb037459aeb4d8bda5f21fb45d8e6958cac"
 dependencies = [
  "ahash 0.8.12",
  "cc",
@@ -8875,9 +8887,9 @@ dependencies = [
 
 [[package]]
 name = "xberg"
-version = "1.0.0-rc.42"
+version = "1.0.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "11a00254a4b2986b442724ac2bdba9fb5b436da7d9e8dfb265d13e8b1ce9b6e9"
+checksum = "8fb349741ea65ad8c5dcf178a413a0054fac4cd1e399a1d91abadc0688dbc81b"
 dependencies = [
  "aes-gcm",
  "ahash 0.8.12",
@@ -8894,6 +8906,7 @@ dependencies = [
  "cfb 0.14.0",
  "chardetng",
  "chrono",
+ "codepage",
  "comrak",
  "crawlberg",
  "dbase",
@@ -8992,6 +9005,7 @@ dependencies = [
  "whatlang",
  "xberg-gliner",
  "xberg-libheif",
+ "xberg-libwpd",
  "xberg-paddle-ocr",
  "xberg-tesseract",
  "zeroize",
@@ -9000,9 +9014,9 @@ dependencies = [
 
 [[package]]
 name = "xberg-gliner"
-version = "1.0.0-rc.42"
+version = "1.0.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "62e67c0be3f48b9a09681fee1873c5463abc5491aab44327359e24180c57d2cd"
+checksum = "eeef925ad620d65ef94bb359f4d7ce3bf960eeb3b141f31e6b432a54cc3306b2"
 dependencies = [
  "getrandom 0.2.17",
  "getrandom 0.3.4",
@@ -9016,9 +9030,9 @@ dependencies = [
 
 [[package]]
 name = "xberg-libheif"
-version = "1.0.0-rc.42"
+version = "1.0.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4f1afdf970190b5a4c754fd037b2cb02c927633f12cdc44f2768058e68e8b5d8"
+checksum = "0e5c57720eceaf8669192e70c2b60e2970b589dd39d8d4315dd8e8c40ada5fb3"
 dependencies = [
  "cfg-if 1.0.4",
  "enumn",
@@ -9027,11 +9041,26 @@ dependencies = [
  "libheif-sys",
 ]
 
+[[package]]
+name = "xberg-libwpd"
+version = "1.0.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "74a0611a3b4d80b95162381848878a928ce3d3abbf50ef043769b9c95edf29c5"
+dependencies = [
+ "cc",
+ "flate2",
+ "libz-sys",
+ "sha2 0.11.0",
+ "tar",
+ "thiserror 2.0.19",
+ "tracing",
+]
+
 [[package]]
 name = "xberg-paddle-ocr"
-version = "1.0.0-rc.42"
+version = "1.0.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6878bc3247ec24ddcb8e57827b007f2ff42a6f35e6c1ea982275951fcdcde54f"
+checksum = "534f7bb90aee032ec2130d503a508e1e5f24ad3cbd305a26031645852cf9dfd2"
 dependencies = [
  "geo-clipper",
  "geo-types",
@@ -9046,7 +9075,7 @@ dependencies = [
 
 [[package]]
 name = "xberg-rb"
-version = "1.0.0-rc.42"
+version = "1.0.1"
 dependencies = [
  "async-trait",
  "crawlberg",
@@ -9061,9 +9090,9 @@ dependencies = [
 
 [[package]]
 name = "xberg-tesseract"
-version = "1.0.0-rc.42"
+version = "1.0.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "019a1b0d14535c8a1eb7822249dd6dfcb83345ef8cd81dd0eff00e769dce1b5f"
+checksum = "0d839c598c71d536069372e7d726130c354f6aee7d1b10adf0490588f6abc79d"
 dependencies = [
  "cc",
  "cmake",

data/ext/xberg_rb/native/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "xberg-rb"
-version = "1.0.0-rc.42"
+version = "1.0.1"
 edition = "2024"
 license = "MIT"
 description = "High-performance document intelligence library"
@@ -56,7 +56,7 @@ xml = ["xberg/xml"]
 
 [dependencies]
 async-trait = "0.1"
-crawlberg = { default-features = false, version = "1.0.9" }
+crawlberg = { default-features = false, version = "1.0.10" }
 html-to-markdown-rs = { default-features = false, version = "3.9" }
 magnus = "0.8"
 rb-sys = ">=0.9, <0.9.128"
@@ -64,10 +64,10 @@ serde = { version = "1", features = ["derive"] }
 serde_json = "1"
 tokio = { version = "1", features = ["rt-multi-thread"] }
 [target.'cfg(not(any(target_os = "windows", all(target_os = "macos", target_arch = "x86_64"))))'.dependencies]
-xberg = { version = "1.0.0-rc.42", features = ["full", "pdf", "ocr", "paddle-ocr", "paddle-ocr-types", "layout-detection", "layout-types", "embeddings", "embedding-presets", "reranker", "reranker-presets", "sparse-embeddings", "sparse-embedding-presets", "late-interaction", "late-interaction-presets", "chunking", "keywords-yake", "keywords-rake", "language-detection", "html", "tree-sitter", "office", "email", "archives", "stopwords", "auto-rotate", "auto-rotate-types", "tokio-runtime", "api", "api-types", "mcp", "liter-llm", "quality", "svg", "transcription", "transcription-types", "classification", "captioning", "ner-onnx", "ner-llm", "diff", "markdown-footnotes", "redaction", "enrichment", "heuristics", "presets", "structured", "url-ingestion"] }
+xberg = { version = "1.0.1", features = ["full", "pdf", "ocr", "paddle-ocr", "paddle-ocr-types", "layout-detection", "layout-types", "embeddings", "embedding-presets", "reranker", "reranker-presets", "sparse-embeddings", "sparse-embedding-presets", "late-interaction", "late-interaction-presets", "chunking", "keywords-yake", "keywords-rake", "language-detection", "html", "tree-sitter", "office", "email", "archives", "stopwords", "auto-rotate", "auto-rotate-types", "tokio-runtime", "api", "api-types", "mcp", "liter-llm", "quality", "svg", "transcription", "transcription-types", "classification", "captioning", "ner-onnx", "ner-llm", "diff", "markdown-footnotes", "redaction", "enrichment", "heuristics", "presets", "structured", "url-ingestion"] }
 
 [target.'cfg(target_os = "windows")'.dependencies]
-xberg = { version = "1.0.0-rc.42", features = ["windows-target"] }
+xberg = { version = "1.0.1", features = ["windows-target"] }
 
 [target.'cfg(all(target_os = "macos", target_arch = "x86_64"))'.dependencies]
-xberg = { version = "1.0.0-rc.42", features = ["macos-intel-target"] }
+xberg = { version = "1.0.1", features = ["macos-intel-target"] }

data/ext/xberg_rb/src/lib.rs

@@ -1,5 +1,5 @@
 // This file is auto-generated by alef. DO NOT EDIT.
-// alef:hash:27b14662489dbd72bffe76d749cce63acd51fb05b9d0eea856452c7d4c92c223
+// alef:hash:ba94fe4cec719c31f8ac484fcc2aba1d30db0f87dea8b1b6d0257417a4a260e2
 // Re-generate with: alef generate
 #![allow(dead_code, unused_imports, unused_variables)]
 #![allow(
@@ -2787,6 +2787,7 @@ impl LateInteractionConfig {
 #[serde(default)]
 #[magnus::wrap(class = "Xberg::LayoutDetectionConfig")]
 pub struct LayoutDetectionConfig {
+    strategy: LayoutStrategy,
     confidence_threshold: Option<f32>,
     apply_heuristics: bool,
     table_model: TableModel,
@@ -2836,6 +2837,10 @@ impl LayoutDetectionConfig {
         let (kwargs_opt,) = args.optional;
         let kwargs = kwargs_opt.unwrap_or_else(|| ruby.hash_new());
         Ok(Self {
+            strategy: kwargs
+                .get(ruby.to_symbol("strategy"))
+                .and_then(|v| LayoutStrategy::try_convert(v).ok())
+                .unwrap_or_default(),
             confidence_threshold: kwargs
                 .get(ruby.to_symbol("confidence_threshold"))
                 .and_then(|v| f32::try_convert(v).ok()),
@@ -2861,6 +2866,10 @@ impl LayoutDetectionConfig {
         })
     }
 
+    fn strategy(&self) -> LayoutStrategy {
+        self.strategy.clone()
+    }
+
     fn confidence_threshold(&self) -> Option<f32> {
         self.confidence_threshold
     }
@@ -21015,6 +21024,8 @@ pub struct PdfMetadata {
     page_count: Option<u32>,
     scanned_confidence: Option<f32>,
     scanned_pages: Option<Vec<u32>>,
+    layout_gated_pages: Option<Vec<u32>>,
+    layout_gate_reasons: Option<Vec<String>>,
 }
 
 unsafe impl IntoValueFromNative for PdfMetadata {}
@@ -21082,6 +21093,12 @@ impl PdfMetadata {
             scanned_pages: kwargs
                 .get(ruby.to_symbol("scanned_pages"))
                 .and_then(|v| <Vec<u32>>::try_convert(v).ok()),
+            layout_gated_pages: kwargs
+                .get(ruby.to_symbol("layout_gated_pages"))
+                .and_then(|v| <Vec<u32>>::try_convert(v).ok()),
+            layout_gate_reasons: kwargs
+                .get(ruby.to_symbol("layout_gate_reasons"))
+                .and_then(|v| <Vec<String>>::try_convert(v).ok()),
         })
     }
 
@@ -21116,6 +21133,14 @@ impl PdfMetadata {
     fn scanned_pages(&self) -> Option<Vec<u32>> {
         self.scanned_pages.clone()
     }
+
+    fn layout_gated_pages(&self) -> Option<Vec<u32>> {
+        self.layout_gated_pages.clone()
+    }
+
+    fn layout_gate_reasons(&self) -> Option<Vec<String>> {
+        self.layout_gate_reasons.clone()
+    }
 }
 
 #[derive(Clone, Debug, serde::Serialize, serde::Deserialize)]
@@ -23190,6 +23215,48 @@ impl magnus::TryConvert for TableOverlapPreference {
 unsafe impl IntoValueFromNative for TableOverlapPreference {}
 unsafe impl TryConvertOwned for TableOverlapPreference {}
 
+#[derive(Clone, Copy, PartialEq, Eq, Debug, serde::Serialize, serde::Deserialize)]
+#[serde(rename_all = "snake_case")]
+pub enum LayoutStrategy {
+    Always,
+    Auto,
+}
+
+impl Default for LayoutStrategy {
+    fn default() -> Self {
+        Self::Always
+    }
+}
+
+impl magnus::IntoValue for LayoutStrategy {
+    fn into_value_with(self, handle: &Ruby) -> magnus::Value {
+        let sym = match self {
+            LayoutStrategy::Always => "always",
+            LayoutStrategy::Auto => "auto",
+        };
+        handle.to_symbol(sym).into_value_with(handle)
+    }
+}
+
+impl magnus::TryConvert for LayoutStrategy {
+    fn try_convert(val: magnus::Value) -> Result<Self, magnus::Error> {
+        let s: String = magnus::TryConvert::try_convert(val)?;
+        // Accept the serde wire name (snake_case), the PascalCase Rust variant name,
+        // and a lowercase fallback so fixtures written in any of those styles work.
+        match s.as_str() {
+            "always" | "Always" => Ok(LayoutStrategy::Always),
+            "auto" | "Auto" => Ok(LayoutStrategy::Auto),
+            other => Err(magnus::Error::new(
+                unsafe { Ruby::get_unchecked() }.exception_arg_error(),
+                format!("invalid LayoutStrategy value: {other}"),
+            )),
+        }
+    }
+}
+
+unsafe impl IntoValueFromNative for LayoutStrategy {}
+unsafe impl TryConvertOwned for LayoutStrategy {}
+
 #[derive(Clone, Copy, PartialEq, Eq, Debug, serde::Serialize, serde::Deserialize)]
 #[serde(rename_all = "snake_case")]
 pub enum CallMode {
@@ -31462,6 +31529,7 @@ impl From<xberg::LateInteractionConfig> for LateInteractionConfig {
 impl From<LayoutDetectionConfig> for xberg::LayoutDetectionConfig {
     fn from(val: LayoutDetectionConfig) -> Self {
         Self {
+            strategy: val.strategy.into(),
             confidence_threshold: val.confidence_threshold,
             apply_heuristics: val.apply_heuristics,
             table_model: val.table_model.into(),
@@ -31477,6 +31545,7 @@ impl From<LayoutDetectionConfig> for xberg::LayoutDetectionConfig {
 impl From<xberg::LayoutDetectionConfig> for LayoutDetectionConfig {
     fn from(val: xberg::LayoutDetectionConfig) -> Self {
         Self {
+            strategy: val.strategy.into(),
             confidence_threshold: val.confidence_threshold,
             apply_heuristics: val.apply_heuristics,
             table_model: val.table_model.into(),
@@ -35987,6 +36056,8 @@ impl From<PdfMetadata> for xberg::pdf::metadata::PdfMetadata {
             page_count: val.page_count,
             scanned_confidence: val.scanned_confidence,
             scanned_pages: val.scanned_pages.map(|v| v.into_iter().collect()),
+            layout_gated_pages: val.layout_gated_pages.map(|v| v.into_iter().collect()),
+            layout_gate_reasons: val.layout_gate_reasons.map(|v| v.into_iter().collect()),
             ..Default::default()
         }
     }
@@ -36004,6 +36075,8 @@ impl From<xberg::pdf::metadata::PdfMetadata> for PdfMetadata {
             page_count: val.page_count,
             scanned_confidence: val.scanned_confidence,
             scanned_pages: val.scanned_pages.map(|v| v.into_iter().collect()),
+            layout_gated_pages: val.layout_gated_pages.map(|v| v.into_iter().collect()),
+            layout_gate_reasons: val.layout_gate_reasons.map(|v| v.into_iter().collect()),
         }
     }
 }
@@ -36667,6 +36740,24 @@ impl From<xberg::core::config::layout::TableOverlapPreference> for TableOverlapP
     }
 }
 
+impl From<LayoutStrategy> for xberg::LayoutStrategy {
+    fn from(val: LayoutStrategy) -> Self {
+        match val {
+            LayoutStrategy::Always => Self::Always,
+            LayoutStrategy::Auto => Self::Auto,
+        }
+    }
+}
+
+impl From<xberg::LayoutStrategy> for LayoutStrategy {
+    fn from(val: xberg::LayoutStrategy) -> Self {
+        match val {
+            xberg::LayoutStrategy::Always => Self::Always,
+            xberg::LayoutStrategy::Auto => Self::Auto,
+        }
+    }
+}
+
 impl From<CallMode> for xberg::CallMode {
     fn from(val: CallMode) -> Self {
         match val {
@@ -39178,6 +39269,8 @@ fn ruby_init(ruby: &Ruby) -> Result<(), Error> {
 
     class.define_singleton_method("new", function!(LayoutDetectionConfig::new, -1))?;
 
+    class.define_method("strategy", method!(LayoutDetectionConfig::strategy, 0))?;
+
     class.define_method(
         "confidence_threshold",
         method!(LayoutDetectionConfig::confidence_threshold, 0),
@@ -42190,6 +42283,10 @@ fn ruby_init(ruby: &Ruby) -> Result<(), Error> {
 
     class.define_method("scanned_pages", method!(PdfMetadata::scanned_pages, 0))?;
 
+    class.define_method("layout_gated_pages", method!(PdfMetadata::layout_gated_pages, 0))?;
+
+    class.define_method("layout_gate_reasons", method!(PdfMetadata::layout_gate_reasons, 0))?;
+
     let class = module.define_class("ChunkClassificationEnrichmentConfig", ruby.class_object())?;
 
     class.define_singleton_method("new", function!(ChunkClassificationEnrichmentConfig::new, -1))?;

data/lib/xberg/native.rb

@@ -1,5 +1,5 @@
 # This file is auto-generated by alef — DO NOT EDIT.
-# alef:hash:27b14662489dbd72bffe76d749cce63acd51fb05b9d0eea856452c7d4c92c223
+# alef:hash:ba94fe4cec719c31f8ac484fcc2aba1d30db0f87dea8b1b6d0257417a4a260e2
 # To regenerate: alef generate
 # To verify freshness: alef verify --exit-code
 # frozen_string_literal: true

data/lib/xberg/version.rb

@@ -1,10 +1,10 @@
 # This file is auto-generated by alef — DO NOT EDIT.
-# alef:hash:27b14662489dbd72bffe76d749cce63acd51fb05b9d0eea856452c7d4c92c223
+# alef:hash:ba94fe4cec719c31f8ac484fcc2aba1d30db0f87dea8b1b6d0257417a4a260e2
 # To regenerate: alef generate
 # To verify freshness: alef verify --exit-code
 # frozen_string_literal: true
 
 module Xberg
   ## The version string for this package.
-  VERSION = "1.0.0.pre.rc.42"
+  VERSION = "1.0.1"
 end

data/lib/xberg.rb

@@ -1,5 +1,5 @@
 # This file is auto-generated by alef — DO NOT EDIT.
-# alef:hash:27b14662489dbd72bffe76d749cce63acd51fb05b9d0eea856452c7d4c92c223
+# alef:hash:ba94fe4cec719c31f8ac484fcc2aba1d30db0f87dea8b1b6d0257417a4a260e2
 # To regenerate: alef generate
 # To verify freshness: alef verify --exit-code
 # frozen_string_literal: true

data/sig/types.rbs

@@ -1,5 +1,5 @@
 # This file is auto-generated by alef — DO NOT EDIT.
-# alef:hash:27b14662489dbd72bffe76d749cce63acd51fb05b9d0eea856452c7d4c92c223
+# alef:hash:ba94fe4cec719c31f8ac484fcc2aba1d30db0f87dea8b1b6d0257417a4a260e2
 # To regenerate: alef generate
 # To verify freshness: alef verify --exit-code
 
@@ -302,6 +302,7 @@ def initialize: (?results: Array[ExtractedDocument], ?errors: Array[ExtractionEr
     end
 
     class LayoutDetectionConfig
+        attr_accessor strategy: LayoutStrategy?
         attr_accessor confidence_threshold: Float?
         attr_accessor apply_heuristics: bool?
         attr_accessor table_model: TableModel?
@@ -309,7 +310,7 @@ def initialize: (?results: Array[ExtractedDocument], ?errors: Array[ExtractionEr
         attr_accessor acceleration: AccelerationConfig?
         attr_accessor enable_chart_understanding: bool?
 
-    def initialize: (?confidence_threshold: Float, ?apply_heuristics: bool, ?table_model: TableModel, ?table_overlap_preference: TableOverlapPreference, ?acceleration: AccelerationConfig, ?enable_chart_understanding: bool) -> void
+    def initialize: (?strategy: LayoutStrategy, ?confidence_threshold: Float, ?apply_heuristics: bool, ?table_model: TableModel, ?table_overlap_preference: TableOverlapPreference, ?acceleration: AccelerationConfig, ?enable_chart_understanding: bool) -> void
     def self.default: () -> LayoutDetectionConfig
     end
 
@@ -2257,8 +2258,10 @@ def initialize: (page_width: Integer, page_height: Integer, detections: Array[La
         attr_accessor page_count: Integer?
         attr_accessor scanned_confidence: Float?
     attr_accessor scanned_pages: Array[Integer]?
+    attr_accessor layout_gated_pages: Array[Integer]?
+    attr_accessor layout_gate_reasons: Array[String]?
 
-def initialize: (?pdf_version: String, ?producer: String, ?is_encrypted: bool, ?width: Integer, ?height: Integer, ?page_count: Integer, ?scanned_confidence: Float, ?scanned_pages: Array[Integer]) -> void
+def initialize: (?pdf_version: String, ?producer: String, ?is_encrypted: bool, ?width: Integer, ?height: Integer, ?page_count: Integer, ?scanned_confidence: Float, ?scanned_pages: Array[Integer], ?layout_gated_pages: Array[Integer], ?layout_gate_reasons: Array[String]) -> void
     end
 
     class ChunkClassificationEnrichmentConfig
@@ -2466,6 +2469,10 @@ def initialize: (?heading_style: HeadingStyle, ?list_indent_type: ListIndentType
         type value = :content | :native | :layout
     end
 
+    class LayoutStrategy
+        type value = :always | :auto
+    end
+
     class CallMode
         type value = :text_only | :vision_only | :text_plus_vision
     end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: xberg
 version: !ruby/object:Gem::Version
-  version: 1.0.0.pre.rc.42
+  version: 1.0.1
 platform: ruby
 authors:
 - Na'aman Hirschfeld <naaman@xberg.io>
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2026-07-27 00:00:00.000000000 Z
+date: 2026-07-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rb_sys