#!/usr/bin/env python3 # -*- coding: utf-8 -*- """ 水利智学系统 - 系统大脑 V1.0 功能:24小时后台运行,全面分析所有代码,建立知识图谱 设计:顶层设计 + 所有文件联动 + 一个大脑调动 """ import os import re import json import time import hashlib import requests from pathlib import Path from datetime import datetime from typing import List, Dict, Set, Tuple # ===== 配置 ===== OLLAMA_URL = "http://localhost:11434/api/chat" MODEL_NAME = "qwen2.5:7b" PROJECT_ROOT = r"F:\水利智学系统\可运行版(解压即用,双击start.bat)" BACKEND_ROOT = os.path.join(PROJECT_ROOT, "backend") OUTPUT_DIR = os.path.join(PROJECT_ROOT, "系统大脑") # 创建输出目录 os.makedirs(OUTPUT_DIR, exist_ok=True) # 要分析的目录 ANALYZE_DIRS = [ ("旧API", "api/v1"), ("新模块", "modules"), ("core", "core"), ("utils", "utils"), ("models", "models"), ("crud", "crud"), ] # ===== 工具函数 ===== def file_hash(file_path: str) -> str: """计算文件哈希值,用于增量分析""" try: with open(file_path, 'rb') as f: return hashlib.md5(f.read()).hexdigest() except: return "" def get_all_py_files(directory: str) -> List[str]: """获取目录下所有.py文件""" root = Path(BACKEND_ROOT) / directory if not root.exists(): return [] files = [] for f in root.rglob("*.py"): if "__pycache__" in str(f): continue files.append(str(f)) return sorted(files) def count_lines(file_path: str) -> int: """统计文件行数""" try: with open(file_path, 'r', encoding='utf-8') as f: return sum(1 for _ in f) except: try: with open(file_path, 'r', encoding='gbk') as f: return sum(1 for _ in f) except: return 0 def extract_imports(content: str) -> Set[str]: """提取文件中的import依赖""" imports = set() # import xxx for match in re.finditer(r'^import\s+([\w.]+)', content, re.MULTILINE): imports.add(match.group(1)) # from xxx import yyy for match in re.finditer(r'^from\s+([\w.]+)\s+import', content, re.MULTILINE): imports.add(match.group(1)) return imports def extract_functions(file_path: str) -> List[Dict]: """提取文件中的所有函数和类,带行号""" try: with open(file_path, 'r', encoding='utf-8') as f: content = f.read() except: try: with open(file_path, 'r', encoding='gbk') as f: content = f.read() except: return [] functions = [] # 匹配类定义 class_pattern = re.compile(r'^class\s+(\w+)\s*[\(\:]', re.MULTILINE) for match in class_pattern.finditer(content): class_name = match.group(1) start_line = content[:match.start()].count('\n') + 1 functions.append({ "type": "class", "name": class_name, "line": start_line, }) # 匹配函数定义 func_pattern = re.compile(r'^(\s*)def\s+(\w+)\s*\(', re.MULTILINE) for match in func_pattern.finditer(content): indent = len(match.group(1)) func_name = match.group(2) start_line = content[:match.start()].count('\n') + 1 if indent <= 4: functions.append({ "type": "method" if indent > 0 else "function", "name": func_name, "line": start_line, "indent": indent, }) return functions # ===== AI分析 ===== def analyze_file_with_ai(file_path: str, functions: List[Dict], dir_name: str) -> Dict: """调用AI分析单个文件""" try: with open(file_path, 'r', encoding='utf-8') as f: content = f.read() except: try: with open(file_path, 'r', encoding='gbk') as f: content = f.read() except: return {"error": "无法读取文件"} # 限制内容长度 max_chars = 8000 if len(content) > max_chars: content = content[:max_chars] + "\n...(截断)" func_list = "\n".join([f"- {f['type']}: {f['name']} (行{f['line']})" for f in functions]) prompt = f"""你是90名顶级专家组成的评审团,对一造备考系统进行全面架构评审。 请从300个维度分析以下代码文件。 模块:{dir_name} 路径:{file_path} 函数清单:{func_list} 代码: ```python {content} ``` 输出JSON,包含10大板块: 【1. 文件基本信息】file_info: file_name, file_path, file_role, business_domain, core_value, file_complexity, file_size, importance, maturity, maintenance_cost 【2. 函数分析】functions[]: name, line, purpose, business_logic, priority, input_params, output, dependencies, error_handling, performance_impact, security_risk, complexity, testability, maintainability, documentation, parameter_validation, return_spec, exception_handling, logging, caching, transaction, concurrency 【3. 业务逻辑】business_analysis: business_rules, user_scenario, input_data, output_data, business_flow, boundary_conditions, exception_scenarios, compliance_requirements, user_experience, data_consistency, business_value, business_priority, business_impact, business_dependency, business_risk, business_change, business_integration, business_report, business_statistics, business_audit 【4. 技术架构】architecture: design_pattern, layered_architecture, modularity, coupling, cohesion, extensibility, reusability, testability, deployability, observability, maintainability, performance_bottleneck, security_arch, data_arch, integration_arch 【5. 新旧对比】comparison: in_new_module, missing_features, extra_features, duplicate_code, code_differences, performance_comparison, security_comparison, maintainability_comparison, extensibility_comparison, ux_comparison, data_consistency_comparison, api_compatibility, db_compatibility, config_compatibility, deploy_compatibility 【6. 融合方案】fusion_plan: fusion_difficulty, fusion_risk, fusion_strategy, fusion_steps, fusion_time, fusion_cost, fusion_benefit, fusion_priority, fusion_dependency, fusion_validation 【7. 用户+教育】user_education: ux_learnability, ux_efficiency, ux_satisfaction, ux_accessibility, ux_mobile_friendly, ux_loading_speed, ux_error_recovery, ux_feedback, ux_navigation, ux_visual_design, learning_effectiveness, retention_rate, mastery_measurement, spaced_repetition, active_recall, feedback_quality, adaptation_level, progress_tracking, motivation_design, outcome_prediction, exam_alignment, question_quality, coverage_rate, weak_point_detection, review_schedule, practice_mode, mock_exam, time_management, stress_management, score_prediction 【8. 运维安全】ops_security: backup_strategy, disaster_recovery, monitoring_alert, log_management, performance_monitoring, security_testing, data_encryption, access_control, compliance_check, vulnerability_scan, deployment_ease, rollback_capability, scalability, resource_usage, cost_efficiency 【9. 战略价值】strategic_value: competitive_advantage, market_fit, differentiation, scalability, monetization_potential, user_growth, brand_value, ecosystem_fit, tech_debt_level, innovation_potential, risk_management, compliance_readiness, internationalization, ai_readiness, data_value, user_insight, product_roadmap, team_capability, time_to_market, roi_potential 【10. 顶层设计+文档】top_design: system_positioning, target_users, core_value_prop, competitive_edge, development_roadmap, requirement_doc, design_doc, architecture_doc, api_doc, deploy_doc, tech_stack_selection, team_collaboration, project_management, quality_assurance, knowledge_management 【11. 用户设计理念】user_design_philosophy: core_motivation, usage_habit, aesthetic_preference, feature_priority, tech_background, pain_points, learning_philosophy, exam_philosophy, product_philosophy, user_philosophy, design_style, interaction_preference, information_density, feedback_style, error_handling_preference 只返回JSON。""" try: response = requests.post( OLLAMA_URL, json={ "model": MODEL_NAME, "messages": [{"role": "user", "content": prompt}], "stream": False, "options": {"temperature": 0.1} }, timeout=120 ) result = response.json() ai_text = result['message']['content'].strip() start = ai_text.find('{') end = ai_text.rfind('}') + 1 if start >= 0 and end > start: return json.loads(ai_text[start:end]) else: return {"error": "AI返回格式错误"} except Exception as e: return {"error": str(e)} # ===== 主流程 ===== def main(): print("=" * 80) print("水利智学系统 - 系统大脑 V1.0") print("24小时后台分析引擎") print("=" * 80) print(f"启动时间: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}") print() # 加载已有分析结果 knowledge_file = os.path.join(OUTPUT_DIR, "系统知识图谱.json") if os.path.exists(knowledge_file): with open(knowledge_file, 'r', encoding='utf-8') as f: knowledge = json.load(f) print(f"已加载历史分析: {len(knowledge.get('files', {}))} 个文件") else: knowledge = { "created_at": datetime.now().isoformat(), "updated_at": datetime.now().isoformat(), "files": {}, "dependencies": {}, "business_domains": {}, } # 扫描所有文件 print("\n【第一步】扫描所有文件...") all_files = {} total_lines = 0 for dir_name, dir_path in ANALYZE_DIRS: files = get_all_py_files(dir_path) all_files[dir_name] = files dir_lines = sum(count_lines(f) for f in files) total_lines += dir_lines print(f" {dir_name}: {len(files)} 个文件, {dir_lines:,} 行") print(f"\n总计: {sum(len(v) for v in all_files.values())} 个文件, {total_lines:,} 行") # 分析新文件或变化的文件 print("\n【第二步】增量分析变化的文件...") analyzed_count = 0 skipped_count = 0 for dir_name, files in all_files.items(): print(f"\n--- {dir_name} ---") for file_path in files: file_name = os.path.relpath(file_path, BACKEND_ROOT) file_hash_val = file_hash(file_path) # 检查是否需要重新分析 if file_name in knowledge["files"]: old_hash = knowledge["files"][file_name].get("hash", "") if old_hash == file_hash_val: skipped_count += 1 continue # 分析文件 functions = extract_functions(file_path) if not functions: knowledge["files"][file_name] = { "hash": file_hash_val, "lines": count_lines(file_path), "file_summary": "无函数", "functions": [] } continue print(f" 分析: {os.path.basename(file_name)} ({len(functions)}个函数)") result = analyze_file_with_ai(file_path, functions, dir_name) # 保存结果 knowledge["files"][file_name] = { "hash": file_hash_val, "lines": count_lines(file_path), "dir": dir_name, "analyzed_at": datetime.now().isoformat(), **result } analyzed_count += 1 # 保存中间结果 if analyzed_count % 5 == 0: knowledge["updated_at"] = datetime.now().isoformat() with open(knowledge_file, 'w', encoding='utf-8') as f: json.dump(knowledge, f, ensure_ascii=False, indent=2) print(f" -> 已保存中间结果") # 避免请求太快 time.sleep(0.5) print(f"\n分析完成: 新分析{analyzed_count}个文件, 跳过{skipped_count}个文件") # 生成依赖关系图 print("\n【第三步】生成依赖关系图...") generate_dependency_graph(knowledge) # 生成业务域分布 print("\n【第四步】生成业务域分布...") generate_business_domains(knowledge) # 保存最终结果 knowledge["updated_at"] = datetime.now().isoformat() with open(knowledge_file, 'w', encoding='utf-8') as f: json.dump(knowledge, f, ensure_ascii=False, indent=2) # 生成融合报告 print("\n【第五步】生成融合报告...") generate_fusion_report(knowledge) print("\n" + "=" * 80) print(f"分析完成!结果保存在: {OUTPUT_DIR}") print(f"完成时间: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}") print("=" * 80) def generate_dependency_graph(knowledge: Dict): """生成文件依赖关系图""" deps_file = os.path.join(OUTPUT_DIR, "文件依赖关系.md") with open(deps_file, 'w', encoding='utf-8') as f: f.write("# 文件依赖关系图\n\n") f.write(f"生成时间: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}\n\n") # 按目录分组 by_dir = {} for file_name, file_info in knowledge["files"].items(): dir_name = file_info.get("dir", "其他") if dir_name not in by_dir: by_dir[dir_name] = [] by_dir[dir_name].append(file_name) for dir_name, files in sorted(by_dir.items()): f.write(f"## {dir_name}\n\n") for file_name in sorted(files): file_info = knowledge["files"][file_name] f.write(f"- `{file_name}` ({file_info.get('lines', 0)}行)\n") f.write("\n") def generate_business_domains(knowledge: Dict): """生成业务域分布""" domain_file = os.path.join(OUTPUT_DIR, "业务域分布.md") domains = {} for file_name, file_info in knowledge["files"].items(): domain = file_info.get("business_domain", "未分类") if domain not in domains: domains[domain] = [] domains[domain].append({ "file": file_name, "summary": file_info.get("file_summary", "") }) with open(domain_file, 'w', encoding='utf-8') as f: f.write("# 业务域分布\n\n") f.write(f"生成时间: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}\n\n") for domain, files in sorted(domains.items()): f.write(f"## {domain} ({len(files)}个文件)\n\n") for item in files: f.write(f"- `{item['file']}`: {item['summary']}\n") f.write("\n") def generate_fusion_report(knowledge: Dict): """生成融合报告""" report_file = os.path.join(OUTPUT_DIR, "融合执行方案.md") p0_tasks = [] p1_tasks = [] missing_features = [] for file_name, file_info in knowledge["files"].items(): funcs = file_info.get("functions", []) for func in funcs: if func.get("priority") == "P0": p0_tasks.append({ "file": file_name, "function": func.get("name"), "line": func.get("line"), "purpose": func.get("purpose"), "action": func.get("fusion_action") }) elif func.get("priority") == "P1": p1_tasks.append({ "file": file_name, "function": func.get("name"), "line": func.get("line"), "purpose": func.get("purpose"), "action": func.get("fusion_action") }) missing = file_info.get("missing_in_new", "") if missing: missing_features.append({ "file": file_name, "missing": missing }) with open(report_file, 'w', encoding='utf-8') as f: f.write("# 融合执行方案\n\n") f.write(f"生成时间: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}\n\n") f.write("## 一、P0核心任务(必须立即融合)\n\n") f.write(f"共{len(p0_tasks)}个任务\n\n") for i, task in enumerate(p0_tasks, 1): f.write(f"{i}. **{task['function']}** (行{task['line']})\n") f.write(f" - 文件: `{task['file']}`\n") f.write(f" - 功能: {task['purpose']}\n") f.write(f" - 动作: {task['action']}\n\n") f.write("---\n\n") f.write("## 二、P1重要任务(本周完成)\n\n") f.write(f"共{len(p1_tasks)}个任务\n\n") for i, task in enumerate(p1_tasks, 1): f.write(f"{i}. **{task['function']}** (行{task['line']})\n") f.write(f" - 文件: `{task['file']}`\n") f.write(f" - 功能: {task['purpose']}\n") f.write(f" - 动作: {task['action']}\n\n") f.write("---\n\n") f.write("## 三、新模块缺失功能\n\n") for item in missing_features: f.write(f"- `{item['file']}`: {item['missing']}\n") if __name__ == "__main__": main()