chore: ci fixes

Author: Goldziher

.github/actions/setup-paddle-ocr-models/action.yml

@@ -148,21 +148,21 @@ runs:
         echo "Checking for PaddleOCR models in $CACHE_DIR"
 
         if [ -f "$CACHE_DIR/det/model.onnx" ]; then
-          SIZE=$(du -b "$CACHE_DIR/det/model.onnx" | cut -f1)
+          SIZE=$(wc -c < "$CACHE_DIR/det/model.onnx" | tr -d ' ')
           AVAILABLE_MODELS+=("det")
           TOTAL_SIZE=$((TOTAL_SIZE + SIZE))
           echo "  ✓ Detection model: $(numfmt --to=iec-i --suffix=B $SIZE 2>/dev/null || echo $SIZE bytes)"
         fi
 
         if [ -f "$CACHE_DIR/cls/model.onnx" ]; then
-          SIZE=$(du -b "$CACHE_DIR/cls/model.onnx" | cut -f1)
+          SIZE=$(wc -c < "$CACHE_DIR/cls/model.onnx" | tr -d ' ')
           AVAILABLE_MODELS+=("cls")
           TOTAL_SIZE=$((TOTAL_SIZE + SIZE))
           echo "  ✓ Classification model: $(numfmt --to=iec-i --suffix=B $SIZE 2>/dev/null || echo $SIZE bytes)"
         fi
 
         if [ -f "$CACHE_DIR/rec/model.onnx" ]; then
-          SIZE=$(du -b "$CACHE_DIR/rec/model.onnx" | cut -f1)
+          SIZE=$(wc -c < "$CACHE_DIR/rec/model.onnx" | tr -d ' ')
           AVAILABLE_MODELS+=("rec")
           TOTAL_SIZE=$((TOTAL_SIZE + SIZE))
           echo "  ✓ Recognition model: $(numfmt --to=iec-i --suffix=B $SIZE 2>/dev/null || echo $SIZE bytes)"

.task/languages/rust.yml

@@ -79,9 +79,10 @@ tasks:
       PDFIUM_VERSION: '{{.PDFIUM_VERSION}}'
     cmds:
       # Note: exclude benchmark-harness on Windows as jemalloc doesn't build with MSVC
-      - cmd: cargo build --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm
+      # Note: exclude kreuzberg-pdfium-render because --all-features enables mutually exclusive static/dynamic linking
+      - cmd: cargo build --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm --exclude kreuzberg-pdfium-render
         platforms: [linux, darwin]
-      - cmd: cargo build --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm --exclude benchmark-harness
+      - cmd: cargo build --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm --exclude benchmark-harness --exclude kreuzberg-pdfium-render
         platforms: [windows]
 
   build:release:
@@ -91,9 +92,9 @@ tasks:
       CMAKE: '{{.CMAKE_PATH}}'
       PDFIUM_VERSION: '{{.PDFIUM_VERSION}}'
     cmds:
-      - cmd: cargo build --release --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm
+      - cmd: cargo build --release --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm --exclude kreuzberg-pdfium-render
         platforms: [linux, darwin]
-      - cmd: cargo build --release --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm --exclude benchmark-harness
+      - cmd: cargo build --release --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm --exclude benchmark-harness --exclude kreuzberg-pdfium-render
         platforms: [windows]
 
   build:ci:
@@ -104,9 +105,9 @@ tasks:
       PDFIUM_VERSION: '{{.PDFIUM_VERSION}}'
       RUSTFLAGS: '-C debuginfo=2'
     cmds:
-      - cmd: cargo build --release --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm
+      - cmd: cargo build --release --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm --exclude kreuzberg-pdfium-render
         platforms: [linux, darwin]
-      - cmd: cargo build --release --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm --exclude benchmark-harness
+      - cmd: cargo build --release --workspace --all-features --exclude kreuzberg-php --exclude kreuzberg-node --exclude kreuzberg-wasm --exclude benchmark-harness --exclude kreuzberg-pdfium-render
         platforms: [windows]
 
   build:profiling:

crates/kreuzberg-cli/src/commands/extract.rs

@@ -104,13 +104,16 @@ pub fn apply_extraction_overrides(
                 Some("easyocr") => ("easyocr", "en"),
                 _ => ("tesseract", "eng"),
             };
+            // Preserve existing paddle_ocr_config and element_config from config file/inline JSON
+            let existing_paddle_config = config.ocr.as_ref().and_then(|o| o.paddle_ocr_config.clone());
+            let existing_element_config = config.ocr.as_ref().and_then(|o| o.element_config.clone());
             config.ocr = Some(OcrConfig {
                 backend: backend.to_string(),
                 language: language.to_string(),
                 tesseract_config: None,
                 output_format: None,
-                paddle_ocr_config: None,
-                element_config: None,
+                paddle_ocr_config: existing_paddle_config,
+                element_config: existing_element_config,
             });
         } else {
             config.ocr = None;