spiderforce4ai 2.6__py3-none-any.whl → 2.6.3__py3-none-any.whl

spiderforce4ai/post_extraction_agent.py

@@ -156,19 +156,34 @@ class PostExtractionAgent:
         self.buffer = PostExtractionBuffer(config.buffer_file)
         self.results: Dict[str, Any] = {}
         self.rate_limiter = RateLimiter()
+        # Convert string path to Path object if needed
+        if isinstance(self.config.output_file, str):
+            self.config.output_file = Path(self.config.output_file)
         self._setup_output()
         
     def _setup_output(self) -> None:
         """Setup output file if combining results."""
         if self.config.combine_output and self.config.output_file:
+            # Ensure parent directory exists
             self.config.output_file.parent.mkdir(parents=True, exist_ok=True)
+            
+            # Load existing results if file exists
             if self.config.output_file.exists():
-                # Backup existing file
-                backup_path = self.config.output_file.with_suffix(f".bak_{int(time.time())}")
-                self.config.output_file.rename(backup_path)
-            self.config.output_file.touch()
-            # Initialize empty results dictionary
-            self.results = {}
+                try:
+                    with open(self.config.output_file, 'r') as f:
+                        self.results = json.load(f)
+                except json.JSONDecodeError:
+                    # If file is corrupted, backup and start fresh
+                    backup_path = self.config.output_file.with_suffix(f".bak_{int(time.time())}")
+                    self.config.output_file.rename(backup_path)
+                    self.results = {}
+            
+            # Create file if it doesn't exist
+            if not self.config.output_file.exists():
+                self.config.output_file.touch()
+                self.results = {}
+            
+            logger.info(f"Initialized output file at {self.config.output_file}")
 
     def _process_single_content(self, url: str, content: str) -> Optional[Dict]:
         """Process a single piece of content through the LLM."""
@@ -263,28 +278,38 @@ class PostExtractionAgent:
                 # Convert Path to string if needed
                 output_file = str(self.config.output_file) if isinstance(self.config.output_file, Path) else self.config.output_file
                 
-                # Load existing results if file exists
-                if Path(output_file).exists():
-                    try:
-                        with open(output_file, 'r') as f:
+                # Ensure output directory exists
+                Path(output_file).parent.mkdir(parents=True, exist_ok=True)
+                
+                # Load existing results or create new
+                try:
+                    if Path(output_file).exists():
+                        with open(output_file, 'r', encoding='utf-8') as f:
                             self.results = json.load(f)
-                    except json.JSONDecodeError:
+                    else:
                         self.results = {}
+                except (json.JSONDecodeError, FileNotFoundError):
+                    self.results = {}
                 
                 # Update results with new data
                 self.results[url] = result
                 
-                # Ensure output directory exists
-                Path(output_file).parent.mkdir(parents=True, exist_ok=True)
-                
                 # Save combined results atomically
                 temp_file = f"{output_file}.tmp"
-                with open(temp_file, 'w') as f:
-                    json.dump(self.results, f, indent=2)
+                with open(temp_file, 'w', encoding='utf-8') as f:
+                    json.dump(self.results, f, indent=2, ensure_ascii=False)
                 
-                # Atomic replace
+                # Atomic replace and cleanup backup files
                 Path(temp_file).replace(output_file)
-                logger.info(f"Updated combined results file with {url}")
+                logger.info(f"Updated combined results file with {url} in {output_file}")
+                
+                # Cleanup all backup files
+                for backup_file in Path(output_file).parent.glob(f"{Path(output_file).stem}.bak_*"):
+                    try:
+                        backup_file.unlink()
+                        logger.info(f"Cleaned up backup file: {backup_file}")
+                    except Exception as e:
+                        logger.warning(f"Failed to remove backup file {backup_file}: {e}")
                 
                 # Cleanup backup files
                 for backup_file in Path(output_file).parent.glob(f"{Path(output_file).stem}.bak_*"):
@@ -345,18 +370,20 @@ class PostExtractionAgent:
                 
                 # Save result synchronously
                 try:
-                    # Save both original and transformed result
-                    if self.config.custom_transform_function:
-                        self._save_result_sync(url, transformed_result)
-                        logger.info(f"Saved transformed results for {url}")
+                    # Always save the result, whether transformed or original
+                    result_to_save = transformed_result if self.config.custom_transform_function else result
+                    if self.config.combine_output and self.config.output_file:
+                        self._save_result_sync(url, result_to_save)
+                        logger.info(f"Saved results for {url} to combined output")
                     else:
-                        self._save_result_sync(url, result)
-                        logger.info(f"Saved original results for {url}")
+                        self._save_result_sync(url, result_to_save)
+                        logger.info(f"Saved individual result for {url}")
                 except Exception as e:
                     error_msg = f"Error saving results for {url}: {str(e)}"
                     logger.error(error_msg)
                     console.print(f"[red]{error_msg}[/red]")
                 
+                # Return the appropriate result
                 return transformed_result if self.config.custom_transform_function else result
             
             # Wait before retry

{spiderforce4ai-2.6.dist-info → spiderforce4ai-2.6.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: spiderforce4ai
-Version: 2.6
+Version: 2.6.3
 Summary: Python wrapper for SpiderForce4AI HTML-to-Markdown conversion service with LLM post-processing
 Home-page: https://petertam.pro
 Author: Piotr Tamulewicz

spiderforce4ai-2.6.3.dist-info/RECORD

@@ -0,0 +1,7 @@
+spiderforce4ai/__init__.py,sha256=BlrRIrtpDUHjmDedqgXP1KbAAsAH9vwFPncUR5VGGyM,44804
+spiderforce4ai/post_extraction_agent.py,sha256=3HB54YrPCbQTUoZWINr7yHvwXwQywmq0f_RBJwKr2gg,19355
+spiderforce4ai-2.6.3.dist-info/METADATA,sha256=L5GCJHggqks18Z31ru5DbDdXT3mdS8pYEDhfdR9igms,9012
+spiderforce4ai-2.6.3.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
+spiderforce4ai-2.6.3.dist-info/entry_points.txt,sha256=ibARQxOlDiL1ho12zbDZt4Uq5RKSIk_qk159ZlZ46hc,59
+spiderforce4ai-2.6.3.dist-info/top_level.txt,sha256=Kth7A21Js7DCp0j5XBBi-FE45SCLouZkeNZU__Yr9Yk,15
+spiderforce4ai-2.6.3.dist-info/RECORD,,

spiderforce4ai-2.6.dist-info/RECORD

@@ -1,7 +0,0 @@
-spiderforce4ai/__init__.py,sha256=BlrRIrtpDUHjmDedqgXP1KbAAsAH9vwFPncUR5VGGyM,44804
-spiderforce4ai/post_extraction_agent.py,sha256=AysrHIoD-IreqbvWqCDxyN7v8EPSdLOG9yxABamTZSg,17827
-spiderforce4ai-2.6.dist-info/METADATA,sha256=JtIZ1-ojRvfm773-yF1a_M_x6eB5kbnb6WT5XT04KDA,9010
-spiderforce4ai-2.6.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
-spiderforce4ai-2.6.dist-info/entry_points.txt,sha256=ibARQxOlDiL1ho12zbDZt4Uq5RKSIk_qk159ZlZ46hc,59
-spiderforce4ai-2.6.dist-info/top_level.txt,sha256=Kth7A21Js7DCp0j5XBBi-FE45SCLouZkeNZU__Yr9Yk,15
-spiderforce4ai-2.6.dist-info/RECORD,,