signalpilot-ai-internal 0.5.0__tar.gz → 0.6.0__tar.gz

{signalpilot_ai_internal-0.5.0 → signalpilot_ai_internal-0.6.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: signalpilot_ai_internal
-Version: 0.5.0
+Version: 0.6.0
 Summary: SignalPilot Agent - Your Jupyter Notebook Assistant
 Project-URL: Homepage, https://signalpilot.ai/
 Author-email: Daniel Schaffield <danielschaffield@gmail.com>
@@ -50,7 +50,9 @@ Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Requires-Python: >=3.10
 Requires-Dist: jupyter-server<3,>=2.4.0
+Requires-Dist: openpyxl
 Requires-Dist: pandas>=2.0.0
+Requires-Dist: pyarrow
 Requires-Dist: schema-search
 Provides-Extra: test
 Requires-Dist: coverage; extra == 'test'

{signalpilot_ai_internal-0.5.0 → signalpilot_ai_internal-0.6.0}/package.json

@@ -1,6 +1,6 @@
 {
   "name": "signalpilot-ai-internal",
-  "version": "0.5.0",
+  "version": "0.6.0",
   "description": "SignalPilot Agent - Your Jupyter Notebook Assistant",
   "keywords": [
     "jupyter",

{signalpilot_ai_internal-0.5.0 → signalpilot_ai_internal-0.6.0}/pyproject.toml

@@ -25,7 +25,9 @@ classifiers = [
 dependencies = [
     "jupyter_server>=2.4.0,<3",
     "pandas>=2.0.0",
-    "schema-search"
+    "pyarrow",
+    "schema-search",
+    "openpyxl"
 ]
 dynamic = ["version", "description", "authors", "urls", "keywords"]
 

{signalpilot_ai_internal-0.5.0 → signalpilot_ai_internal-0.6.0}/signalpilot_ai_internal/_version.py

@@ -1,4 +1,4 @@
 # This file is auto-generated by Hatchling. As such, do not:
 #   - modify
 #   - track in version control e.g. be sure to add to .gitignore
-__version__ = VERSION = '0.5.0'
+__version__ = VERSION = '0.6.0'

{signalpilot_ai_internal-0.5.0 → signalpilot_ai_internal-0.6.0}/signalpilot_ai_internal/cache_service.py

@@ -3,6 +3,7 @@ Persistent caching service for SignalPilot AI.
 Handles OS-specific cache directory management and robust file operations.
 """
 
+import hashlib
 import json
 import os
 import platform
@@ -12,7 +13,7 @@ import threading
 import time
 import uuid
 from pathlib import Path
-from typing import Any, Dict, Optional
+from typing import Any, Dict, List, Optional
 
 
 class CacheDirectoryManager:
@@ -537,8 +538,150 @@ class PersistentCacheService:
         return info
 
 
+class FileScanCacheManager:
+    """Dedicated cache manager for file scanning operations with individual file caching"""
+    
+    def __init__(self):
+        self.cache_dir = CacheDirectoryManager.find_usable_cache_directory()
+        self.file_scans_dir = None
+        self.scanned_directories_file = None
+        self._lock = threading.RLock()
+        
+        if self.cache_dir:
+            self.file_scans_dir = self.cache_dir / "file_scans"
+            self.scanned_directories_file = self.file_scans_dir / "scanned_directories.json"
+            
+            # Initialize file scans directory
+            try:
+                self.file_scans_dir.mkdir(parents=True, exist_ok=True)
+                print(f"File scan cache directory initialized: {self.file_scans_dir}")
+            except Exception as e:
+                print(f"ERROR: Failed to create file scans directory: {e}")
+                self.file_scans_dir = None
+                self.scanned_directories_file = None
+        else:
+            print("WARNING: File scan cache manager running without persistent storage!")
+    
+    def _get_file_cache_path(self, file_path: str) -> Path:
+        """Generate cache file path using MD5 hash of absolute file path"""
+        if not self.file_scans_dir:
+            raise ValueError("File scans directory not available")
+        
+        # Create hash from absolute file path
+        file_hash = hashlib.md5(file_path.encode()).hexdigest()
+        return self.file_scans_dir / f"{file_hash}.json"
+    
+    def is_available(self) -> bool:
+        """Check if file scan cache is available"""
+        return self.file_scans_dir is not None and self.file_scans_dir.exists()
+    
+    def get_file_entry(self, file_path: str) -> Optional[Dict[str, Any]]:
+        """Read individual file cache entry"""
+        if not self.is_available():
+            return None
+        
+        try:
+            cache_path = self._get_file_cache_path(file_path)
+            return RobustFileOperations.safe_read_json(cache_path, None)
+        except Exception as e:
+            print(f"Error reading file cache for {file_path}: {e}")
+            return None
+    
+    def set_file_entry(self, file_path: str, entry: Dict[str, Any]) -> bool:
+        """Write individual file cache entry"""
+        if not self.is_available():
+            return False
+        
+        try:
+            cache_path = self._get_file_cache_path(file_path)
+            return RobustFileOperations.safe_write_json(cache_path, entry)
+        except Exception as e:
+            print(f"Error writing file cache for {file_path}: {e}")
+            return False
+    
+    def delete_file_entry(self, file_path: str) -> bool:
+        """Delete individual file cache entry"""
+        if not self.is_available():
+            return False
+        
+        try:
+            cache_path = self._get_file_cache_path(file_path)
+            if cache_path.exists():
+                cache_path.unlink()
+                return True
+            return True  # File doesn't exist, consider it deleted
+        except Exception as e:
+            print(f"Error deleting file cache for {file_path}: {e}")
+            return False
+    
+    def get_scanned_directories(self) -> List[Dict[str, Any]]:
+        """Read scanned directories list"""
+        if not self.scanned_directories_file:
+            return []
+        
+        try:
+            return RobustFileOperations.safe_read_json(self.scanned_directories_file, [])
+        except Exception as e:
+            print(f"Error reading scanned directories: {e}")
+            return []
+    
+    def set_scanned_directories(self, directories: List[Dict[str, Any]]) -> bool:
+        """Write scanned directories list"""
+        if not self.scanned_directories_file:
+            return False
+        
+        try:
+            return RobustFileOperations.safe_write_json(self.scanned_directories_file, directories)
+        except Exception as e:
+            print(f"Error writing scanned directories: {e}")
+            return False
+    
+    def clear_all_file_entries(self) -> bool:
+        """Clear all cached file entries"""
+        if not self.is_available():
+            return False
+        
+        try:
+            # Remove all .json files except scanned_directories.json
+            for cache_file in self.file_scans_dir.glob("*.json"):
+                if cache_file.name != "scanned_directories.json":
+                    cache_file.unlink()
+            return True
+        except Exception as e:
+            print(f"Error clearing file entries: {e}")
+            return False
+    
+    def get_cache_stats(self) -> Dict[str, Any]:
+        """Return cache statistics"""
+        stats = {
+            "available": self.is_available(),
+            "file_scans_directory": str(self.file_scans_dir) if self.file_scans_dir else None,
+            "total_file_entries": 0,
+            "total_cache_size": 0,
+            "scanned_directories_count": 0
+        }
+        
+        if self.is_available():
+            try:
+                # Count file entries (excluding scanned_directories.json)
+                file_entries = list(self.file_scans_dir.glob("*.json"))
+                file_entries = [f for f in file_entries if f.name != "scanned_directories.json"]
+                stats["total_file_entries"] = len(file_entries)
+                stats["total_cache_size"] = sum(f.stat().st_size for f in file_entries if f.exists())
+                
+                # Count scanned directories
+                directories = self.get_scanned_directories()
+                stats["scanned_directories_count"] = len(directories)
+                
+            except Exception as e:
+                stats["error"] = str(e)
+        
+        return stats
+
+
 # Global cache service instance
 _cache_service = None
+_file_scan_cache_manager = None
 
 
 def get_cache_service() -> PersistentCacheService:
@@ -547,3 +690,11 @@ def get_cache_service() -> PersistentCacheService:
     if _cache_service is None:
         _cache_service = PersistentCacheService()
     return _cache_service
+
+
+def get_file_scan_cache_manager() -> FileScanCacheManager:
+    """Get the global file scan cache manager instance"""
+    global _file_scan_cache_manager
+    if _file_scan_cache_manager is None:
+        _file_scan_cache_manager = FileScanCacheManager()
+    return _file_scan_cache_manager