Import ~2000 common concepts

manage_concepts.py

@@ -0,0 +1,87 @@
+#!/usr/bin/env python3
+"""
+Script for managing SFOSR concepts (export/import).
+"""
+
+import argparse
+import sys
+import os
+
+# Ensure the script can find the sfosr_core module
+script_dir = os.path.dirname(os.path.abspath(__file__))
+project_root = script_dir # Assumes script is in the root
+sys.path.insert(0, project_root)
+
+try:
+    from sfosr_core.sfosr_database import SFOSRDatabase
+except ImportError as e:
+    print(f"Error: Could not import SFOSRDatabase. Make sure sfosr_core is accessible.")
+    print(f"PYTHONPATH: {sys.path}")
+    sys.exit(1)
+
+DEFAULT_DB_PATH = "sfosr.db"
+DEFAULT_JSON_PATH_EXPORT = "concepts_export.json" # Default for export
+DEFAULT_JSON_PATH_IMPORT = "concepts_to_import.json" # Default for import
+
+def main():
+    parser = argparse.ArgumentParser(description="Manage SFOSR concepts.")
+    subparsers = parser.add_subparsers(dest='action', help='Action to perform', required=True)
+
+    # Export command
+    parser_export = subparsers.add_parser('export', help='Export concepts to JSON')
+    parser_export.add_argument(
+        '--db-path', 
+        default=DEFAULT_DB_PATH, 
+        help=f"Path to the SFOSR database file (default: {DEFAULT_DB_PATH})"
+    )
+    parser_export.add_argument(
+        '--json-path', 
+        default=DEFAULT_JSON_PATH_EXPORT, 
+        help=f"Path to the output JSON file (default: {DEFAULT_JSON_PATH_EXPORT})"
+    )
+
+    # Import command
+    parser_import = subparsers.add_parser('import', help='Import concepts from JSON (adds only new concepts)')
+    parser_import.add_argument(
+        '--db-path', 
+        default=DEFAULT_DB_PATH, 
+        help=f"Path to the SFOSR database file (default: {DEFAULT_DB_PATH})"
+    )
+    parser_import.add_argument(
+        '--json-path', 
+        default=DEFAULT_JSON_PATH_IMPORT, 
+        help=f"Path to the input JSON file (default: {DEFAULT_JSON_PATH_IMPORT})"
+    )
+
+    args = parser.parse_args()
+
+    if not os.path.exists(args.db_path):
+        print(f"Error: Database file not found at {args.db_path}")
+        sys.exit(1)
+        
+    db = SFOSRDatabase(db_path=args.db_path)
+
+    if args.action == 'export':
+        print(f"Exporting concepts from {args.db_path} to {args.json_path}...")
+        db.export_concepts_to_json(args.json_path)
+    
+    elif args.action == 'import':
+        if not os.path.exists(args.json_path):
+             print(f"Error: Import JSON file not found at {args.json_path}")
+             sys.exit(1)
+             
+        print(f"Importing concepts from {args.json_path} to {args.db_path}...")
+        try:
+            added, skipped = db.import_concepts_from_json(args.json_path)
+            # Optional: Add more feedback based on counts
+        except AttributeError:
+            print(f"Error: import_concepts_from_json method not found in SFOSRDatabase class in {db.__class__.__module__}.")
+            sys.exit(1)
+        except Exception as e:
+            print(f"An unexpected error occurred during import: {e}")
+            import traceback
+            traceback.print_exc() # Print stack trace for unexpected errors
+            sys.exit(1)
+
+if __name__ == "__main__":
+    main() 

sfosr.db

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3ab8d2f8291292b3d49448099b3b085e72a83825fd8b20a37e534355495777a8
-size 118784
+oid sha256:2dced7d5030004004ac3532c01117075552aa621e9caa6396e7f6b8ea45bf248
+size 319488

sfosr_core/sfosr_database.py

@@ -449,4 +449,89 @@ class SFOSRDatabase:
                 }
                 vectors.append(vector)
             
-            return vectors 
+            return vectors 
+
+    def export_concepts_to_json(self, filepath: str) -> None:
+        """
+        Экспортирует все концепты из БД в JSON файл.
+        
+        Args:
+            filepath: Путь к файлу для сохранения JSON.
+        """
+        concepts = self.get_all_concepts() # Reuse existing method
+        try:
+            with open(filepath, 'w', encoding='utf-8') as f:
+                json.dump(concepts, f, ensure_ascii=False, indent=2)
+            print(f"Concepts successfully exported to {filepath}")
+        except IOError as e:
+            print(f"Error writing to file {filepath}: {e}")
+        except Exception as e:
+            print(f"An unexpected error occurred during export: {e}") 
+
+    def import_concepts_from_json(self, filepath: str) -> Tuple[int, int]:
+        """
+        Импортирует концепты из JSON файла, добавляя только те, которых нет в БД.
+        
+        Args:
+            filepath: Путь к JSON файлу со списком концептов.
+                      Каждый концепт должен быть словарем с ключами 
+                      'name', 'description', 'domain', 'level'. 'id' игнорируется.
+                      
+        Returns:
+            Кортеж (added_count, skipped_count): Количество добавленных и пропущенных концептов.
+        """
+        added_count = 0
+        skipped_count = 0
+        try:
+            with open(filepath, 'r', encoding='utf-8') as f:
+                concepts_to_import = json.load(f)
+            
+            if not isinstance(concepts_to_import, list):
+                print(f"Error: JSON file {filepath} should contain a list of concepts.")
+                return 0, 0
+
+            existing_names = self.get_all_concept_names() # Получаем существующие имена
+            
+            with self as conn: # Используем одно соединение для всех вставок
+                cursor = conn.cursor()
+                for concept in concepts_to_import:
+                    name = concept.get('name')
+                    description = concept.get('description')
+                    domain = concept.get('domain')
+                    level = concept.get('level')
+                    
+                    if not all([name, description, domain, level]):
+                        print(f"Warning: Skipping concept due to missing fields: {concept}")
+                        skipped_count += 1
+                        continue
+                        
+                    if name in existing_names:
+                        # print(f"Skipping existing concept: {name}") # Optional: uncomment for verbose logging
+                        skipped_count += 1
+                    else:
+                        try:
+                            cursor.execute("""
+                                INSERT INTO concepts (name, description, domain, level)
+                                VALUES (?, ?, ?, ?)
+                            """, (name, description, domain, level))
+                            added_count += 1
+                            existing_names.add(name) # Добавляем в сет, чтобы не добавить дубль из самого файла
+                        except sqlite3.Error as e:
+                            print(f"Error inserting concept {name}: {e}")
+                            skipped_count += 1
+                            # Optionally rollback or handle error differently
+                
+                conn.commit() # Коммитим все изменения разом
+
+            print(f"Import finished. Added: {added_count}, Skipped (existing or invalid): {skipped_count}")
+            return added_count, skipped_count
+
+        except FileNotFoundError:
+            print(f"Error: Import file not found at {filepath}")
+            return 0, 0
+        except json.JSONDecodeError as e:
+            print(f"Error decoding JSON from {filepath}: {e}")
+            return 0, 0
+        except Exception as e:
+            print(f"An unexpected error occurred during import: {e}")
+            return 0, 0