spiderforce4ai 0.1.6__tar.gz → 0.1.7__tar.gz

{spiderforce4ai-0.1.6 → spiderforce4ai-0.1.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: spiderforce4ai
-Version: 0.1.6
+Version: 0.1.7
 Summary: Python wrapper for SpiderForce4AI HTML-to-Markdown conversion service
 Home-page: https://petertam.pro
 Author: Piotr Tamulewicz

{spiderforce4ai-0.1.6 → spiderforce4ai-0.1.7}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "spiderforce4ai"
-version = "0.1.6"
+version = "0.1.7"
 description = "Python wrapper for SpiderForce4AI HTML-to-Markdown conversion service"
 readme = "README.md"
 authors = [{name = "Piotr Tamulewicz", email = "pt@petertam.pro"}]

{spiderforce4ai-0.1.6 → spiderforce4ai-0.1.7}/setup.py

@@ -3,7 +3,7 @@ from setuptools import setup, find_packages
 
 setup(
     name="spiderforce4ai",
-    version="0.1.6",
+    version="0.1.7",
     author="Piotr Tamulewicz",
     author_email="pt@petertam.pro",
     description="Python wrapper for SpiderForce4AI HTML-to-Markdown conversion service",

{spiderforce4ai-0.1.6 → spiderforce4ai-0.1.7}/spiderforce4ai/__init__.py

@@ -86,6 +86,31 @@ class CrawlConfig:
             payload["remove_selectors_regex"] = self.remove_selectors_regex
         return payload
 
+
+def _send_webhook_sync(result: CrawlResult, config: CrawlConfig) -> None:
+    """Synchronous version of webhook sender for parallel processing."""
+    if not config.webhook_url:
+        return
+
+    payload = {
+        "url": result.url,
+        "status": result.status,
+        "markdown": result.markdown if result.status == "success" else None,
+        "error": result.error if result.status == "failed" else None,
+        "timestamp": result.timestamp,
+        "config": config.to_dict()
+    }
+
+    try:
+        response = requests.post(
+            config.webhook_url,
+            json=payload,
+            timeout=config.webhook_timeout
+        )
+        response.raise_for_status()
+    except Exception as e:
+        print(f"Warning: Failed to send webhook for {result.url}: {str(e)}")
+
 # Module level function for multiprocessing
 def _process_url_parallel(args: Tuple[str, str, CrawlConfig]) -> CrawlResult:
     """Process a single URL for parallel processing."""
@@ -99,12 +124,15 @@ def _process_url_parallel(args: Tuple[str, str, CrawlConfig]) -> CrawlResult:
         
         response = requests.post(endpoint, json=payload, timeout=config.timeout)
         if response.status_code != 200:
-            return CrawlResult(
+            result = CrawlResult(
                 url=url,
                 status="failed",
                 error=f"HTTP {response.status_code}: {response.text}",
                 config=config.to_dict()
             )
+            # Send webhook for failed result
+            _send_webhook_sync(result, config)
+            return result
         
         markdown = response.text
         
@@ -114,24 +142,32 @@ def _process_url_parallel(args: Tuple[str, str, CrawlConfig]) -> CrawlResult:
             with open(filepath, 'w', encoding='utf-8') as f:
                 f.write(markdown)
         
-        # Add delay if configured
-        if config.request_delay:
-            time.sleep(config.request_delay)
-        
-        return CrawlResult(
+        result = CrawlResult(
             url=url,
             status="success",
             markdown=markdown,
             config=config.to_dict()
         )
+        
+        # Send webhook for successful result
+        _send_webhook_sync(result, config)
+        
+        # Add delay if configured
+        if config.request_delay:
+            time.sleep(config.request_delay)
+        
+        return result
             
     except Exception as e:
-        return CrawlResult(
+        result = CrawlResult(
             url=url,
             status="failed",
             error=str(e),
             config=config.to_dict()
         )
+        # Send webhook for error result
+        _send_webhook_sync(result, config)
+        return result
 
 class SpiderForce4AI:
     """Main class for interacting with SpiderForce4AI service."""
@@ -424,4 +460,5 @@ class SpiderForce4AI:
 
     def __exit__(self, exc_type, exc_val, exc_tb):
         """Sync context manager exit."""
-        self._executor.shutdown(wait=True)
+        self._executor.shutdown(wait=True)
+

{spiderforce4ai-0.1.6 → spiderforce4ai-0.1.7}/spiderforce4ai.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: spiderforce4ai
-Version: 0.1.6
+Version: 0.1.7
 Summary: Python wrapper for SpiderForce4AI HTML-to-Markdown conversion service
 Home-page: https://petertam.pro
 Author: Piotr Tamulewicz