#!/usr/bin/env python3
"""知识库自动检查脚本（Lint）

自动修复：
- 索引一致性：缺失的文章补条目，失效的标记 [MISSING]
- 内部链接：断链修复（同一文件名存在 → 自动修正路径）
- 原始资料引用验证

报告（不自动修）：
- 孤立页面
- 缺少交叉引用的相关文章
"""
import os, re, json
from datetime import datetime

VAULT = "/home/gxd/蝶雅供应链"
WIKI = os.path.join(VAULT, "知识库")
RAW = os.path.join(VAULT, "原始资料")
INDEX_FILE = os.path.join(WIKI, "索引.md")
LOG_FILE = os.path.join(WIKI, "操作日志.md")

issues = []
auto_fixed = 0

def log(msg):
    print(f"  {msg}")

def get_md_files(directory):
    """获取目录下所有 md 文件（递归）"""
    files = {}
    for root, dirs, fnames in os.walk(directory):
        for f in fnames:
            if f.endswith('.md') and f not in ('索引.md', '操作日志.md'):
                rel = os.path.relpath(os.path.join(root, f), directory)
                files[f] = rel
    return files

def check_index_consistency():
    """检查索引一致性"""
    global auto_fixed
    if not os.path.exists(INDEX_FILE):
        issues.append("索引.md 不存在")
        return
    
    # 获取所有 wiki 文章
    wiki_files = get_md_files(WIKI)
    
    with open(INDEX_FILE, 'r', encoding='utf-8') as f:
        content = f.read()
    
    # 解析索引中的文章路径
    indexed_paths = set()
    for line in content.split('\n'):
        if '](' in line and line.startswith('|'):
            m = re.search(r'\]\(([^)]+)\)', line)
            if m:
                path = m.group(1)
                indexed_paths.add(path)
    
    # 检查：文件存在但索引中没有
    for fname, relpath in wiki_files.items():
        if relpath not in indexed_paths:
            log(f"⚠️ 文章存在但未在索引中: {relpath}")
            # 尝试自动补索引（需要知道主题）
            topic = relpath.split('/')[0] if '/' in relpath else None
            if topic:
                # 读取文章获取摘要
                summary = ""
                with open(os.path.join(WIKI, relpath), 'r', encoding='utf-8') as f:
                    for line in f:
                        if line.startswith('## 概述'):
                            next_line = next(f, '').strip()
                            summary = next_line[:50]
                            break
                        if line.startswith('> 来源') or line.startswith('> Source'):
                            continue
                
                # 生成新的索引行
                today = datetime.now().strftime('%Y-%m-%d')
                new_line = f"| [{fname[:-3]}]({relpath}) | {summary or '(无摘要)'} | {today} |"
                
                # 找到对应主题的表格最后一行，追加
                topic_header = f"## {topic}"
                if topic_header in content:
                    # 找到 topic 区域末尾，插入新行
                    lines = content.split('\n')
                    insert_pos = -1
                    in_topic = False
                    for i, line in enumerate(lines):
                        if line == topic_header:
                            in_topic = True
                            continue
                        if in_topic and line.startswith('## '):
                            break
                        if in_topic and line.startswith('|') and '暂无' in line:
                            # 替换"暂无"行
                            lines[i] = new_line
                            insert_pos = -2  # flag for replaced
                            break
                        if in_topic and line.startswith('| ['):
                            insert_pos = i + 1
                    
                    if insert_pos == -2:
                        # 已替换
                        pass
                    elif insert_pos > 0:
                        lines.insert(insert_pos, new_line)
                    content = '\n'.join(lines)
                    auto_fixed += 1
    
    # 检查：索引中有但文件不存在
    for path in indexed_paths:
        full_path = os.path.join(WIKI, path)
        if not os.path.exists(full_path):
            log(f"⚠️ 索引中的文章已删除: {path}")
            # 标记为 MISSING（避免重复追加）
            if '[MISSING]' not in content.split(f"]({path})")[1][:30]:
                content = content.replace(f"]({path})", f"]({path}) [MISSING]")
                auto_fixed += 1
            auto_fixed += 1
    
    # 写入更新后的索引
    with open(INDEX_FILE, 'w', encoding='utf-8') as f:
        f.write(content)

def check_internal_links():
    """检查文章内部的 Wiki 链接"""
    global auto_fixed
    wiki_files = {}
    for root, dirs, fnames in os.walk(WIKI):
        for f in fnames:
            if f.endswith('.md'):
                rel = os.path.relpath(os.path.join(root, f), WIKI)
                wiki_files[f] = rel
    
    for root, dirs, fnames in os.walk(WIKI):
        for fname in fnames:
            if not fname.endswith('.md') or fname in ('索引.md', '操作日志.md'):
                continue
            fpath = os.path.join(root, fname)
            with open(fpath, 'r', encoding='utf-8') as f:
                content = f.read()
            
            # 查找 markdown 链接
            for m in re.finditer(r'\[([^\]]+)\]\(([^)]+)\)', content):
                link_text = m.group(1)
                link_target = m.group(2)
                
                # 跳过外部链接
                if link_target.startswith('http://') or link_target.startswith('https://'):
                    continue

                # 跳过 file:// URI（本机文件链接）和页内锚点（#xxx）
                if link_target.startswith('file://') or link_target.startswith('#'):
                    continue
                
                # 检查路径是否有效
                # 如果是相对路径，基于当前文件目录解析
                base_dir = os.path.dirname(fpath)
                abs_target = os.path.normpath(os.path.join(base_dir, link_target))
                
                if not os.path.exists(abs_target):
                    target_file = os.path.basename(link_target)
                    if target_file in wiki_files:
                        # 找到了正确路径
                        correct_path = os.path.relpath(
                            os.path.join(WIKI, wiki_files[target_file]),
                            os.path.dirname(fpath)
                        )
                        if correct_path != link_target:
                            content = content.replace(
                                f']({link_target})', 
                                f']({correct_path})'
                            )
                            log(f"🔧 修复链接: {fname} → {link_target} → {correct_path}")
                            auto_fixed += 1
                    else:
                        issues.append(f"断链: {fname} → {link_target} (文件不存在)")
            
            with open(fpath, 'w', encoding='utf-8') as f:
                f.write(content)

def check_orphan_pages():
    """检查孤立页面（没有其他文章引用）"""
    all_content = ""
    for root, dirs, fnames in os.walk(WIKI):
        for fname in fnames:
            if fname.endswith('.md') and fname not in ('索引.md', '操作日志.md'):
                fpath = os.path.join(root, fname)
                with open(fpath, 'r', encoding='utf-8') as f:
                    all_content += f.read() + "\n"
    
    for root, dirs, fnames in os.walk(WIKI):
        for fname in fnames:
            if fname.endswith('.md') and fname not in ('索引.md', '操作日志.md'):
                # 检查是否有其他文章引用了它
                name_without_ext = fname[:-3]
                if name_without_ext not in all_content.replace(fname, ''):
                    rel = os.path.relpath(os.path.join(root, fname), WIKI)
                    if rel not in ('供应链流程/知识库搭建架构.md', '供应链流程/2026-06-20-知识库搭建全记录.md'):
                        issues.append(f"孤立页面: {rel} (没有入链)")

def update_log():
    """更新操作日志"""
    count_issues = len(issues)
    if count_issues == 0 and auto_fixed == 0:
        return
    
    today = datetime.now().strftime("%Y-%m-%d")
    with open(LOG_FILE, 'a', encoding='utf-8') as f:
        f.write(f"\n## [{today}] 检查 | {count_issues} 个问题，{auto_fixed} 处自动修复\n")
        for issue in issues:
            f.write(f"- {issue}\n")

if __name__ == '__main__':
    print(f"🔍 知识库检查开始...")
    check_index_consistency()
    check_internal_links()
    check_orphan_pages()
    update_log()
    print(f"\n✅ 检查完成: {len(issues)} 个问题, {auto_fixed} 处自动修复")
    if issues:
        for issue in issues:
            print(f"  ⚠️  {issue}")