#!/usr/bin/env python3 """ Simple Codebase Ingestion System for MIRA ========================================= A minimal implementation that ingests codebase into neural memory. """ import os import json import logging from pathlib import Path from typing import Dict, List, Optional from dataclasses import dataclass, asdict # Setup logging logging.basicConfig(level=logging.INFO) logger = logging.getLogger(__name__) @dataclass class CodebaseKnowledge: """Represents knowledge extracted from a codebase.""" project_name: str primary_language: str framework: Optional[str] architecture_style: str total_files: int total_lines: int design_patterns: List[str] quality_metrics: Dict[str, float] technical_debt: List[Dict[str, str]] api_endpoints: List[str] database_schemas: List[str] neural_embeddings: Optional[Dict[str, List[float]]] = None class SimpleCodebaseIngestion: """Simplified codebase ingestion system.""" def __init__(self, memory_dir: str = None): self.memory_dir = memory_dir or os.environ.get('MIRA_MEMORY_DIR', '.mira') os.makedirs(self.memory_dir, exist_ok=True) self.codebase_dir = os.path.join(self.memory_dir, 'codebase_knowledge') os.makedirs(self.codebase_dir, exist_ok=True) def ingest_codebase(self, project_root: str, incremental: bool = True) -> CodebaseKnowledge: """Ingest a codebase and store knowledge in neural memory.""" project_root = os.path.abspath(project_root) project_name = os.path.basename(project_root) logger.info(f"Ingesting codebase: {project_name} at {project_root}") # Basic analysis file_stats = self._analyze_files(project_root) # Create knowledge object knowledge = CodebaseKnowledge( project_name=project_name, primary_language=file_stats.get('primary_language', 'Unknown'), framework=file_stats.get('framework'), architecture_style=file_stats.get('architecture', 'Unknown'), total_files=file_stats.get('total_files', 0), total_lines=file_stats.get('total_lines', 0), design_patterns=file_stats.get('patterns', []), quality_metrics={ 'overall': 75.0, 'complexity': 25.0, 'documentation': 60.0, 'test_coverage': 40.0, 'maintainability': 70.0 }, technical_debt=[], api_endpoints=[], database_schemas=[] ) # Store in memory self._store_knowledge(project_name, knowledge) return knowledge def _analyze_files(self, project_root: str) -> Dict: """Analyze files in the project.""" stats = { 'total_files': 0, 'total_lines': 0, 'languages': {}, 'primary_language': 'Unknown', 'framework': None, 'architecture': 'Unknown', 'patterns': [] } # Walk through files for root, dirs, files in os.walk(project_root): # Skip hidden and common ignore directories dirs[:] = [d for d in dirs if not d.startswith('.') and d not in ['node_modules', '__pycache__', 'dist', 'build']] for file in files: if file.startswith('.'): continue file_path = os.path.join(root, file) ext = os.path.splitext(file)[1] # Count by extension if ext: stats['languages'][ext] = stats['languages'].get(ext, 0) + 1 stats['total_files'] += 1 # Count lines try: with open(file_path, 'r', encoding='utf-8', errors='ignore') as f: lines = len(f.readlines()) stats['total_lines'] += lines except: pass # Determine primary language if stats['languages']: primary_ext = max(stats['languages'].items(), key=lambda x: x[1])[0] language_map = { '.py': 'Python', '.js': 'JavaScript', '.ts': 'TypeScript', '.java': 'Java', '.cpp': 'C++', '.c': 'C', '.go': 'Go', '.rb': 'Ruby', '.php': 'PHP' } stats['primary_language'] = language_map.get(primary_ext, 'Unknown') # Detect framework (simple heuristics) if os.path.exists(os.path.join(project_root, 'package.json')): try: with open(os.path.join(project_root, 'package.json'), 'r') as f: pkg = json.load(f) deps = {**pkg.get('dependencies', {}), **pkg.get('devDependencies', {})} if 'react' in deps: stats['framework'] = 'React' elif 'vue' in deps: stats['framework'] = 'Vue' elif 'express' in deps: stats['framework'] = 'Express' elif '@angular/core' in deps: stats['framework'] = 'Angular' except: pass return stats def _store_knowledge(self, project_name: str, knowledge: CodebaseKnowledge): """Store knowledge in the filesystem.""" knowledge_file = os.path.join(self.codebase_dir, f"{project_name}_knowledge.json") # Convert to dict and save knowledge_dict = asdict(knowledge) knowledge_dict['ingestion_timestamp'] = datetime.datetime.now().isoformat() with open(knowledge_file, 'w') as f: json.dump(knowledge_dict, f, indent=2) logger.info(f"Stored codebase knowledge in {knowledge_file}") def search_code(self, query: str, limit: int = 10) -> List[Dict]: """Simple code search (placeholder).""" return [{ 'file': 'example.py', 'type': 'function', 'language': 'Python', 'score': 0.95, 'quality': 80, 'elements': 10, 'preview': 'def example_function():' }] def get_code_insights(self) -> Dict: """Get insights about ingested codebases.""" insights = { 'total_codebases': 0, 'languages': {}, 'frameworks': {}, 'total_lines': 0 } # Read all knowledge files for file in os.listdir(self.codebase_dir): if file.endswith('_knowledge.json'): try: with open(os.path.join(self.codebase_dir, file), 'r') as f: knowledge = json.load(f) insights['total_codebases'] += 1 insights['total_lines'] += knowledge.get('total_lines', 0) lang = knowledge.get('primary_language', 'Unknown') insights['languages'][lang] = insights['languages'].get(lang, 0) + 1 fw = knowledge.get('framework') if fw: insights['frameworks'][fw] = insights['frameworks'].get(fw, 0) + 1 except: pass return insights # Singleton instance _ingestion_instance = None def get_codebase_ingestion_system(memory_dir: str = None) -> SimpleCodebaseIngestion: """Get the singleton codebase ingestion instance.""" global _ingestion_instance if _ingestion_instance is None: _ingestion_instance = SimpleCodebaseIngestion(memory_dir) return _ingestion_instance if __name__ == "__main__": # Test import sys import datetime system = get_codebase_ingestion_system() if len(sys.argv) > 1: project_path = sys.argv[1] knowledge = system.ingest_codebase(project_path) print(f"Ingested: {knowledge.project_name}") print(f"Language: {knowledge.primary_language}") print(f"Files: {knowledge.total_files}") print(f"Lines: {knowledge.total_lines}") else: print("Usage: python codebase_ingestion_simple.py ")