Gracefully handle missing PaddleOCRVL

python_api/ocr_api_server.py

@@ -16,7 +16,11 @@ import logging
 import traceback
 from pathlib import Path
 from flask import Flask, request, jsonify
-from paddleocr import PaddleOCR, PaddleOCRVL
+from paddleocr import PaddleOCR
+try:
+    from paddleocr import PaddleOCRVL  # type: ignore
+except Exception:
+    PaddleOCRVL = None
 
 # 配置日志
 logging.basicConfig(
@@ -93,6 +97,10 @@ def init_models():
     logger.info("=" * 60)
     logger.info("正在初始化 PaddleOCRVL...")
     logger.info("=" * 60)
+    if PaddleOCRVL is None:
+        logger.warning("PaddleOCRVL not available in installed paddleocr. Skipping VL pipeline.")
+        vl_pipeline = None
+    else:
         try:
             vl_kwargs = {
                 "use_seal_recognition": True,
@@ -220,10 +228,11 @@ def ocr_pdf():
 
         # 处理 PDF（传递 verbose 参数）
         try:
+            ocr_model = 'paddleocr_vl' if vl_pipeline else 'ppocr_v5'
             result = process_single_pdf_standalone(
                 Path(pdf_path),
                 Path(output_dir),
-                ocr_model='paddleocr_vl',
+                ocr_model=ocr_model,
                 vl_pipeline=vl_pipeline,
                 verbose=verbose  # 新增：传递verbose参数
             )
@@ -298,12 +307,13 @@ def ocr_image():
             return jsonify({'success': False, 'error': f'Image not found: {image_path}'}), 404
 
         logger.info(f"处理图像: {image_path}")
-
+        # Use PaddleOCRVL if available, otherwise fallback to PaddleOCR
-        # 使用 PaddleOCRVL 识别
+        if vl_pipeline:
-        if not vl_pipeline:
-            return jsonify({'success': False, 'error': 'PaddleOCRVL not initialized'}), 500
-
             result = vl_pipeline.ocr(image_path)
+        else:
+            if not ocr_pipeline:
+                return jsonify({'success': False, 'error': 'OCR pipeline not initialized'}), 500
+            result = ocr_pipeline.ocr(image_path)
 
         # 提取文本
         texts = []

src/main/resources/python-api/ocr_api_server.py

@@ -15,7 +15,11 @@ import json
 import logging
 from pathlib import Path
 from flask import Flask, request, jsonify
-from paddleocr import PaddleOCR, PaddleOCRVL
+from paddleocr import PaddleOCR
+try:
+    from paddleocr import PaddleOCRVL  # type: ignore
+except Exception:
+    PaddleOCRVL = None
 
 # 配置日志
 logging.basicConfig(
@@ -92,6 +96,10 @@ def init_models():
     logger.info("=" * 60)
     logger.info("正在初始化 PaddleOCRVL...")
     logger.info("=" * 60)
+    if PaddleOCRVL is None:
+        logger.warning("PaddleOCRVL not available in installed paddleocr. Skipping VL pipeline.")
+        vl_pipeline = None
+    else:
         try:
             vl_kwargs = {
                 "use_seal_recognition": True,
@@ -219,10 +227,11 @@ def ocr_pdf():
 
         # 处理 PDF（传递 verbose 参数）
         try:
+            ocr_model = 'paddleocr_vl' if vl_pipeline else 'ppocr_v5'
             result = process_single_pdf_standalone(
                 Path(pdf_path),
                 Path(output_dir),
-                ocr_model='paddleocr_vl',
+                ocr_model=ocr_model,
                 vl_pipeline=vl_pipeline,
                 verbose=verbose  # 新增：传递verbose参数
             )
@@ -280,12 +289,13 @@ def ocr_image():
             return jsonify({'success': False, 'error': f'Image not found: {image_path}'}), 404
 
         logger.info(f"处理图像: {image_path}")
-
+        # Use PaddleOCRVL if available, otherwise fallback to PaddleOCR
-        # 使用 PaddleOCRVL 识别
+        if vl_pipeline:
-        if not vl_pipeline:
-            return jsonify({'success': False, 'error': 'PaddleOCRVL not initialized'}), 500
-
             result = vl_pipeline.ocr(image_path)
+        else:
+            if not ocr_pipeline:
+                return jsonify({'success': False, 'error': 'OCR pipeline not initialized'}), 500
+            result = ocr_pipeline.ocr(image_path)
 
         # 提取文本
         texts = []