paddleocr-skills 1.0.0 → 1.1.0

package/templates/{paddleocr-vl/scripts/paddleocr-vl → paddleocr-vl-1.5/scripts/paddleocr-vl-1.5}/smoke_test.py

@@ -1,199 +1,199 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-"""
-Smoke Test for PaddleOCR-VL
-
-Verifies that VL_API_URL and VL_TOKEN are correctly configured
-and that the API is accessible.
-
-Usage:
-    python scripts/paddleocr-vl/smoke_test.py
-    python scripts/paddleocr-vl/smoke_test.py --test-url "URL"  # Optional custom test
-"""
-
-import argparse
-import json
-import sys
-from pathlib import Path
-
-# Add current directory to Python path
-script_dir = Path(__file__).parent
-sys.path.insert(0, str(script_dir))
-
-
-def main():
-    parser = argparse.ArgumentParser(description='PaddleOCR-VL smoke test')
-    parser.add_argument(
-        '--test-url',
-        help='Optional: Custom document URL for testing'
-    )
-    parser.add_argument(
-        '--skip-api-test',
-        action='store_true',
-        help='Skip API connectivity test, only check configuration'
-    )
-    args = parser.parse_args()
-
-    print("=" * 60)
-    print("PaddleOCR-VL - Smoke Test")
-    print("=" * 60)
-    print()
-
-    # Check configuration
-    print("\n[1/3] Checking configuration...")
-
-    from _lib import Config
-
-    try:
-        Config.load_env()
-
-        # Get config values
-        api_url = Config.get_vl_api_url()
-        token = Config.get_vl_token()
-        timeout_ms = Config.get_timeout_ms()
-        max_retry = Config.get_max_retry()
-        cache_ttl = Config.get_cache_ttl_sec()
-
-        print(f"+ VL_API_URL: {api_url}")
-
-        # Mask token for display
-        if len(token) > 12:
-            masked_token = token[:8] + "..." + token[-4:]
-        else:
-            masked_token = "***"
-        print(f"+ VL_TOKEN: {masked_token}")
-        print(f"+ Timeout: {timeout_ms}ms")
-        print(f"+ Max Retry: {max_retry}")
-        print(f"+ Cache TTL: {cache_ttl}s")
-        print()
-
-    except Exception as e:
-        print(f"\nConfiguration error: {e}")
-        return 1
-
-    # Check imports and dependencies
-    print("[2/3] Checking dependencies...")
-    print()
-
-    try:
-        import httpx
-        print(f"+ httpx: {httpx.__version__}")
-
-        from dotenv import load_dotenv
-        print("+ python-dotenv: installed")
-
-        from _lib import VLClient, QualityEvaluator, SimpleCache
-        print("+ VLClient: OK")
-        print("+ QualityEvaluator: OK")
-        print("+ SimpleCache: OK")
-        print()
-
-    except ImportError as e:
-        print(f"X Dependency error: {e}")
-        print()
-        print("Please install dependencies:")
-        print("  pip install -r scripts/paddleocr-vl/requirements.txt")
-        print()
-        return 1
-
-    # Test API connectivity
-    if args.skip_api_test:
-        print("[3/3] Skipping API connectivity test (--skip-api-test)")
-        print()
-        print("=" * 60)
-        print("Configuration Check Complete!")
-        print("=" * 60)
-        print()
-        print("Your PaddleOCR-VL is configured correctly.")
-        print()
-        print("To test with a real document, run:")
-        print('  python scripts/paddleocr-vl/vl_caller.py --file-url "URL"')
-        print()
-        return 0
-
-    print("[3/3] Testing API connectivity...")
-    print()
-
-    # Use provided test URL or default
-    test_url = args.test_url or "https://paddleocr.bj.bcebos.com/dataset/document_layout_sample.png"
-
-    print(f"Test document: {test_url}")
-    print("Calling PaddleOCR-VL API...")
-    print()
-
-    from _lib import make_api_request, QualityEvaluator
-
-    try:
-        result = make_api_request(
-            file_url=test_url,
-            timeout_ms=timeout_ms,
-            use_cache=False  # Force fresh request for testing
-        )
-
-        if not result.get("ok", False):
-            error = result.get("error", {})
-            print(f"X API call failed: {error.get('message', 'Unknown error')}")
-            print(f"   Error code: {error.get('code', 'UNKNOWN')}")
-            print()
-            return 1
-
-        print("+ API call successful!")
-        print()
-
-        # Evaluate quality
-        quality = QualityEvaluator.evaluate(result)
-
-        print("Response Quality:")
-        print(f"  - Overall Confidence: {quality['overall_confidence']:.2f} / 1.00")
-        print(f"  - Quality Level: {quality['quality_level']}")
-
-        if quality.get('region_stats'):
-            stats = quality['region_stats']
-            print(f"  - Regions Detected: {stats.get('total_regions', 0)}")
-
-            if stats.get('by_type'):
-                print("  - Region Types:")
-                for region_type, count in sorted(stats['by_type'].items()):
-                    print(f"      {region_type}: {count}")
-
-        if quality.get('warnings'):
-            print("\n  Warnings:")
-            for warning in quality['warnings']:
-                print(f"    ! {warning}")
-
-        print()
-
-        # Show sample content
-        result_data = result.get("result", {})
-        full_text = result_data.get("full_text", "")
-
-        if full_text:
-            preview_length = min(200, len(full_text))
-            print(f"Content Preview ({preview_length} chars):")
-            print(f"  {full_text[:preview_length]}...")
-            print()
-
-    except Exception as e:
-        print(f"X API call failed: {e}")
-        print()
-        return 1
-
-    # Success
-    print("=" * 60)
-    print("All Tests Passed!")
-    print("=" * 60)
-    print()
-    print("Your PaddleOCR-VL setup is working correctly.")
-    print()
-    print("Next steps:")
-    print('  - Parse documents: python scripts/paddleocr-vl/vl_caller.py --file-url "URL"')
-    print('  - Show quality info: Add --show-quality flag')
-    print('  - Pretty JSON: Add --pretty flag')
-    print()
-
-    return 0
-
-
-if __name__ == '__main__':
-    exit_code = main()
-    sys.exit(exit_code)
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Smoke Test for PaddleOCR-VL 1.5
+
+Verifies that VL_API_URL and VL_TOKEN are correctly configured
+and that the API is accessible.
+
+Usage:
+    python scripts/paddleocr-vl-1.5/smoke_test.py
+    python scripts/paddleocr-vl-1.5/smoke_test.py --test-url "URL"  # Optional custom test
+"""
+
+import argparse
+import json
+import sys
+from pathlib import Path
+
+# Add current directory to Python path
+script_dir = Path(__file__).parent
+sys.path.insert(0, str(script_dir))
+
+
+def main():
+    parser = argparse.ArgumentParser(description='PaddleOCR-VL 1.5 smoke test')
+    parser.add_argument(
+        '--test-url',
+        help='Optional: Custom document URL for testing'
+    )
+    parser.add_argument(
+        '--skip-api-test',
+        action='store_true',
+        help='Skip API connectivity test, only check configuration'
+    )
+    args = parser.parse_args()
+
+    print("=" * 60)
+    print("PaddleOCR-VL 1.5 - Smoke Test")
+    print("=" * 60)
+    print()
+
+    # Check configuration
+    print("\n[1/3] Checking configuration...")
+
+    from _lib import Config
+
+    try:
+        Config.load_env()
+
+        # Get config values
+        api_url = Config.get_vl_api_url()
+        token = Config.get_vl_token()
+        timeout_ms = Config.get_timeout_ms()
+        max_retry = Config.get_max_retry()
+        cache_ttl = Config.get_cache_ttl_sec()
+
+        print(f"+ VL_API_URL: {api_url}")
+
+        # Mask token for display
+        if len(token) > 12:
+            masked_token = token[:8] + "..." + token[-4:]
+        else:
+            masked_token = "***"
+        print(f"+ VL_TOKEN: {masked_token}")
+        print(f"+ Timeout: {timeout_ms}ms")
+        print(f"+ Max Retry: {max_retry}")
+        print(f"+ Cache TTL: {cache_ttl}s")
+        print()
+
+    except Exception as e:
+        print(f"\nConfiguration error: {e}")
+        return 1
+
+    # Check imports and dependencies
+    print("[2/3] Checking dependencies...")
+    print()
+
+    try:
+        import httpx
+        print(f"+ httpx: {httpx.__version__}")
+
+        from dotenv import load_dotenv
+        print("+ python-dotenv: installed")
+
+        from _lib import VLClient, QualityEvaluator, SimpleCache
+        print("+ VLClient: OK")
+        print("+ QualityEvaluator: OK")
+        print("+ SimpleCache: OK")
+        print()
+
+    except ImportError as e:
+        print(f"X Dependency error: {e}")
+        print()
+        print("Please install dependencies:")
+        print("  pip install -r scripts/paddleocr-vl-1.5/requirements.txt")
+        print()
+        return 1
+
+    # Test API connectivity
+    if args.skip_api_test:
+        print("[3/3] Skipping API connectivity test (--skip-api-test)")
+        print()
+        print("=" * 60)
+        print("Configuration Check Complete!")
+        print("=" * 60)
+        print()
+        print("Your PaddleOCR-VL 1.5 is configured correctly.")
+        print()
+        print("To test with a real document, run:")
+        print('  python scripts/paddleocr-vl-1.5/vl_caller.py --file-url "URL"')
+        print()
+        return 0
+
+    print("[3/3] Testing API connectivity...")
+    print()
+
+    # Use provided test URL or default
+    test_url = args.test_url or "https://paddleocr.bj.bcebos.com/dataset/document_layout_sample.png"
+
+    print(f"Test document: {test_url}")
+    print("Calling PaddleOCR-VL 1.5 API...")
+    print()
+
+    from _lib import make_api_request, QualityEvaluator
+
+    try:
+        result = make_api_request(
+            file_url=test_url,
+            timeout_ms=timeout_ms,
+            use_cache=False  # Force fresh request for testing
+        )
+
+        if not result.get("ok", False):
+            error = result.get("error", {})
+            print(f"X API call failed: {error.get('message', 'Unknown error')}")
+            print(f"   Error code: {error.get('code', 'UNKNOWN')}")
+            print()
+            return 1
+
+        print("+ API call successful!")
+        print()
+
+        # Evaluate quality
+        quality = QualityEvaluator.evaluate(result)
+
+        print("Response Quality:")
+        print(f"  - Overall Confidence: {quality['overall_confidence']:.2f} / 1.00")
+        print(f"  - Quality Level: {quality['quality_level']}")
+
+        if quality.get('region_stats'):
+            stats = quality['region_stats']
+            print(f"  - Regions Detected: {stats.get('total_regions', 0)}")
+
+            if stats.get('by_type'):
+                print("  - Region Types:")
+                for region_type, count in sorted(stats['by_type'].items()):
+                    print(f"      {region_type}: {count}")
+
+        if quality.get('warnings'):
+            print("\n  Warnings:")
+            for warning in quality['warnings']:
+                print(f"    ! {warning}")
+
+        print()
+
+        # Show sample content
+        result_data = result.get("result", {})
+        full_text = result_data.get("full_text", "")
+
+        if full_text:
+            preview_length = min(200, len(full_text))
+            print(f"Content Preview ({preview_length} chars):")
+            print(f"  {full_text[:preview_length]}...")
+            print()
+
+    except Exception as e:
+        print(f"X API call failed: {e}")
+        print()
+        return 1
+
+    # Success
+    print("=" * 60)
+    print("All Tests Passed!")
+    print("=" * 60)
+    print()
+    print("Your PaddleOCR-VL 1.5 setup is working correctly.")
+    print()
+    print("Next steps:")
+    print('  - Parse documents: python scripts/paddleocr-vl-1.5/vl_caller.py --file-url "URL"')
+    print('  - Show quality info: Add --show-quality flag')
+    print('  - Pretty JSON: Add --pretty flag')
+    print()
+
+    return 0
+
+
+if __name__ == '__main__':
+    exit_code = main()
+    sys.exit(exit_code)