Downgrade kreuzberg because of license change

Author: ScriptSmith

Cargo.lock

@@ -350,7 +350,7 @@ google-cloud-token = { version = "0.1", optional = true }
 hickory-resolver = { version = "0.25", features = ["tokio", "system-config"], optional = true }
 hostname = { version = "0.4.2", optional = true }
 jsonschema = { version = "0.29", optional = true }
-kreuzberg = { version = "4.8", default-features = false, features = ["tokio-runtime", "bundled-pdfium", "office", "excel", "ocr"], optional = true }
+kreuzberg = { version = "~4.7", default-features = false, features = ["tokio-runtime", "bundled-pdfium", "office", "excel", "ocr"], optional = true }
 metrics = { version = "0.24", optional = true }
 metrics-exporter-prometheus = { version = "0.16", optional = true }
 open = { version = "5.3.3", optional = true }

Cargo.toml

@@ -2191,16 +2191,20 @@ fn build_kreuzberg_config(config: &DocumentExtractionConfig) -> kreuzberg::Extra
         kreuzberg_config.ocr = Some(kreuzberg::OcrConfig {
             backend: "tesseract".to_string(),
             language: config.ocr_language.clone(),
-            ..Default::default()
+            tesseract_config: None,
+            output_format: None,
+            paddle_ocr_config: None,
+            element_config: None,
         });
         kreuzberg_config.force_ocr = config.force_ocr;
     }
 
     // Configure PDF-specific options
     kreuzberg_config.pdf_options = Some(kreuzberg::PdfConfig {
         extract_images: config.pdf_extract_images,
+        passwords: None,
         extract_metadata: true,
-        ..Default::default()
+        hierarchy: None,
     });
 
     // Configure image extraction settings (includes DPI for OCR)
@@ -2212,7 +2216,6 @@ fn build_kreuzberg_config(config: &DocumentExtractionConfig) -> kreuzberg::Extra
             auto_adjust_dpi: true,
             min_dpi: 72,
             max_dpi: 600,
-            ..Default::default()
         });
     }