microwind Git分析器技能
概述
Git历史是项目的时间线。正确读取它以了解代码如何演进。误读历史会在调试时浪费时间。
核心原则: Git告诉你什么改变了以及为什么改变。
何时使用
始终:
- 理解代码演进
- 查找错误何时引入
- 跟踪功能开发
- 理解重构
- 代码审查
- 性能分析
触发短语:
- "分析Git历史"
- "查找错误引入点"
- "代码演进分析"
- "Git提交历史"
- "分支分析"
- "代码回溯"
Git分析功能
历史分析
- 提交历史查看
- 分支演进跟踪
- 合并历史分析
- 标签历史查看
- 作者贡献统计
代码分析
- 文件变更历史
- 代码行数统计
- 热点文件识别
- 代码复杂度分析
- 技术债务检测
团队分析
- 开发者活跃度
- 贡献分布统计
- 协作模式分析
- 代码审查效率
- 工作时间分析
常见Git问题
提交信息不规范
问题:
提交信息不清晰或不规范
错误示例:
fix bug
update
temp commit
wip
解决方案:
使用规范的提交格式:
feat: 添加用户登录功能
fix: 修复登录验证错误
docs: 更新API文档
refactor: 重构用户服务模块
分支管理混乱
问题:
分支命名不规范，合并策略混乱
错误示例:
- feature1
- bug-fix-2
- test-branch
- master分支直接开发
解决方案:
- main/master: 主分支
- develop: 开发分支
- feature/xxx: 功能分支
- hotfix/xxx: 热修复分支
- release/xxx: 发布分支
合并冲突频发
问题:
频繁的合并冲突
原因:
- 功能分支过长时间不合并
- 多人修改同一文件
- 缺乏代码审查
解决方案:
1. 定期合并主分支
2. 小步提交，频繁集成
3. 建立代码审查流程
4. 使用功能开关
代码实现示例
Git分析器
import subprocess
import json
import re
from typing import Dict, List, Any, Optional, Tuple
from dataclasses import dataclass
from datetime import datetime
from collections import defaultdict, Counter
import os
@dataclass
class GitCommit:
"""Git提交信息"""
hash: str
author: str
email: str
date: datetime
message: str
files_changed: List[str]
insertions: int
deletions: int
branches: List[str]
@dataclass
class GitFile:
"""Git文件信息"""
path: str
size: int
last_modified: datetime
last_author: str
change_count: int
total_insertions: int
total_deletions: int
@dataclass
class GitAuthor:
"""Git作者信息"""
name: str
email: str
commits: int
insertions: int
deletions: int
files_touched: int
first_commit: datetime
last_commit: datetime
active_days: int
@dataclass
class GitAnalysis:
"""Git分析结果"""
total_commits: int
total_authors: int
total_files: int
date_range: Tuple[datetime, datetime]
top_authors: List[GitAuthor]
hot_files: List[GitFile]
commit_frequency: Dict[str, int]
class GitAnalyzer:
"""Git分析器"""
def __init__(self, repo_path: str = "."):
self.repo_path = repo_path
self.git_cmd = ["git", "-C", repo_path]
def _run_git_command(self, args: List[str]) -> str:
"""执行Git命令"""
try:
result = subprocess.run(
self.git_cmd + args,
capture_output=True,
text=True,
check=True
)
return result.stdout.strip()
except subprocess.CalledProcessError as e:
raise Exception(f"Git命令执行失败: {e.stderr}")
def get_commits(self, limit: int = 100, since: Optional[str] = None) -> List[GitCommit]:
"""获取提交历史"""
args = [
"log",
f"--max-count={limit}",
"--pretty=format:%H|%an|%ae|%ad|%s",
"--date=iso",
"--numstat",
"--name-only"
]
if since:
args.append(f"--since={since}")
output = self._run_git_command(args)
return self._parse_commits(output)
def _parse_commits(self, output: str) -> List[GitCommit]:
"""解析提交信息"""
commits = []
lines = output.split('\n')
i = 0
while i < len(lines):
if not lines[i]:
i += 1
continue
# 解析提交头信息
header = lines[i].split('|')
if len(header) >= 5:
commit_hash = header[0]
author = header[1]
email = header[2]
date = datetime.fromisoformat(header[3].replace(' ', 'T'))
message = '|'.join(header[4:])
i += 1
# 解析文件统计信息
insertions = 0
deletions = 0
files_changed = []
while i < len(lines) and lines[i]:
line = lines[i]
if '\t' in line:
parts = line.split('\t')
if len(parts) == 3:
try:
ins = int(parts[0]) if parts[0] != '-' else 0
dels = int(parts[1]) if parts[1] != '-' else 0
file_path = parts[2]
insertions += ins
deletions += dels
files_changed.append(file_path)
except ValueError:
pass
i += 1
# 获取分支信息
branches = self._get_commit_branches(commit_hash)
commits.append(GitCommit(
hash=commit_hash,
author=author,
email=email,
date=date,
message=message,
files_changed=files_changed,
insertions=insertions,
deletions=deletions,
branches=branches
))
else:
i += 1
return commits
def _get_commit_branches(self, commit_hash: str) -> List[str]:
"""获取提交所在的分支"""
try:
output = self._run_git_command([
"branch", "--contains", commit_hash
])
branches = []
for line in output.split('\n'):
branch = line.strip().replace('* ', '')
if branch and branch != 'HEAD':
branches.append(branch)
return branches
except:
return []
def get_file_history(self, file_path: str) -> List[GitCommit]:
"""获取文件历史"""
output = self._run_git_command([
"log",
"--pretty=format:%H|%an|%ae|%ad|%s",
"--date=iso",
"--numstat",
file_path
])
return self._parse_commits(output)
def get_authors(self) -> List[GitAuthor]:
"""获取作者统计"""
output = self._run_git_command([
"log",
"--pretty=format:%an|%ae|%ad",
"--date=iso",
"--numstat"
])
author_stats = defaultdict(lambda: {
'commits': 0,
'insertions': 0,
'deletions': 0,
'files': set(),
'dates': [],
'first_commit': None,
'last_commit': None
})
lines = output.split('\n')
i = 0
while i < len(lines):
if not lines[i]:
i += 1
continue
if '|' in lines[i]:
# 提交信息
parts = lines[i].split('|')
if len(parts) >= 3:
author = parts[0]
email = parts[1]
date = datetime.fromisoformat(parts[2].replace(' ', 'T'))
stats = author_stats[author]
stats['commits'] += 1
stats['dates'].append(date)
if not stats['first_commit'] or date < stats['first_commit']:
stats['first_commit'] = date
if not stats['last_commit'] or date > stats['last_commit']:
stats['last_commit'] = date
i += 1
# 文件统计
while i < len(lines) and lines[i] and '\t' in lines[i]:
file_parts = lines[i].split('\t')
if len(file_parts) >= 3:
file_path = file_parts[2]
stats['files'].add(file_path)
try:
ins = int(file_parts[0]) if file_parts[0] != '-' else 0
dels = int(file_parts[1]) if file_parts[1] != '-' else 0
stats['insertions'] += ins
stats['deletions'] += dels
except ValueError:
pass
i += 1
else:
i += 1
# 构建作者对象
authors = []
for author_name, stats in author_stats.items():
active_days = len(set(d.date() for d in stats['dates']))
authors.append(GitAuthor(
name=author_name,
email="", # 需要额外获取
commits=stats['commits'],
insertions=stats['insertions'],
deletions=stats['deletions'],
files_touched=len(stats['files']),
first_commit=stats['first_commit'],
last_commit=stats['last_commit'],
active_days=active_days
))
return sorted(authors, key=lambda x: x.commits, reverse=True)
def get_hot_files(self, limit: int = 10) -> List[GitFile]:
"""获取热点文件"""
output = self._run_git_command([
"log",
"--name-only",
"--pretty=format:"
])
file_stats = defaultdict(lambda: {
'count': 0,
'last_modified': None,
'last_author': None,
'insertions': 0,
'deletions': 0
})
lines = output.split('\n')
current_commit = None
for line in lines:
if line.strip():
if not line.startswith(' '):
# 新的提交
current_commit = line
else:
# 文件路径
file_path = line.strip()
if file_path and file_path != '':
stats = file_stats[file_path]
stats['count'] += 1
# 这里可以添加更多统计信息
# 获取文件详细信息
hot_files = []
for file_path, stats in file_stats.items():
try:
# 获取文件大小
full_path = os.path.join(self.repo_path, file_path)
if os.path.exists(full_path):
size = os.path.getsize(full_path)
else:
size = 0
# 获取最后修改信息
last_commit_info = self._get_last_file_commit(file_path)
hot_files.append(GitFile(
path=file_path,
size=size,
last_modified=last_commit_info.get('date', datetime.now()),
last_author=last_commit_info.get('author', ''),
change_count=stats['count'],
total_insertions=stats['insertions'],
total_deletions=stats['deletions']
))
except:
continue
return sorted(hot_files, key=lambda x: x.change_count, reverse=True)[:limit]
def _get_last_file_commit(self, file_path: str) -> Dict[str, Any]:
"""获取文件的最后提交信息"""
try:
output = self._run_git_command([
"log",
"-1",
"--pretty=format:%an|%ad",
"--date=iso",
file_path
])
if output:
parts = output.split('|')
return {
'author': parts[0] if len(parts) > 0 else '',
'date': datetime.fromisoformat(parts[1].replace(' ', 'T')) if len(parts) > 1 else datetime.now()
}
except:
pass
return {'author': '', 'date': datetime.now()}
def analyze_repository(self) -> GitAnalysis:
"""分析整个仓库"""
# 获取基本统计
total_commits = int(self._run_git_command(["rev-list", "--count", "HEAD"]))
# 获取作者数量
authors = self.get_authors()
total_authors = len(authors)
# 获取文件数量
output = self._run_git_command(["ls-files"])
total_files = len(output.split('\n')) if output else 0
# 获取日期范围
first_date_str = self._run_git_command(["log", "--reverse", "--pretty=format:%ad", "--date=iso"]).split('\n')[0]
last_date_str = self._run_git_command(["log", "-1", "--pretty=format:%ad", "--date=iso"])
first_date = datetime.fromisoformat(first_date_str.replace(' ', 'T'))
last_date = datetime.fromisoformat(last_date_str.replace(' ', 'T'))
# 获取热点文件
hot_files = self.get_hot_files()
# 获取提交频率
commit_frequency = self._get_commit_frequency()
return GitAnalysis(
total_commits=total_commits,
total_authors=total_authors,
total_files=total_files,
date_range=(first_date, last_date),
top_authors=authors[:10],
hot_files=hot_files,
commit_frequency=commit_frequency
)
def _get_commit_frequency(self) -> Dict[str, int]:
"""获取提交频率"""
output = self._run_git_command([
"log",
"--pretty=format:%ad",
"--date=format:%Y-%m-%d"
])
dates = output.split('\n')
return dict(Counter(dates))
def find_bug_introduction(self, bug_description: str) -> List[GitCommit]:
"""查找错误引入点"""
# 使用git bisect查找错误引入点
# 这里简化实现，实际需要更复杂的逻辑
commits = self.get_commits(limit=200)
suspicious_commits = []
for commit in commits:
# 简单的关键词匹配
if any(keyword in commit.message.lower() for keyword in ['fix', 'bug', 'error', 'issue']):
suspicious_commits.append(commit)
return suspicious_commits
def generate_analysis_report(self, analysis: GitAnalysis) -> str:
"""生成分析报告"""
report = ["=== Git仓库分析报告 ===\n"]
# 基本统计
report.append("=== 基本统计 ===")
report.append(f"总提交数: {analysis.total_commits}")
report.append(f"总作者数: {analysis.total_authors}")
report.append(f"总文件数: {analysis.total_files}")
report.append(f"时间范围: {analysis.date_range[0].strftime('%Y-%m-%d')} 到 {analysis.date_range[1].strftime('%Y-%m-%d')}")
# 计算项目年龄
project_age = (analysis.date_range[1] - analysis.date_range[0]).days
report.append(f"项目年龄: {project_age} 天")
if project_age > 0:
avg_commits_per_day = analysis.total_commits / project_age
report.append(f"平均每天提交: {avg_commits_per_day:.2f}")
report.append("")
# 顶级贡献者
report.append("=== 顶级贡献者 ===")
for i, author in enumerate(analysis.top_authors[:10], 1):
report.append(f"{i}. {author.name}")
report.append(f" 提交数: {author.commits}")
report.append(f" 代码行数: +{author.insertions} -{author.deletions}")
report.append(f" 活跃天数: {author.active_days}")
report.append("")
# 热点文件
report.append("=== 热点文件 ===")
for i, file in enumerate(analysis.hot_files[:10], 1):
report.append(f"{i}. {file.path}")
report.append(f" 修改次数: {file.change_count}")
report.append(f" 文件大小: {file.size} bytes")
report.append(f" 最后修改: {file.last_modified.strftime('%Y-%m-%d')}")
report.append("")
# 提交频率
report.append("=== 最近提交频率 ===")
recent_dates = sorted(analysis.commit_frequency.items(), reverse=True)[:10]
for date, count in recent_dates:
report.append(f"{date}: {count} 次提交")
return '\n'.join(report)
# 使用示例
def main():
analyzer = GitAnalyzer("/path/to/repo")
# 分析仓库
analysis = analyzer.analyze_repository()
# 生成报告
report = analyzer.generate_analysis_report(analysis)
print(report)
# 查找错误引入点
bug_commits = analyzer.find_bug_introduction("login error")
print(f"\n找到 {len(bug_commits)} 个可能相关的提交")
if __name__ == "__main__":
main()
Git代码审查工具
import re
from typing import List, Dict, Any
class GitCodeReviewer:
"""Git代码审查工具"""
def __init__(self, repo_path: str = "."):
self.analyzer = GitAnalyzer(repo_path)
self.review_rules = self._initialize_review_rules()
def review_commit(self, commit_hash: str) -> Dict[str, Any]:
"""审查单个提交"""
commits = self.analyzer.get_commits(limit=1)
if not commits or commits[0].hash != commit_hash:
# 获取特定提交
output = self.analyzer._run_git_command([
"show",
"--pretty=format:%H|%an|%ae|%ad|%s",
"--date=iso",
"--numstat",
commit_hash
])
commits = self.analyzer._parse_commits(output)
if not commits:
return {"error": "提交不存在"}
commit = commits[0]
# 审查提交信息
message_issues = self._review_commit_message(commit.message)
# 审查代码变更
code_issues = self._review_code_changes(commit)
# 审查文件变更
file_issues = self._review_file_changes(commit)
return {
"commit": commit,
"issues": {
"message": message_issues,
"code": code_issues,
"files": file_issues
},
"score": self._calculate_review_score(message_issues, code_issues, file_issues)
}
def _review_commit_message(self, message: str) -> List[str]:
"""审查提交信息"""
issues = []
# 检查长度
if len(message) > 72:
issues.append("提交信息过长，建议不超过72个字符")
# 检查格式
if not re.match(r'^(feat|fix|docs|style|refactor|test|chore)(\(.+\))?: .+', message):
issues.append("提交信息格式不规范，建议使用类型: 描述格式")
# 检查是否包含必要信息
if len(message) < 10:
issues.append("提交信息过于简单，缺少详细描述")
# 检查是否包含敏感信息
sensitive_patterns = [
r'password',
r'secret',
r'key',
r'token',
r'credential'
]
for pattern in sensitive_patterns:
if re.search(pattern, message, re.IGNORECASE):
issues.append(f"提交信息可能包含敏感信息: {pattern}")
return issues
def _review_code_changes(self, commit: GitCommit) -> List[str]:
"""审查代码变更"""
issues = []
# 检查变更量
total_changes = commit.insertions + commit.deletions
if total_changes > 1000:
issues.append("单次提交变更量过大，建议拆分为多个小提交")
# 检查删除比例
if commit.deletions > commit.insertions * 2:
issues.append("删除代码过多，可能需要更好的重构策略")
# 检查文件数量
if len(commit.files_changed) > 20:
issues.append("修改文件过多，可能影响多个功能模块")
return issues
def _review_file_changes(self, commit: GitCommit) -> List[str]:
"""审查文件变更"""
issues = []
# 检查敏感文件
sensitive_files = [
'config.py',
'.env',
'secrets.json',
'private.key',
'id_rsa'
]
for file_path in commit.files_changed:
for sensitive in sensitive_files:
if sensitive in file_path:
issues.append(f"修改了敏感文件: {file_path}")
break
# 检查测试文件
test_files = [f for f in commit.files_changed if 'test' in f.lower()]
if not test_files and len(commit.files_changed) > 5:
issues.append("大量代码变更但没有测试文件修改")
return issues
def _calculate_review_score(self, message_issues: List[str],
code_issues: List[str],
file_issues: List[str]) -> int:
"""计算审查分数"""
total_issues = len(message_issues) + len(code_issues) + len(file_issues)
# 基础分数100，每个问题扣10分
score = max(0, 100 - total_issues * 10)
return score
def review_branch(self, branch_name: str) -> Dict[str, Any]:
"""审查分支"""
# 获取分支的所有提交
output = self.analyzer._run_git_command([
"log",
f"{branch_name}..HEAD",
"--pretty=format:%H"
])
commit_hashes = output.split('\n') if output else []
branch_reviews = []
total_score = 0
for commit_hash in commit_hashes:
if commit_hash.strip():
review = self.review_commit(commit_hash.strip())
branch_reviews.append(review)
total_score += review.get('score', 0)
avg_score = total_score / len(branch_reviews) if branch_reviews else 0
return {
"branch": branch_name,
"commits": branch_reviews,
"total_commits": len(branch_reviews),
"average_score": avg_score,
"recommendation": self._get_branch_recommendation(avg_score)
}
def _get_branch_recommendation(self, score: float) -> str:
"""获取分支建议"""
if score >= 80:
return "可以合并"
elif score >= 60:
return "建议修复问题后合并"
else:
return "不建议合并，需要重大改进"
def _initialize_review_rules(self) -> Dict[str, Any]:
"""初始化审查规则"""
return {
"max_commit_size": 1000,
"max_files_per_commit": 20,
"max_message_length": 72,
"required_message_format": r'^(feat|fix|docs|style|refactor|test|chore)(\(.+\))?: .+',
"sensitive_files": [
'config.py', '.env', 'secrets.json', 'private.key', 'id_rsa'
]
}
# 使用示例
def main():
reviewer = GitCodeReviewer("/path/to/repo")
# 审查最新提交
latest_commit = reviewer.analyzer._run_git_command(["rev-parse", "HEAD"])
review_result = reviewer.review_commit(latest_commit)
print("=== 代码审查报告 ===")
print(f"提交: {review_result['commit'].hash}")
print(f"作者: {review_result['commit'].author}")
print(f"分数: {review_result['score']}")
if review_result['issues']['message']:
print("\n提交信息问题:")
for issue in review_result['issues']['message']:
print(f" - {issue}")
if review_result['issues']['code']:
print("\n代码变更问题:")
for issue in review_result['issues']['code']:
print(f" - {issue}")
if review_result['issues']['files']:
print("\n文件变更问题:")
for issue in review_result['issues']['files']:
print(f" - {issue}")
if __name__ == "__main__":
main()
Git最佳实践
提交规范
- 原子提交: 每个提交只做一件事
- 清晰信息: 使用描述性的提交信息
- 格式统一: 遵循团队约定的格式
- 及时提交: 频繁提交，小步快跑
分支策略
- 主分支保护: 主分支只接受合并
- 功能分支: 每个功能使用独立分支
- 定期同步: 定期合并主分支更新
- 清理分支: 及时删除已合并的分支
代码审查
- 强制审查: 所有代码必须经过审查
- 审查清单: 使用标准化的审查清单
- 及时反馈: 快速响应审查请求
- 建设性意见: 提供具体的改进建议
相关技能
- code-review - 代码审查
- version-control - 版本控制
- project-management - 项目管理
- team-collaboration - 团队协作