fix: resolve OCR track converter data structure mismatch

**Problem**: OCR track was producing empty output files (0 pages, 0 elements)
despite successful OCR extraction (27 text regions detected).

**Root Causes**:
1. Converter expected `text_regions` inside `layout_data`, but
   `process_file_traditional` returns it at top level
2. Converter expected `ocr_dimensions` to be a list, but single-page
   documents return it as dict `{'width': W, 'height': H}`

**Solution**:
- Add `_extract_from_traditional_ocr()` method to handle top-level
  `text_regions` structure from `process_file_traditional`
- Handle both dict (single-page) and list (multi-page) formats for
  `ocr_dimensions`
- Update `_extract_pages()` to check for `text_regions` key before
  `layout_data` key

**Verification**:
- Before: img1.png → 0 pages, 0 elements, 0 characters
- After: img1.png → 1 page, 27 elements, 278 characters
- Output files now properly generated (JSON: 13KB, MD: 498B, PDF: 23KB)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

backend/app/services/ocr_to_unified_converter.py

@@ -84,7 +84,7 @@ class OCRToUnifiedConverter:
         except Exception as e:
             logger.error(f"Error converting OCR results: {e}")
             import traceback
-            traceback.print_exc()
+            logger.error(f"Traceback: {traceback.format_exc()}")
 
             # Return minimal document with error
             return UnifiedDocument(
@@ -127,6 +127,9 @@ class OCRToUnifiedConverter:
         # Check if we have enhanced results from PPStructureEnhanced
         if 'enhanced_results' in ocr_results:
             pages = self._extract_from_enhanced_results(ocr_results['enhanced_results'])
+        # Check for traditional OCR results with text_regions at top level (from process_file_traditional)
+        elif 'text_regions' in ocr_results:
+            pages = self._extract_from_traditional_ocr(ocr_results)
         # Check for traditional layout_data structure
         elif 'layout_data' in ocr_results:
             pages = self._extract_from_layout_data(ocr_results['layout_data'])
@@ -237,6 +240,96 @@ class OCRToUnifiedConverter:
 
         return pages
 
+    def _extract_from_traditional_ocr(self, ocr_results: Dict[str, Any]) -> List[Page]:
+        """
+        Extract pages from traditional OCR results (process_file_traditional).
+
+        This handles the structure where text_regions and images_metadata are at
+        the top level of ocr_results, not nested inside layout_data.
+        """
+        pages = []
+
+        # Get text regions and page dimensions
+        text_regions = ocr_results.get('text_regions', [])
+        ocr_dimensions = ocr_results.get('ocr_dimensions', [])
+        total_pages = ocr_results.get('total_pages', 1)
+
+        # Group elements by page
+        elements_by_page = {}
+
+        # Process text regions
+        for text_region in text_regions:
+            page_num = text_region.get('page', 1)
+            if page_num not in elements_by_page:
+                elements_by_page[page_num] = []
+
+            element = self._convert_text_region(text_region)
+            if element:
+                elements_by_page[page_num].append(element)
+
+        # Process images
+        for img_meta in ocr_results.get('images_metadata', []):
+            page_num = img_meta.get('page', 1)
+            if page_num not in elements_by_page:
+                elements_by_page[page_num] = []
+
+            element = self._convert_image_metadata(img_meta)
+            if element:
+                elements_by_page[page_num].append(element)
+
+        # Process tables from layout_data if available
+        if 'layout_data' in ocr_results and isinstance(ocr_results['layout_data'], dict):
+            for table_data in ocr_results['layout_data'].get('tables', []):
+                page_num = table_data.get('page', 1)
+                if page_num not in elements_by_page:
+                    elements_by_page[page_num] = []
+
+                element = self._convert_table_data(table_data)
+                if element:
+                    elements_by_page[page_num].append(element)
+
+        # Create pages
+        max_page = max(elements_by_page.keys()) if elements_by_page else total_pages
+        for page_num in range(1, max_page + 1):
+            elements = elements_by_page.get(page_num, [])
+
+            # Get page dimensions
+            # Handle both dict (single page) and list (multiple pages) formats
+            if isinstance(ocr_dimensions, dict):
+                # Single page format: {'width': W, 'height': H}
+                page_width = ocr_dimensions.get('width', 0)
+                page_height = ocr_dimensions.get('height', 0)
+            elif isinstance(ocr_dimensions, list):
+                # Multi-page format: [{'page': 1, 'width': W, 'height': H}, ...]
+                page_dims = next((d for d in ocr_dimensions if isinstance(d, dict) and d.get('page') == page_num), None)
+                if page_dims:
+                    page_width = page_dims.get('width', 0)
+                    page_height = page_dims.get('height', 0)
+                else:
+                    page_width = 0
+                    page_height = 0
+            else:
+                # Default dimensions if not available
+                page_width = 0
+                page_height = 0
+
+            # Determine reading order based on position
+            reading_order = self._calculate_reading_order(elements)
+
+            page = Page(
+                page_number=page_num,
+                dimensions=Dimensions(
+                    width=page_width,
+                    height=page_height
+                ),
+                elements=elements,
+                metadata={'reading_order': reading_order}
+            )
+
+            pages.append(page)
+
+        return pages
+
     def _convert_pp3_element(
         self,
         elem_data: Dict[str, Any],