README.md

@@ -0,0 +1,224 @@
+# AI Code Quality Scanner - 飞书通知版
+
+一个自动化代码质量扫描系统，在代码提交时自动扫描并发送报告到飞书。
+
+## 功能特性
+
+- 🤖 自动监听 Gitea 代码提交事件
+- 🔍 多维度代码质量扫描（语法、风格、安全）
+- 📊 生成 Markdown 格式扫描报告
+- 📱 实时推送飞书机器人通知
+
+## 系统架构
+
+```
+┌─────────────┐     Webhook      ┌──────────────────┐
+│   Gitea     │ ───────────────► │  Webhook Server  │
+│   代码仓库   │                  │    (Flask)       │
+└─────────────┘                  └────────┬─────────┘
+                                          │
+                                          ▼
+                                 ┌──────────────────┐
+                                 │  Code Scanner    │
+                                 │  - ESLint        │
+                                 │  - Pylint        │
+                                 │  - SonarQube     │
+                                 └────────┬─────────┘
+                                          │
+                                          ▼
+                                 ┌──────────────────┐
+                                 │  Report Generator│
+                                 │  - Markdown      │
+                                 └────────┬─────────┘
+                                          │
+                                          ▼
+                                 ┌──────────────────┐
+                                 │  Feishu Bot      │
+                                 │  - Webhook       │
+                                 └──────────────────┘
+```
+
+## 快速开始
+
+### 1. 安装依赖
+
+```bash
+pip install -r requirements.txt
+```
+
+### 2. 配置飞书机器人
+
+1. 打开飞书群聊 → 设置 → 群机器人
+2. 添加机器人 → 选择"自定义机器人"
+3. 获取 Webhook 地址
+4. 配置 `config.yaml`
+
+### 3. 配置 Gitea Webhook
+
+#### 方式一：Push 时扫描（原有方式）
+
+1. 进入 Gitea 仓库 → 设置 → Webhooks
+2. 添加 Webhook：
+   - 目标 URL: `http://你的服务器IP:5000/webhook/gitea`
+   - 触发事件: Push
+   - 密钥: 配置 `config.yaml` 中的 secret
+
+#### 方式二：PR 创建时扫描（推荐）
+
+1. 进入 Gitea 仓库 → 设置 → Webhooks
+2. 添加 Webhook：
+   - 目标 URL: `http://你的服务器IP:5000/webhook/gitea`
+   - 触发事件: Pull Request
+   - 密钥: 配置 `config.yaml` 中的 secret
+
+**支持的 PR 事件：**
+- `opened` - 创建新 PR
+- `reopened` - 重新打开 PR
+- `synchronize` - PR 中的提交有更新
+- `ready_for_review` - PR 标记为准备好审查
+
+### 4. 运行服务
+
+```bash
+python app.py
+```
+
+## 配置说明
+
+所有配置在 `config.yaml` 中：
+
+```yaml
+server:
+  host: "0.0.0.0"
+  port: 5000
+  debug: true
+
+gitea:
+  base_url: "http://localhost:3000"
+  # Webhook 签名密钥
+  webhook_secret: "your_webhook_secret"
+
+feishu:
+  # 飞书机器人 Webhook 地址
+  webhook_url: "https://open.feishu.cn/open-apis/bot/v2/hook/xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
+  # 消息推送 secret（可选，用于签名）
+  secret: "your_feishu_secret"
+
+scanner:
+  # 支持的语言
+  languages:
+    - python
+    - javascript
+    - typescript
+  # 扫描阈值
+  max_issues: 10
+  # 是否启用详细扫描
+  detailed: true
+
+report:
+  # 报告保存路径
+  output_dir: "./reports"
+  # 是否保留报告文件
+  keep_files: true
+```
+
+## 项目结构
+
+```
+code-scanner/
+├── app.py                    # 主应用入口
+├── config.yaml               # 配置文件
+├── requirements.txt          # Python 依赖
+├── README.md                 # 项目说明
+├── scanner/
+│   ├── __init__.py
+│   ├── base.py               # 扫描器基类
+│   ├── python_scanner.py     # Python 代码扫描
+│   ├── js_scanner.py         # JavaScript/TypeScript 扫描
+│   └── security_scanner.py   # 安全扫描
+├── report/
+│   ├── __init__.py
+│   └── generator.py          # Markdown 报告生成
+├── notify/
+│   ├── __init__.py
+│   └── feishu.py             # 飞书通知
+├── webhook/
+│   ├── __init__.py
+│   └── handler.py            # Webhook 处理
+└── reports/                  # 报告输出目录
+```
+
+## 支持的扫描工具
+
+### Python
+- **Pylint** - 代码风格和错误检查
+- **Flake8** - Python 代码检查
+- **Bandit** - 安全漏洞扫描
+
+### JavaScript/TypeScript
+- **ESLint** - JavaScript/TypeScript 检查
+- **Prettier** - 代码格式化
+
+## 飞书消息效果
+
+扫描完成后，将收到类似以下消息：
+
+### Push 扫描消息
+
+```
+📊 代码质量扫描报告
+
+仓库: my-project
+分支: main
+提交: abc1234
+提交者: developer@example.com
+
+✅ 扫描通过 (0 issues)
+或
+⚠️ 发现问题 (5 issues)
+```
+
+### PR 扫描消息
+
+```
+📊 PR 代码质量扫描报告
+
+仓库: my-project
+源分支: feature-xxx → 目标分支: main
+PR链接: https://gitea.example.com/user/project/pulls/123
+提交: abc1234
+提交者: developer@example.com
+
+✅ 扫描通过 (0 issues)
+或
+⚠️ 发现问题 (5 issues)
+```
+
+## Docker 部署
+
+```dockerfile
+FROM python:3.11-slim
+
+WORKDIR /app
+COPY requirements.txt .
+RUN pip install -r requirements.txt
+
+COPY . .
+EXPOSE 5000
+
+CMD ["python", "app.py"]
+```
+
+## 环境变量
+
+也可以通过环境变量配置：
+
+```bash
+export FEISHU_WEBHOOK_URL="https://open.feishu.cn/..."
+export GITEA_WEBHOOK_SECRET="secret"
+export SCANNER_MAX_ISSUES=10
+```
+
+## 许可证
+
+MIT License

app.py

@@ -16,6 +16,7 @@ from scanner.python_scanner import PythonScanner
 from scanner.js_scanner import JavaScriptScanner
 from scanner.security_scanner import SecurityScanner
 from scanner.ai_reviewer import AIReviewer
+from scanner.diff_parser import merge_issues_with_code
 from report.generator import ReportGenerator
 from notify.feishu import FeishuNotifier
 from gitea_client import GiteaClient
@@ -232,6 +233,23 @@ def handle_pull_request(payload: Dict[str, Any]) -> Tuple[Dict, int]:
                     clone_url, source_sha, source_branch
                 )
 
+            # 获取 PR 的代码差异，用于将问题与代码片段关联
+            pr_diff = None
+            if '/' in repo_name:
+                repo_owner, repo_name_only = repo_name.split('/', 1)
+            else:
+                repo_owner = 'Bosch_Demo'
+                repo_name_only = repo_name
+
+            try:
+                pr_diff = gitea_client.get_pull_request_diff(repo_owner, repo_name_only, pr_number)
+                logger.info(f"已获取 PR #{pr_number} 的 diff，长度: {len(pr_diff) if pr_diff else 0}")
+            except Exception as e:
+                logger.warning(f"获取 PR diff 失败: {e}")
+
+            # 将问题与代码片段关联
+            scan_details_with_code = merge_issues_with_code(scan_results, pr_diff or '')
+
             # 生成报告
             commit_message = f'PR #{pr_number}: {pr_title}'
             report = report_generator.generate(
@@ -259,7 +277,7 @@ def handle_pull_request(payload: Dict[str, Any]) -> Tuple[Dict, int]:
                 'target_branch': target_branch,
                 'author': author
             }
-            PRScanDB.save_pr_scan(pr_info_for_db, scan_results, report.get('file_path'))
+            PRScanDB.save_pr_scan(pr_info_for_db, scan_results, report.get('file_path'), scan_details_with_code)
 
             logger.info(f'PR #{pr_number} 扫描完成')
 
@@ -455,12 +473,146 @@ def api_get_pr(pr_id):
             except:
                 pass
 
+        # 返回带代码片段的扫描详情
+        if pr.get('scan_details_with_code') and isinstance(pr['scan_details_with_code'], str):
+            try:
+                pr['scan_details_with_code'] = json.loads(pr['scan_details_with_code'])
+            except:
+                pass
+
         return jsonify(pr)
     except Exception as e:
         logger.error(f'获取 PR 详情失败: {str(e)}')
         return jsonify({'error': str(e)}), 500
 
 
+@app.route('/api/prs/<int:pr_id>/diff')
+def api_get_pr_diff(pr_id):
+    """获取 PR 的代码差异"""
+    try:
+        pr = PRScanDB.get_pr_by_id(pr_id)
+        if not pr:
+            return jsonify({'error': 'PR not found'}), 404
+
+        repo_name = pr.get('repo_name', '')
+        pr_number = pr.get('pr_number', 0)
+
+        if not repo_name or not pr_number:
+            return jsonify({'error': 'PR 信息不完整'}), 400
+
+        # 解析 owner 和 repo
+        if '/' in repo_name:
+            owner, repo = repo_name.split('/', 1)
+        else:
+            owner = 'Bosch_Demo'  # 默认
+            repo = repo_name
+
+        logger.info(f"获取 PR #{pr_number} ({owner}/{repo}) 的 diff")
+
+        # 获取 diff
+        diff = gitea_client.get_pull_request_diff(owner, repo, pr_number)
+        if diff is None:
+            return jsonify({'error': '获取 diff 失败'}), 500
+
+        return jsonify({
+            'diff': diff,
+            'pr_number': pr_number,
+            'repo_name': repo_name
+        })
+
+    except Exception as e:
+        logger.error(f'获取 PR diff 失败: {str(e)}')
+        return jsonify({'error': str(e)}), 500
+
+
+@app.route('/api/prs/<int:pr_id>/files')
+def api_get_pr_files(pr_id):
+    """获取 PR 变更文件列表（用于左侧树状展示）"""
+    try:
+        pr = PRScanDB.get_pr_by_id(pr_id)
+        if not pr:
+            return jsonify({'error': 'PR not found'}), 404
+        repo_name = pr.get('repo_name', '')
+        pr_number = pr.get('pr_number', 0)
+        if not repo_name or not pr_number:
+            return jsonify({'error': 'PR 信息不完整'}), 400
+        if '/' in repo_name:
+            owner, repo = repo_name.split('/', 1)
+        else:
+            owner, repo = 'Bosch_Demo', repo_name
+        files = gitea_client.get_pull_request_files(owner, repo, pr_number)
+        if files is None:
+            return jsonify({'error': '获取文件列表失败'}), 500
+        return jsonify({'files': files, 'repo_name': repo_name})
+    except Exception as e:
+        logger.error(f'获取 PR 文件列表失败: {str(e)}')
+        return jsonify({'error': str(e)}), 500
+
+
+@app.route('/api/prs/<int:pr_id>/file')
+def api_get_pr_file_content(pr_id):
+    """获取 PR 中某文件在源分支上的完整内容"""
+    try:
+        path = request.args.get('path')
+        if not path:
+            return jsonify({'error': '缺少 path 参数'}), 400
+        pr = PRScanDB.get_pr_by_id(pr_id)
+        if not pr:
+            return jsonify({'error': 'PR not found'}), 404
+        repo_name = pr.get('repo_name', '')
+        pr_number = pr.get('pr_number', 0)
+        if not repo_name or not pr_number:
+            return jsonify({'error': 'PR 信息不完整'}), 400
+        if '/' in repo_name:
+            owner, repo = repo_name.split('/', 1)
+        else:
+            owner, repo = 'Bosch_Demo', repo_name
+        pr_info = gitea_client.get_pull_request(owner, repo, pr_number)
+        if not pr_info:
+            return jsonify({'error': '获取 PR 信息失败'}), 500
+        head_ref = pr_info.get('head', {}).get('ref') or pr_info.get('head_branch') or pr.get('source_branch')
+        if not head_ref:
+            return jsonify({'error': '无法确定源分支'}), 400
+        content = gitea_client.get_file_contents(owner, repo, path, head_ref)
+        if content is None:
+            return jsonify({'error': '文件不存在或无法读取'}), 404
+
+        # 获取该文件的扫描问题（PR 创建时已扫描并存入 scan_details_with_code）
+        scan_issues = []
+        path_norm = path.replace('\\', '/').strip()
+        scan_details = pr.get('scan_details_with_code')
+        if isinstance(scan_details, str):
+            try:
+                scan_details = json.loads(scan_details)
+            except Exception:
+                scan_details = None
+        if scan_details and scan_details.get('scanners'):
+            for scanner in scan_details['scanners']:
+                for issue in scanner.get('issues', []):
+                    issue_file = (issue.get('file') or '').replace('\\', '/').strip()
+                    if not issue_file:
+                        continue
+                    # 匹配：精确相等或一端包含另一端（兼容 basename 或完整路径）
+                    if path_norm == issue_file or path_norm.endswith(issue_file) or issue_file.endswith(path_norm):
+                        sev = (issue.get('severity') or 'info')
+                        if isinstance(sev, str):
+                            sev = sev.lower()
+                        scanner_name = scanner.get('name', '')
+                        scanner_display = {'python': 'Python', 'javascript': 'JavaScript', 'security': 'Security'}.get(scanner_name, scanner_name)
+                        scan_issues.append({
+                            'scanner': scanner_display,
+                            'severity': sev,
+                            'line': int(issue.get('line') or 0),
+                            'message': (issue.get('message') or issue.get('description') or '').strip(),
+                            'code_context': issue.get('code_context')
+                        })
+
+        return jsonify({'path': path, 'content': content, 'scan_issues': scan_issues})
+    except Exception as e:
+        logger.error(f'获取文件内容失败: {str(e)}')
+        return jsonify({'error': str(e)}), 500
+
+
 @app.route('/api/prs/<int:pr_id>/merge', methods=['POST'])
 def api_merge_pr(pr_id):
     """合并 PR"""

db.py

@@ -40,6 +40,7 @@ def init_db():
             state TEXT DEFAULT 'pending',
             scan_status TEXT DEFAULT 'pending',
             scan_result TEXT,
+            scan_details_with_code TEXT,
             issues_count INTEGER DEFAULT 0,
             security_issues INTEGER DEFAULT 0,
             ai_review TEXT,
@@ -72,16 +73,17 @@ class PRScanDB:
     """PR 扫描结果数据库操作类"""
     
     @staticmethod
-    def save_pr_scan(pr_info: Dict[str, Any], scan_results: Dict[str, Any], 
-                     report_path: str = None) -> int:
+    def save_pr_scan(pr_info: Dict[str, Any], scan_results: Dict[str, Any],
+                     report_path: str = None, scan_details_with_code: Dict = None) -> int:
         """
         保存 PR 扫描结果
-        
+
         Args:
             pr_info: PR 信息
             scan_results: 扫描结果
             report_path: 报告文件路径
-            
+            scan_details_with_code: 带代码片段的扫描详情
+
         Returns:
             扫描记录 ID
         """
@@ -116,6 +118,7 @@ class PRScanDB:
                     author = ?,
                     scan_status = ?,
                     scan_result = ?,
+                    scan_details_with_code = ?,
                     issues_count = ?,
                     security_issues = ?,
                     ai_review = ?,
@@ -129,6 +132,7 @@ class PRScanDB:
                 pr_info.get('author'),
                 'completed',
                 json.dumps(scan_results, ensure_ascii=False),
+                json.dumps(scan_details_with_code, ensure_ascii=False) if scan_details_with_code else None,
                 issues_count,
                 security_issues,
                 json.dumps(scan_results.get('ai', {}), ensure_ascii=False),
@@ -143,9 +147,9 @@ class PRScanDB:
                 INSERT INTO pr_scans (
                     pr_number, repo_name, pr_title, pr_url,
                     source_branch, target_branch, author,
-                    state, scan_status, scan_result,
+                    state, scan_status, scan_result, scan_details_with_code,
                     issues_count, security_issues, ai_review, report_path
-                ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
+                ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
             ''', (
                 pr_info.get('pr_number'),
                 pr_info.get('repo_name'),
@@ -157,6 +161,7 @@ class PRScanDB:
                 'open',
                 'completed',
                 json.dumps(scan_results, ensure_ascii=False),
+                json.dumps(scan_details_with_code, ensure_ascii=False) if scan_details_with_code else None,
                 issues_count,
                 security_issues,
                 json.dumps(scan_results.get('ai', {}), ensure_ascii=False),

gitea_client.py

@@ -6,7 +6,7 @@ Gitea API 客户端
 """
 import logging
 import requests
-from typing import Dict, Any, Optional
+from typing import Dict, Any, Optional, List
 
 logger = logging.getLogger(__name__)
 
@@ -167,3 +167,101 @@ class GiteaClient:
         if pr_info:
             return pr_info.get('mergeable', False) and pr_info.get('state') == 'open'
         return False
+
+    def get_pull_request_diff(self, owner: str, repo: str, pr_number: int) -> Optional[str]:
+        """
+        获取 Pull Request 的代码差异
+
+        Args:
+            owner: 仓库所有者
+            repo: 仓库名称
+            pr_number: PR 编号
+
+        Returns:
+            diff 文本，失败返回 None
+        """
+        url = f"{self.base_url}/api/v1/repos/{owner}/{repo}/pulls/{pr_number}.diff"
+
+        try:
+            response = requests.get(
+                url,
+                headers=self._get_headers(),
+                timeout=30
+            )
+
+            if response.status_code == 200:
+                logger.info(f"成功获取 PR #{pr_number} 的 diff")
+                return response.text
+            else:
+                logger.error(f"获取 PR #{pr_number} diff 失败: {response.status_code}")
+                return None
+
+        except Exception as e:
+            logger.error(f"获取 PR #{pr_number} diff 异常: {str(e)}")
+            return None
+
+    def get_pull_request_files(self, owner: str, repo: str, pr_number: int) -> Optional[List[Dict[str, Any]]]:
+        """
+        获取 PR 中修改的文件列表
+
+        Args:
+            owner: 仓库所有者
+            repo: 仓库名称
+            pr_number: PR 编号
+
+        Returns:
+            文件列表，失败返回 None
+        """
+        url = f"{self.base_url}/api/v1/repos/{owner}/{repo}/pulls/{pr_number}/files"
+
+        try:
+            response = requests.get(
+                url,
+                headers=self._get_headers(),
+                timeout=30
+            )
+
+            if response.status_code == 200:
+                logger.info(f"成功获取 PR #{pr_number} 的文件列表")
+                return response.json()
+            else:
+                logger.error(f"获取 PR #{pr_number} 文件列表失败: {response.status_code}")
+                return None
+
+        except Exception as e:
+            logger.error(f"获取 PR #{pr_number} 文件列表异常: {str(e)}")
+            return None
+
+    def get_file_contents(self, owner: str, repo: str, filepath: str, ref: str) -> Optional[str]:
+        """
+        获取仓库中指定文件在给定 ref（分支/commit）下的内容
+
+        Args:
+            owner: 仓库所有者
+            repo: 仓库名称
+            filepath: 文件路径
+            ref: 分支名或 commit SHA
+
+        Returns:
+            文件内容文本，失败返回 None
+        """
+        import base64
+        import urllib.parse
+        encoded_path = urllib.parse.quote(filepath, safe='')
+        url = f"{self.base_url}/api/v1/repos/{owner}/{repo}/contents/{encoded_path}?ref={urllib.parse.quote(ref)}"
+        try:
+            response = requests.get(
+                url,
+                headers=self._get_headers(),
+                timeout=30
+            )
+            if response.status_code == 200:
+                data = response.json()
+                if data.get('encoding') == 'base64' and data.get('content'):
+                    return base64.b64decode(data['content']).decode('utf-8', errors='replace')
+                return None
+            logger.error(f"获取文件 {filepath} 失败: {response.status_code}")
+            return None
+        except Exception as e:
+            logger.error(f"获取文件内容异常: {str(e)}")
+            return None

install.sh

@@ -0,0 +1,77 @@
+#!/bin/bash
+# AI Code Quality Scanner 安装脚本
+
+echo "========================================="
+echo "  AI Code Quality Scanner 安装脚本"
+echo "========================================="
+
+# 检查 Python 版本
+if ! command -v python3 &> /dev/null; then
+    echo "❌ 错误: 未找到 Python3，请先安装 Python 3.8+"
+    exit 1
+fi
+
+PYTHON_VERSION=$(python3 -c 'import sys; print(".".join(map(str, sys.version_info[:2])))')
+echo "✅ Python 版本: $PYTHON_VERSION"
+
+# 创建虚拟环境（可选）
+if [ ! -d "venv" ]; then
+    echo "📦 创建虚拟环境..."
+    python3 -m venv venv
+fi
+
+# 激活虚拟环境
+source venv/bin/activate
+
+# 安装依赖
+echo "📦 安装 Python 依赖..."
+pip install --upgrade pip
+pip install -r requirements.txt
+
+# 创建必要的目录
+echo "📁 创建必要的目录..."
+mkdir -p reports
+mkdir -p /tmp/code_scanner_clones
+
+# 检查并安装代码扫描工具（可选）
+echo "🛠️ 检查代码扫描工具..."
+
+# Pylint (Python)
+if command -v pylint &> /dev/null || python -m pylint --version &> /dev/null; then
+    echo "  ✅ Pylint 已安装"
+else
+    echo "  ⚠️ Pylint 未安装 (pip install pylint)"
+fi
+
+# Flake8 (Python)
+if command -v flake8 &> /dev/null || python -m flake8 --version &> /dev/null; then
+    echo "  ✅ Flake8 已安装"
+else
+    echo "  ⚠️ Flake8 未安装 (pip install flake8)"
+fi
+
+# Bandit (Python 安全扫描)
+if command -v bandit &> /dev/null || python -m bandit --version &> /dev/null; then
+    echo "  ✅ Bandit 已安装"
+else
+    echo "  ⚠️ Bandit 未安装 (pip install bandit)"
+fi
+
+# Node.js 和 npm (JavaScript 扫描)
+if command -v node &> /dev/null; then
+    NODE_VERSION=$(node --version)
+    echo "  ✅ Node.js 版本: $NODE_VERSION"
+else
+    echo "  ⚠️ Node.js 未安装 (JavaScript 扫描需要)"
+fi
+
+echo ""
+echo "========================================="
+echo "  安装完成！"
+echo "========================================="
+echo ""
+echo "下一步操作:"
+echo "1. 编辑 config.yaml 配置飞书机器人和 Gitea"
+echo "2. 运行: python app.py"
+echo "3. 在 Gitea 中配置 Webhook"
+echo ""

scanner/diff_parser.py

@@ -0,0 +1,172 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Diff 解析器 - 将扫描问题与代码片段关联
+"""
+import re
+import logging
+from typing import Dict, List, Any, Optional
+from dataclasses import dataclass, field
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class CodeChunk:
+    """代码块"""
+    file_path: str
+    old_content: str = ""
+    new_content: str = ""
+    old_start: int = 0
+    new_start: int = 0
+    hunks: List[Dict] = field(default_factory=list)
+
+
+class DiffParser:
+    """Diff 解析器"""
+
+    def __init__(self, diff_text: str):
+        self.diff_text = diff_text
+        self.files: Dict[str, CodeChunk] = {}
+        self._parse()
+
+    def _parse(self):
+        """解析 diff 文本"""
+        if not self.diff_text:
+            return
+
+        current_chunk = None
+        lines = self.diff_text.split('\n')
+        for line in lines:
+            diff_match = re.match(r'diff --git a/(.+) b/(.+)', line)
+            if diff_match:
+                file_path = diff_match.group(1)
+                current_chunk = CodeChunk(file_path=file_path)
+                self.files[file_path] = current_chunk
+                continue
+
+            hunk_match = re.match(r'@@ -(\d+),?\d* \+(\d+),?\d* @@', line)
+            if hunk_match and current_chunk:
+                current_chunk.old_start = int(hunk_match.group(1))
+                current_chunk.new_start = int(hunk_match.group(2))
+                continue
+
+            if current_chunk and line:
+                if line.startswith('+') and not line.startswith('+++'):
+                    current_chunk.new_content += line[1:] + '\n'
+                elif line.startswith('-') and not line.startswith('---'):
+                    current_chunk.old_content += line[1:] + '\n'
+                elif line.startswith(' '):
+                    current_chunk.old_content += line[1:] + '\n'
+                    current_chunk.new_content += line[1:] + '\n'
+
+    def get_file_content(self, file_path: str) -> Optional[CodeChunk]:
+        return self.files.get(file_path)
+
+    def get_line_context(self, file_path: str, line_number: int, context_lines: int = 3) -> Optional[Dict[str, Any]]:
+        chunk = self.files.get(file_path)
+        if not chunk:
+            return None
+
+        new_lines = chunk.new_content.split('\n')
+        if line_number > len(new_lines):
+            return None
+
+        start = max(0, line_number - context_lines - 1)
+        end = min(len(new_lines), line_number + context_lines)
+
+        context = []
+        for i in range(start, end):
+            code = new_lines[i].rstrip('\n')
+            is_current_line = (i == line_number - 1)
+            context.append({
+                'line_number': chunk.new_start + i,
+                'code': code,
+                'is_issue_line': is_current_line
+            })
+
+        return {
+            'file': file_path,
+            'line': line_number,
+            'context': context
+        }
+
+
+def merge_issues_with_code(scan_results: Dict[str, Any], diff: str) -> Dict[str, Any]:
+    """将扫描问题与代码片段关联"""
+    if not diff:
+        return scan_results
+
+    parser = DiffParser(diff)
+    enriched_results = {
+        'scanners': [],
+        'summary': scan_results.get('summary', {}),
+        'total_issues': scan_results.get('total_issues', 0)
+    }
+
+    for scanner_name, scanner_data in scan_results.items():
+        if scanner_name in ['summary', 'total_issues', 'ai']:
+            continue
+
+        if isinstance(scanner_data, dict):
+            enriched_scanner = {
+                'name': scanner_name,
+                'issues': [],
+                'file_count': scanner_data.get('file_count', 0),
+                'total_issues': scanner_data.get('total_issues', 0)
+            }
+
+            issues = scanner_data.get('issues', [])
+            for issue in issues:
+                enriched_issue = enrich_issue_with_code(issue, parser)
+                enriched_scanner['issues'].append(enriched_issue)
+
+            enriched_results['scanners'].append(enriched_scanner)
+
+    if 'ai' in scan_results:
+        enriched_results['ai'] = scan_results['ai']
+
+    return enriched_results
+
+
+def enrich_issue_with_code(issue: Dict[str, Any], parser: DiffParser) -> Dict[str, Any]:
+    """为单个问题添加代码片段"""
+    enriched = issue.copy()
+
+    file_path = issue.get('file', '')
+    line_number = issue.get('line', 0)
+
+    if not file_path:
+        return enriched
+
+    if not line_number:
+        desc = issue.get('description', '') or issue.get('message', '')
+        line_match = re.search(r'line[:#]?\s*(\d+)', desc, re.IGNORECASE)
+        if line_match:
+            line_number = int(line_match.group(1))
+
+    matched_path = None
+    for path in parser.files.keys():
+        if file_path.endswith(path) or path.endswith(file_path) or file_path in path:
+            matched_path = path
+            break
+
+    if matched_path:
+        enriched['file'] = matched_path
+    if matched_path and line_number:
+        context = parser.get_line_context(matched_path, line_number)
+        if context:
+            enriched['code_context'] = context
+
+    if 'code_context' not in enriched and matched_path:
+        chunk = parser.get_file_content(matched_path)
+        if chunk and chunk.new_content:
+            lines = chunk.new_content.split('\n')[:10]
+            enriched['code_context'] = {
+                'file': matched_path,
+                'line': line_number or 1,
+                'preview': '\n'.join(lines),
+                'has_more': len(chunk.new_content.split('\n')) > 10
+            }
+
+    return enriched

web/index.html

@@ -8,6 +8,11 @@
     <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/bootstrap-icons@1.11.0/font/bootstrap-icons.css">
     <style>
         body { background-color: #f5f7fa; }
+        /* Diff 语法高亮 */
+        #detail-code-diff .diff-add { color: #98c379; background: rgba(72, 120, 80, 0.2); }
+        #detail-code-diff .diff-del { color: #e06c75; background: rgba(180, 80, 80, 0.2); }
+        #detail-code-diff .diff-info { color: #61afef; }
+        #detail-code-diff .diff-header { color: #e5c07b; }
         .sidebar { 
             min-height: 100vh; 
             background: linear-gradient(180deg, #1a1a2e 0%, #16213e 100%);
@@ -43,6 +48,38 @@
         .issue-high { color: #d32f2f; }
         .issue-medium { color: #f57c00; }
         .issue-low { color: #388e3c; }
+        /* PR 详情 - Git 风格：左侧文件树 + 右侧文件内容 */
+        .pr-detail-file-layout { display: flex; height: 500px; border: 1px solid #dee2e6; border-radius: 6px; overflow: hidden; }
+        .pr-detail-file-tree { width: 280px; min-width: 280px; overflow-y: auto; background: #f8f9fa; border-right: 1px solid #dee2e6; }
+        .pr-detail-file-tree .tree-root { padding: 8px 0; }
+        .pr-detail-file-tree .tree-node { padding: 2px 8px; font-size: 13px; cursor: pointer; border-radius: 4px; }
+        .pr-detail-file-tree .tree-node:hover { background: #e9ecef; }
+        .pr-detail-file-tree .tree-node.active { background: #0d6efd; color: #fff; }
+        .pr-detail-file-tree .tree-folder { font-weight: 600; color: #495057; }
+        .pr-detail-file-tree .tree-file { padding-left: 12px; }
+        /* 代码在中间，右侧缺陷标注，虚线连接代码行与标注（参考审阅风格） */
+        .pr-detail-file-right { flex: 1; display: flex; overflow: hidden; min-width: 0; }
+        .pr-detail-file-wrapper { width: 100%; height: 100%; min-width: 0; overflow: hidden; }
+        .pr-detail-code-view { width: 100%; height: 100%; overflow: auto; background: #1e1e1e; font-family: 'Consolas', 'Monaco', monospace; font-size: 13px; line-height: 1.5; }
+        .pr-detail-code-view .code-line { display: flex; min-height: 1.5em; align-items: stretch; }
+        .pr-detail-code-view .code-line .line-gutter { width: 48px; min-width: 48px; flex-shrink: 0; display: flex; align-items: center; justify-content: flex-end; padding-right: 8px; color: #6c757d; background: #252526; user-select: none; }
+        .pr-detail-code-view .code-line .line-gutter .line-num { margin-right: 4px; }
+        .pr-detail-code-view .code-line .line-gutter .gutter-icon { width: 14px; height: 14px; flex-shrink: 0; }
+        .pr-detail-code-view .code-line .line-gutter .gutter-icon.icon-error { color: #f14c4c; }
+        .pr-detail-code-view .code-line .line-gutter .gutter-icon.icon-warning { color: #cca700; }
+        .pr-detail-code-view .code-line .line-content { flex: 1; min-width: 0; color: #d4d4d4; padding: 0 12px; white-space: pre-wrap; word-break: break-all; overflow-wrap: break-word; }
+        .pr-detail-code-view .code-line.line-has-issue .line-content { background: rgba(220, 53, 69, 0.1); border-left: 2px solid #dc3545; padding-left: 10px; }
+        /* 虚线连接区：代码与标注之间 */
+        .pr-detail-code-view .code-line .line-connector { width: 20px; min-width: 20px; flex-shrink: 0; position: relative; background: #1e1e1e; }
+        .pr-detail-code-view .code-line.line-has-issue .line-connector::before { content: ''; position: absolute; left: 0; right: 0; top: 50%; margin-top: -1px; height: 0; border-bottom: 1px dashed #f14c4c; }
+        .pr-detail-code-view .code-line .line-annotation { width: 180px; min-width: 180px; flex-shrink: 0; padding: 6px 8px; font-size: 11px; background: #252526; color: #9d9d9d; border-left: 1px solid #3c3c3c; white-space: normal; word-break: break-word; display: flex; align-items: center; font-family: inherit; }
+        .pr-detail-code-view .code-line.line-has-issue .line-annotation { background: #fff8f8; border-left: 2px solid #dc3545; color: #c62828; }
+        .pr-detail-code-view .code-line .line-annotation .anno-text { line-height: 1.4; }
+        .pr-detail-code-view .code-line .line-annotation .anno-icon { color: #dc3545; margin-right: 6px; flex-shrink: 0; }
+        .pr-detail-file-placeholder { color: #6c757d; padding: 24px; text-align: center; width: 100%; }
+        /* PR 详情弹窗加宽：占屏幕大部分宽度，代码区随之扩大 */
+        .modal-dialog-pr-wide { max-width: 92%; width: 92%; margin-left: auto; margin-right: auto; }
+        @media (min-width: 1400px) { .modal-dialog-pr-wide { max-width: 1400px; width: 92%; } }
     </style>
 </head>
 <body>
@@ -208,9 +245,9 @@
         </div>
     </div>
 
-    <!-- PR 详情模态框 -->
+    <!-- PR 详情模态框：加宽整体以扩大代码区域 -->
     <div class="modal fade" id="prDetailModal" tabindex="-1">
-        <div class="modal-dialog modal-xl">
+        <div class="modal-dialog modal-dialog-pr-wide">
             <div class="modal-content">
                 <div class="modal-header">
                     <h5 class="modal-title" id="detail-pr-title">PR 详情</h5>
@@ -229,16 +266,20 @@
                             <p><strong>安全漏洞:</strong> <span id="detail-security"></span></p>
                         </div>
                     </div>
-                    <ul class="nav nav-tabs" id="detail-tabs" role="tablist">
-                        <li class="nav-item"><button class="nav-link active" data-bs-toggle="tab" data-bs-target="#tab-scan-result">扫描结果</button></li>
-                        <li class="nav-item"><button class="nav-link" data-bs-toggle="tab" data-bs-target="#tab-ai-review">AI 审查</button></li>
-                    </ul>
-                    <div class="tab-content mt-3">
-                        <div class="tab-pane fade show active" id="tab-scan-result">
-                            <pre id="detail-scan-result" style="max-height: 400px; overflow: auto; background: #f8f9fa; padding: 15px; border-radius: 5px;"></pre>
-                        </div>
-                        <div class="tab-pane fade" id="tab-ai-review">
-                            <div id="detail-ai-review" style="white-space: pre-wrap; background: #f8f9fa; padding: 15px; border-radius: 5px;"></div>
+                    <!-- 仅保留文件 Tab：左侧文件树 + 右侧完整文件内容，最右侧为问题标注 -->
+                    <div class="mt-3">
+                        <div class="pr-detail-file-layout">
+                            <div class="pr-detail-file-tree">
+                                <div id="pr-file-tree-loading" class="p-3 text-center text-muted">加载文件列表中...</div>
+                                <div id="pr-file-tree" class="tree-root" style="display: none;"></div>
+                            </div>
+                            <div class="pr-detail-file-right">
+                                <div id="pr-file-content-placeholder" class="pr-detail-file-placeholder">点击左侧文件查看完整内容</div>
+                                <div id="pr-file-content-wrapper" class="pr-detail-file-wrapper">
+                                    <div class="pr-detail-code-view" id="pr-file-content"></div>
+                                </div>
+                                <div id="pr-file-content-error" class="pr-detail-file-placeholder" style="display: none; color: #dc3545;"></div>
+                            </div>
                         </div>
                     </div>
                 </div>
@@ -259,6 +300,44 @@
     <script>
         let currentPRId = null;
         let currentPRInfo = null;
+        let currentDiffData = null;
+        let currentAIReview = null;
+
+        // 全局：Markdown 简易格式化
+        function formatMarkdown(text) {
+            if (!text) return '';
+            return String(text)
+                .replace(/^### (.+)$/gm, '<h5 class="mt-3 mb-2">$1</h5>')
+                .replace(/^## (.+)$/gm, '<h4 class="mt-3 mb-2">$1</h4>')
+                .replace(/^# (.+)$/gm, '<h3 class="mt-3 mb-2">$1</h3>')
+                .replace(/^\*\*(.+):\*\*$/gm, '<strong>$1:</strong>')
+                .replace(/^- /gm, '<br>• ')
+                .replace(/\n\n/g, '</p><p>')
+                .replace(/\n/g, '<br>');
+        }
+
+        // 将 AI 审查对象格式化为 HTML（优点/问题/优化 或 raw_review）
+        function formatReviewObject(review) {
+            if (!review || typeof review !== 'object') return '';
+            if (review.raw_review) return formatMarkdown(review.raw_review);
+            let html = '';
+            if (review['优点'] && review['优点'].length) {
+                html += '<div class="mb-2"><strong class="text-success">优点</strong><ul class="mb-0">';
+                review['优点'].forEach(s => { html += '<li>' + escapeHtml(s) + '</li>'; });
+                html += '</ul></div>';
+            }
+            if (review['问题'] && review['问题'].length) {
+                html += '<div class="mb-2"><strong class="text-danger">问题</strong><ul class="mb-0">';
+                review['问题'].forEach(s => { html += '<li>' + escapeHtml(s) + '</li>'; });
+                html += '</ul></div>';
+            }
+            if (review['优化'] && review['优化'].length) {
+                html += '<div class="mb-2"><strong class="text-primary">优化</strong><ul class="mb-0">';
+                review['优化'].forEach(s => { html += '<li>' + escapeHtml(s) + '</li>'; });
+                html += '</ul></div>';
+            }
+            return html || formatMarkdown(JSON.stringify(review, null, 2));
+        }
 
         // 页面切换
         function showPage(page) {
@@ -355,36 +434,513 @@
                 document.getElementById('detail-security').innerHTML = (pr.security_issues || 0) > 0 ?
                     '<span class="text-danger">' + pr.security_issues + '</span>' : '0';
 
-                // 扫描结果
-                let scanResult = pr.scan_result;
-                if (typeof scanResult === 'string') {
-                    try { scanResult = JSON.parse(scanResult); } catch(e) {}
-                }
-                document.getElementById('detail-scan-result').textContent = JSON.stringify(scanResult, null, 2);
-
-                // AI 审查
-                let aiReview = pr.ai_review;
-                if (typeof aiReview === 'string') {
-                    try { aiReview = JSON.parse(aiReview); } catch(e) {}
-                }
-                if (aiReview && aiReview.review) {
-                    document.getElementById('detail-ai-review').textContent = aiReview.review;
-                } else {
-                    document.getElementById('detail-ai-review').textContent = '无 AI 审查结果';
-                }
-
                 // 根据状态显示/隐藏按钮
                 const canOperate = pr.state === 'open';
                 document.getElementById('btn-merge').style.display = canOperate ? 'inline-block' : 'none';
                 document.getElementById('btn-reject').style.display = canOperate ? 'inline-block' : 'none';
 
                 // 显示模态框
-                new bootstrap.Modal(document.getElementById('prDetailModal')).show();
+                const modal = new bootstrap.Modal(document.getElementById('prDetailModal'));
+                modal.show();
+
+                // 加载文件树（左侧树，点击文件在右侧显示完整内容+标注）
+                loadPRFileTree(id);
             } catch (e) {
                 alert('加载 PR 详情失败: ' + e.message);
             }
         }
 
+        // 加载 PR 文件列表并渲染左侧树，点击文件在右侧显示完整内容
+        async function loadPRFileTree(prId) {
+            const loadingEl = document.getElementById('pr-file-tree-loading');
+            const treeEl = document.getElementById('pr-file-tree');
+            const placeholderEl = document.getElementById('pr-file-content-placeholder');
+            const wrapperEl = document.getElementById('pr-file-content-wrapper');
+            const contentEl = document.getElementById('pr-file-content');
+            const errorEl = document.getElementById('pr-file-content-error');
+
+            if (loadingEl) loadingEl.style.display = 'block';
+            if (treeEl) treeEl.style.display = 'none';
+            if (placeholderEl) placeholderEl.style.display = 'block';
+            if (wrapperEl) wrapperEl.style.display = 'none';
+            if (contentEl) contentEl.innerHTML = '';
+            if (errorEl) errorEl.style.display = 'none';
+
+            try {
+                const res = await fetch('/api/prs/' + prId + '/files');
+                const data = await res.json();
+                if (data.error || !data.files || !data.files.length) {
+                    if (treeEl) {
+                        treeEl.style.display = 'block';
+                        treeEl.innerHTML = '<div class="p-3 text-muted small">' + (data.error || '暂无变更文件') + '</div>';
+                    }
+                    if (loadingEl) loadingEl.style.display = 'none';
+                    return;
+                }
+                const files = data.files;
+                const tree = buildFileTree(files);
+                if (treeEl) {
+                    treeEl.innerHTML = renderFileTreeNodes(tree, prId);
+                    treeEl.style.display = 'block';
+                }
+                if (loadingEl) loadingEl.style.display = 'none';
+            } catch (e) {
+                if (treeEl) {
+                    treeEl.style.display = 'block';
+                    treeEl.innerHTML = '<div class="p-3 text-danger small">加载失败: ' + escapeHtml(e.message) + '</div>';
+                }
+                if (loadingEl) loadingEl.style.display = 'none';
+            }
+        }
+
+        // 将扁平文件列表转为树结构 { name, children?, path? }
+        function buildFileTree(files) {
+            const root = { name: '', children: {} };
+            for (const f of files) {
+                const path = f.filename || f.path || f;
+                const parts = typeof path === 'string' ? path.split('/') : [String(path)];
+                let cur = root;
+                for (let i = 0; i < parts.length; i++) {
+                    const isFile = i === parts.length - 1;
+                    const name = parts[i];
+                    if (isFile) {
+                        if (!cur.children[name]) cur.children[name] = { name, path, file: true, status: f.status };
+                    } else {
+                        if (!cur.children[name]) cur.children[name] = { name, children: {} };
+                        cur = cur.children[name];
+                    }
+                }
+            }
+            return root;
+        }
+
+        // 递归渲染树节点 HTML
+        function renderFileTreeNodes(node, prId) {
+            const entries = Object.keys(node.children || {}).sort((a, b) => {
+                const aa = node.children[a];
+                const ab = node.children[b];
+                const aIsDir = !aa.file;
+                const bIsDir = !ab.file;
+                if (aIsDir !== bIsDir) return aIsDir ? -1 : 1;
+                return a.localeCompare(b);
+            });
+            if (entries.length === 0) return '';
+            let html = '';
+            for (const key of entries) {
+                const child = node.children[key];
+                if (child.file) {
+                    const status = child.status || '';
+                    const statusCls = status === 'added' ? 'text-success' : status === 'removed' ? 'text-danger' : 'text-secondary';
+                    html += '<div class="tree-node tree-file ' + statusCls + '" data-path="' + escapeHtml(child.path) + '" data-pr-id="' + prId + '"><i class="bi bi-file-code me-1"></i>' + escapeHtml(key) + '</div>';
+                } else {
+                    html += '<div class="tree-folder tree-node">' + escapeHtml(key) + '</div>';
+                    html += '<div style="padding-left: 12px;">' + renderFileTreeNodes(child, prId) + '</div>';
+                }
+            }
+            return html;
+        }
+
+        // 点击文件时加载并显示内容
+        document.addEventListener('click', function(ev) {
+            const node = ev.target.closest('.tree-file');
+            if (!node) return;
+            ev.preventDefault();
+            const path = node.getAttribute('data-path');
+            const prId = node.getAttribute('data-pr-id');
+            if (!path || !prId) return;
+            document.querySelectorAll('.pr-detail-file-tree .tree-node.active').forEach(el => el.classList.remove('active'));
+            node.classList.add('active');
+            loadPRFileContent(prId, path);
+        });
+
+        async function loadPRFileContent(prId, path) {
+            const placeholderEl = document.getElementById('pr-file-content-placeholder');
+            const wrapperEl = document.getElementById('pr-file-content-wrapper');
+            const contentEl = document.getElementById('pr-file-content');
+            const errorEl = document.getElementById('pr-file-content-error');
+
+            if (placeholderEl) placeholderEl.style.display = 'none';
+            if (errorEl) errorEl.style.display = 'none';
+
+            try {
+                const res = await fetch('/api/prs/' + prId + '/file?path=' + encodeURIComponent(path));
+                const data = await res.json();
+
+                if (data.error) {
+                    errorEl.textContent = data.error;
+                    errorEl.style.display = 'block';
+                    return;
+                }
+
+                // 行号 -> 问题列表
+                const lineIssues = {};
+                const issues = data.scan_issues || [];
+                for (const issue of issues) {
+                    const line = issue.line || 1;
+                    if (!lineIssues[line]) lineIssues[line] = [];
+                    lineIssues[line].push(issue);
+                }
+
+                // 代码在中间，右侧缺陷标注，虚线连接代码行与标注
+                const lines = (data.content || '').split('\n');
+                let html = '';
+                for (let i = 0; i < lines.length; i++) {
+                    const lineNum = i + 1;
+                    const rowIssues = lineIssues[lineNum] || [];
+                    const hasIssue = rowIssues.length > 0;
+                    const rowClass = hasIssue ? 'code-line line-has-issue' : 'code-line';
+
+                    const lineText = lines[i];
+                    const displayText = lineText === '' ? '\u00A0' : lineText;
+
+                    let gutterHtml = '<span class="line-num">' + lineNum + '</span>';
+                    if (hasIssue) {
+                        const first = rowIssues[0];
+                        const sev = first.severity || 'info';
+                        const iconClass = (sev === 'error' || sev === 'high') ? 'gutter-icon icon-error' : 'gutter-icon icon-warning';
+                        gutterHtml += '<span class="' + iconClass + '"><i class="bi bi-exclamation-circle-fill" style="font-size:12px;"></i></span>';
+                    }
+
+                    const reasonText = hasIssue ? rowIssues.map(function(iss) {
+                        return (iss.scanner ? '[' + iss.scanner + '] ' : '') + (iss.message || '');
+                    }).join('；') : '';
+
+                    html += '<div class="' + rowClass + '">';
+                    html += '<div class="line-gutter">' + gutterHtml + '</div>';
+                    html += '<span class="line-content">' + escapeHtml(displayText) + '</span>';
+                    html += '<div class="line-connector"></div>';
+                    html += '<div class="line-annotation">';
+                    if (hasIssue && reasonText) {
+                        html += '<span class="anno-icon"><i class="bi bi-exclamation-triangle-fill"></i></span>';
+                        html += '<span class="anno-text">' + escapeHtml(reasonText) + '</span>';
+                    }
+                    html += '</div>';
+                    html += '</div>';
+                }
+                contentEl.innerHTML = html;
+
+                if (wrapperEl) wrapperEl.style.display = 'block';
+            } catch (e) {
+                errorEl.textContent = '加载失败: ' + e.message;
+                errorEl.style.display = 'block';
+            }
+        }
+
+        // 渲染带代码片段的扫描结果
+        function renderScanDetailsWithCode(scanDetails) {
+            let html = '';
+
+            // 显示汇总信息
+            if (scanDetails.summary) {
+                html += '<div class="alert alert-info mb-3">';
+                html += `<strong>总问题数:</strong> ${scanDetails.total_issues} | `;
+                html += `<strong>扫描器:</strong> ${scanDetails.scanners ? scanDetails.scanners.length : 0}`;
+                html += '</div>';
+            }
+
+            // 遍历每个扫描器
+            if (scanDetails.scanners) {
+                for (const scanner of scanDetails.scanners) {
+                    html += `<div class="card mb-3">
+                        <div class="card-header bg-light">
+                            <strong>${scanner.name}</strong>
+                            <span class="badge bg-${scanner.total_issues > 0 ? 'danger' : 'success'}">
+                                ${scanner.total_issues} 个问题
+                            </span>
+                        </div>
+                        <div class="card-body">`;
+
+                    if (scanner.issues && scanner.issues.length > 0) {
+                        for (const issue of scanner.issues) {
+                            html += `<div class="issue-item mb-3 p-2 border rounded">`;
+                            html += `<div class="fw-bold text-${issue.severity === 'error' ? 'danger' : 'warning'}">`;
+                            html += `<i class="bi bi-exclamation-triangle"></i> ${issue.message || issue.description || '问题'}`;
+                            html += ` <small class="text-muted">(${issue.file}:${issue.line || '?'})</small>`;
+                            html += `</div>`;
+
+                            // 显示代码片段
+                            if (issue.code_context) {
+                                const ctx = issue.code_context;
+                                html += `<div class="code-context mt-2" style="background: #1e1e1e; color: #d4d4d4; padding: 10px; border-radius: 4px; font-size: 12px; overflow-x: auto;">`;
+                                html += `<div class="text-muted small mb-1">文件: ${ctx.file || issue.file}</div>`;
+
+                                if (ctx.context) {
+                                    // 显示上下文代码
+                                    for (const line of ctx.context) {
+                                        const lineClass = line.is_issue_line ? 'background: rgba(255, 200, 0, 0.2);' : '';
+                                        const prefix = line.is_issue_line ? '👉 ' : '   ';
+                                        html += `<div style="${lineClass}white-space: pre;">${prefix}${line.line_number}: ${escapeHtml(line.code)}</div>`;
+                                    }
+                                } else if (ctx.preview) {
+                                    // 显示文件预览
+                                    html += `<pre style="margin: 0;">${escapeHtml(ctx.preview)}</pre>`;
+                                    if (ctx.has_more) {
+                                        html += `<div class="text-muted small">... (更多内容)</div>`;
+                                    }
+                                }
+                                html += `</div>`;
+                            }
+
+                            html += `</div>`;
+                        }
+                    } else {
+                        html += '<div class="text-muted">没有问题</div>';
+                    }
+
+                    html += '</div></div>';
+                }
+            }
+
+            return html;
+        }
+
+        // HTML 转义
+        function escapeHtml(text) {
+            const div = document.createElement('div');
+            div.textContent = text;
+            return div.innerHTML;
+        }
+
+        // 加载 PR 代码差异
+        async function loadPRDiff(prId) {
+            try {
+                const response = await fetch('/api/prs/' + prId + '/diff');
+                const data = await response.json();
+
+                const loadingEl = document.getElementById('detail-code-diff-loading');
+                const diffEl = document.getElementById('detail-code-diff');
+
+                if (loadingEl) loadingEl.style.display = 'none';
+
+                if (data.error) {
+                    if (diffEl) {
+                        diffEl.style.display = 'block';
+                        diffEl.textContent = '加载失败: ' + data.error;
+                    }
+                    return;
+                }
+
+                // 格式化 diff 显示
+                const diff = data.diff || '';
+                if (!diff) {
+                    if (diffEl) {
+                        diffEl.style.display = 'block';
+                        diffEl.textContent = '没有代码差异';
+                    }
+                    return;
+                }
+
+                // 添加语法高亮样式
+                const formattedDiff = diff
+                    .replace(/^diff --git.*$/gm, match => '\n' + match)
+                    .replace(/^@@.*@@$/gm, match => '\n' + match + '\n')
+                    .replace(/^\+.*$/gm, match => match)
+                    .replace(/^-.*$/gm, match => match);
+
+                if (diffEl) {
+                    diffEl.style.display = 'block';
+                    diffEl.textContent = diff;
+                }
+
+            } catch (e) {
+                const loadingEl = document.getElementById('detail-code-diff-loading');
+                if (loadingEl) loadingEl.style.display = 'none';
+                if (diffEl) {
+                    diffEl.style.display = 'block';
+                    diffEl.textContent = '加载失败: ' + e.message;
+                }
+            }
+        }
+
+        // 加载 AI 审查结果：一行一行展示（每行左侧代码片段 + 右侧审查结果）
+        async function loadAIReview(prId) {
+            try {
+                const prResponse = await fetch('/api/prs/' + prId);
+                const pr = await prResponse.json();
+                if (pr.error) {
+                    showAIReviewError('加载失败: ' + pr.error);
+                    return;
+                }
+
+                let aiReview = pr.ai_review;
+                if (typeof aiReview === 'string') {
+                    try { aiReview = JSON.parse(aiReview); } catch(e) {}
+                }
+                currentAIReview = aiReview;
+
+                const diffResponse = await fetch('/api/prs/' + prId + '/diff');
+                const diffData = await diffResponse.json();
+                currentDiffData = diffData.diff || '';
+
+                const loadingEl = document.getElementById('ai-review-loading');
+                const emptyEl = document.getElementById('ai-review-empty');
+                const contentEl = document.getElementById('ai-review-content');
+                const summaryEl = document.getElementById('ai-review-summary');
+                const rowsEl = document.getElementById('ai-review-rows');
+
+                if (!loadingEl || !emptyEl || !contentEl || !rowsEl) return;
+
+                loadingEl.style.display = 'none';
+
+                // 兼容处理：优先使用 reviews 数组，若无则尝试从 summary 文本中解析文件审查
+                let reviews = [];
+                if (aiReview && Array.isArray(aiReview.reviews)) {
+                    reviews = aiReview.reviews;
+                } else if (aiReview && aiReview.summary && typeof aiReview.summary === 'string') {
+                    // 尝试从 summary 文本中解析每个文件的审查结果
+                    // 格式如: "### 📄 app.py\n\n**优点:**\n- ..."
+                    const summaryText = aiReview.summary;
+                    // 按 "### 📄 " 分割，提取每个文件的审查
+                    const fileBlocks = summaryText.split(/(?=### 📄 )/);
+                    for (const block of fileBlocks) {
+                        const match = block.match(/### 📄 ([^\n]+)\n([\s\S]*)/);
+                        if (match) {
+                            const fileName = match[1].trim();
+                            const reviewContent = match[2].trim();
+                            // 将文本格式转为对象结构
+                            const reviewObj = parseReviewTextToObject(reviewContent);
+                            reviews.push({ file: fileName, review: reviewObj });
+                        }
+                    }
+                    // 如果解析不出文件，则将整个 summary 作为整体审查
+                    if (reviews.length === 0) {
+                        reviews = [{ file: '整体审查结果', review: { 'raw_review': summaryText } }];
+                    }
+                }
+
+                // 辅助函数：将审查文本转为对象结构
+                function parseReviewTextToObject(text) {
+                    const obj = {};
+                    // 匹配 "**优点:**" / "**需要改进:**" / "**优化建议:**" 等
+                    const sections = { '优点': [], '问题': [], '优化': [] };
+                    let currentSection = null; // '优点' | '问题' | '优化'
+                    const lines = text.split('\n');
+                    for (const line of lines) {
+                        let found = false;
+                        if (/优点/.test(line) && /\*\*/.test(line)) {
+                            currentSection = '优点';
+                            found = true;
+                        } else if (/(需要改进|问题)/.test(line) && /\*\*/.test(line)) {
+                            currentSection = '问题';
+                            found = true;
+                        } else if (/(优化建议|优化)/.test(line) && /\*\*/.test(line)) {
+                            currentSection = '优化';
+                            found = true;
+                        }
+                        if (!found && currentSection) {
+                            const m = line.match(/^- (.+)/);
+                            if (m) {
+                                sections[currentSection].push(m[1]);
+                            }
+                        }
+                    }
+                    // 清理空数组
+                    if (sections['优点'].length) obj['优点'] = sections['优点'];
+                    if (sections['问题'].length) obj['问题'] = sections['问题'];
+                    if (sections['优化'].length) obj['优化'] = sections['优化'];
+                    // 如果解析不出结构，直接返回原始文本
+                    return Object.keys(obj).length ? obj : { 'raw_review': text };
+                }
+
+                if (reviews.length === 0) {
+                    emptyEl.style.display = 'block';
+                    contentEl.style.display = 'none';
+                    return;
+                }
+
+                emptyEl.style.display = 'none';
+                contentEl.style.display = 'block';
+
+                if (summaryEl) {
+                    summaryEl.textContent = aiReview.summary || ('已审查 ' + reviews.length + ' 个文件');
+                }
+
+                let rowsHtml = '';
+                const diff = currentDiffData || '';
+
+                for (const item of reviews) {
+                    const fileDisplay = item.file || (item.path ? item.path.replace(/^.*[/\\]/, '') : '');
+                    const fileForDiff = item.file || fileDisplay;
+                    const codeSnippet = diff ? extractFileFromDiff(diff, fileForDiff) : '';
+                    const codeDisplay = codeSnippet || '（该文件无 diff 或未在 PR 中修改）';
+                    const reviewHtml = formatReviewObject(item.review);
+
+                    rowsHtml += `
+                        <div class="card mb-3 border">
+                            <div class="card-header py-2 bg-light d-flex justify-content-between align-items-center">
+                                <span><i class="bi bi-file-code me-1"></i> ${escapeHtml(fileDisplay)}</span>
+                            </div>
+                            <div class="card-body p-0">
+                                <div class="row g-0">
+                                    <div class="col-md-6 border-end">
+                                        <div class="p-2 bg-dark text-light">
+                                            <pre class="mb-0 p-2" style="background:#1e1e1e;color:#d4d4d4;font-size:12px;max-height:280px;overflow:auto;white-space:pre;">${escapeHtml(codeDisplay)}</pre>
+                                        </div>
+                                    </div>
+                                    <div class="col-md-6">
+                                        <div class="p-3" style="min-height:120px;">${reviewHtml || '—'}</div>
+                                    </div>
+                                </div>
+                            </div>
+                        </div>`;
+                }
+
+                rowsEl.innerHTML = rowsHtml;
+            } catch (e) {
+                showAIReviewError('加载失败: ' + e.message);
+            }
+        }
+
+        // 显示 AI 审查错误信息
+        function showAIReviewError(message) {
+            const loadingEl = document.getElementById('ai-review-loading');
+            const emptyEl = document.getElementById('ai-review-empty');
+            const contentEl = document.getElementById('ai-review-content');
+
+            if (loadingEl) loadingEl.style.display = 'none';
+            if (contentEl) contentEl.style.display = 'none';
+            if (emptyEl) {
+                emptyEl.style.display = 'block';
+                emptyEl.textContent = message;
+            }
+        }
+
+        // 从 diff 中提取指定文件的代码
+        function extractFileFromDiff(diff, targetFile) {
+            const lines = diff.split('\n');
+            let inTargetFile = false;
+            let result = [];
+            let currentFile = '';
+
+            for (const line of lines) {
+                // 检测新文件开始
+                const diffMatch = line.match(/diff --git a\/(.+) b\/(.+)/);
+                if (diffMatch) {
+                    currentFile = diffMatch[1];
+                    inTargetFile = currentFile.includes(targetFile) || targetFile.includes(currentFile);
+                    continue;
+                }
+
+                // 如果在目标文件中，收集代码行
+                if (inTargetFile) {
+                    // 遇到新文件，停止
+                    if (line.startsWith('diff --git')) break;
+                    // 跳过 diff 和 hunk 头部
+                    if (line.startsWith('@@')) {
+                        result.push(line);
+                        continue;
+                    }
+                    // 添加实际代码（去掉 +, -,   前缀，但保留内容）
+                    if (line.startsWith('+') || line.startsWith('-') || line.startsWith(' ')) {
+                        result.push(line);
+                    }
+                }
+            }
+
+            return result.join('\n');
+        }
+
+        // 监听 AI 审查标签点击
         // 合并 PR
         async function mergePR() {
             if (!currentPRInfo) return;
@@ -443,6 +999,10 @@
         }
 
         // 初始化
+        window.onerror = function(msg, url, line, col, error) {
+            console.error('Global error:', msg, 'at line', line);
+            return false;
+        };
         loadDashboard();
     </script>
 </body>

快速开始指南.md

@@ -0,0 +1,230 @@
+# 快速开始指南
+
+本文档将帮助你快速部署 AI Code Quality Scanner 并配置 Gitea Webhook 和飞书通知。
+
+## 环境要求
+
+- Python 3.8+
+- Git
+- Node.js 和 npm（用于 JavaScript/TypeScript 扫描，可选）
+- Docker 和 Docker Compose（可选，用于容器化部署）
+
+## 步骤 1：配置修改
+
+### 修改 `config.yaml`
+
+首先编辑 `config.yaml` 文件，配置以下内容：
+
+```yaml
+server:
+  host: "0.0.0.0"
+  port: 5000  # Webhook 服务端口
+
+gitea:
+  base_url: "http://服务器IP:3000"  # 你的 Gitea 地址
+  webhook_secret: "your_secret_key"  # Webhook 签名密钥
+
+feishu:
+  webhook_url: "https://open.feishu.cn/open-apis/bot/v2/hook/xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"  # 飞书 Webhook 地址
+  secret: ""  # 飞书签名密钥（可选）
+```
+
+### 获取飞书 Webhook 地址
+
+1. 打开飞书群聊
+2. 点击右上角「...」→「设置」→「群机器人」
+3. 点击「添加机器人」→「自定义机器人」
+4. 设置机器人名称，点击「添加」
+5. 复制 Webhook 地址
+6. （可选）开启「签名校验」，复制 secret
+
+### 获取 Gitea Webhook 密钥
+
+1. 在 Gitea 仓库页面，点击「仓库设置」→「Webhooks」
+2. 点击「添加 Webhook」→「Gitea」
+3. 填写以下信息：
+   - 目标 URL: `http://你的服务器IP:5000/webhook/gitea`
+   - 密钥: 自定义一个密钥（如 `my_secret_key`），需要与 config.yaml 中的 `webhook_secret` 一致
+4. 点击「添加 Webhook」
+
+## 步骤 2：安装依赖
+
+### 方式 A：本地安装（Windows/Mac/Linux）
+
+```bash
+# Windows
+install.bat
+
+# Mac/Linux
+chmod +x install.sh
+./install.sh
+```
+
+### 方式 B：Docker 部署
+
+```bash
+# 构建并运行
+docker-compose up -d
+
+# 查看日志
+docker-compose logs -f
+```
+
+## 步骤 3：启动服务
+
+```bash
+# 激活虚拟环境（如果使用虚拟环境）
+# Windows
+call venv\Scripts\activate.bat
+
+# Mac/Linux
+source venv/bin/activate
+
+# 启动服务
+python app.py
+```
+
+服务启动后，访问 `http://localhost:5000` 可以看到健康检查响应。
+
+## 步骤 4：测试
+
+### 测试 Webhook
+
+在 Gitea 仓库中进行一次代码提交，应该能看到：
+
+1. 服务端日志显示收到 Webhook 请求
+2. 代码被克隆到临时目录
+3. 扫描工具运行
+4. 飞书群聊收到通知
+
+### 测试手动扫描
+
+```bash
+curl -X POST http://localhost:5000/scan/manual \
+  -H "Content-Type: application/json" \
+  -d '{"repo_url": "https://github.com/username/repo.git", "branch": "main"}'
+```
+
+## 配置说明
+
+### 扫描工具说明
+
+| 工具 | 语言 | 功能 |
+|------|------|------|
+| Pylint | Python | 代码风格和错误检查 |
+| Flake8 | Python | Python 代码检查 |
+| Bandit | Python | 安全漏洞扫描 |
+| ESLint | JavaScript/TypeScript | JS/TS 代码检查 |
+
+### 配置文件选项
+
+```yaml
+server:
+  host: "0.0.0.0"      # 监听地址
+  port: 5000           # 监听端口
+  debug: true          # 调试模式
+
+gitea:
+  base_url: "http://localhost:3000"   # Gitea 地址
+  webhook_secret: "secret"             # Webhook 签名密钥
+
+feishu:
+  webhook_url: "https://..."           # 飞书 Webhook
+  secret: ""                           # 飞书签名密钥
+
+scanner:
+  languages:
+    - python
+    - javascript
+    - typescript
+  max_issues: 10          # 最大问题数量
+  detailed: true          # 详细扫描模式
+  temp_clone_dir: "/tmp/code_scanner_clones"  # 临时目录
+
+report:
+  output_dir: "./reports"    # 报告保存目录
+  keep_files: true           # 是否保留报告文件
+```
+
+## 常见问题
+
+### Q: 扫描时间很长怎么办？
+
+A: 系统会浅克隆仓库（只获取最新提交），首次扫描后会有缓存。如果仍需优化，可以：
+- 减少扫描的文件类型
+- 调整 `max_issues` 参数
+
+### Q: 飞书消息发送失败？
+
+A: 检查：
+1. Webhook 地址是否正确
+2. 是否开启了签名校验（如果开启了，需要配置 secret）
+3. 网络是否可达
+
+### Q: 扫描不到代码？
+
+A: 检查：
+1. 仓库 URL 是否可公开访问
+2. 私有仓库需要配置 Git 凭证
+3. 确认分支名称正确
+
+### Q: 如何访问 Gitea 私有仓库？
+
+A: 在环境变量中配置 Git 凭证：
+```bash
+export GIT_USERNAME=your_username
+export GIT_PASSWORD=your_password
+```
+
+或者在 Git 克隆 URL 中包含凭证：
+```
+http://username:password@gitea-server.com/user/repo.git
+```
+
+## 系统架构图
+
+```
+用户提交代码
+    │
+    ▼
+Gitea Webhook ──────────────────────┐
+    │                                │
+    ▼                                │
+Webhook 服务                        │
+(Flask :5000)                       │
+    │                                │
+    ├──────────┬──────────┬─────────┘
+    ▼          ▼          ▼
+ Python    JS/TS     Security
+ Scanner   Scanner    Scanner
+    │          │          │
+    └──────────┴──────────┘
+              │
+              ▼
+        Report Generator
+        (Markdown 报告)
+              │
+              ▼
+         Feishu Bot
+         (发送通知)
+```
+
+## 目录结构
+
+```
+code-scanner/
+├── app.py                    # 主应用
+├── config.yaml               # 配置文件
+├── requirements.txt           # 依赖
+├── Dockerfile               # Docker 镜像
+├── docker-compose.yml       # Docker Compose
+├── install.bat              # Windows 安装脚本
+├── install.sh              # Linux 安装脚本
+├── README.md               # 项目说明
+├── 快速开始指南.md           # 本文档
+├── webhook/                 # Webhook 处理
+├── scanner/                 # 代码扫描器
+├── report/                  # 报告生成
+├── notify/                  # 飞书通知
+└── reports/                # 报告输出
+```