Feishu Work Archive

#!/bin/bash
# 使用 agent-browser 批量导出飞书文档

PROFILE_DIR="$HOME/.openclaw/workspace/skills/feishu-work-archive/agent-browser-profile"
BATCHES_FILE="$HOME/.openclaw/workspace/skills/feishu-work-archive/batches.json"
PROGRESS_FILE="$HOME/.openclaw/workspace/skills/feishu-work-archive/progress-agent.json"
OUTPUT_DIR="$HOME/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档"

# 初始化进度
if [ ! -f "$PROGRESS_FILE" ]; then
  echo '{"completed":0,"failed":[],"results":[]}' > "$PROGRESS_FILE"
fi

# 读取进度
COMPLETED=$(jq -r '.completed' "$PROGRESS_FILE")
TOTAL=$(jq '[.batches[][]] | length' "$BATCHES_FILE")

echo "进度: $COMPLETED/$TOTAL"
echo "首次运行需要手动登录飞书..."
echo ""

# 打开飞书首页让用户登录
agent-browser --profile "$PROFILE_DIR" open https://ottno-cd.feishu.cn/drive/home/

echo ""
echo "请在浏览器中登录飞书，登录完成后按回车继续..."
read

echo "开始批量导出..."

#!/usr/bin/env node
/**
 * 飞书工作文档归档工具
 * 用途：导出工作飞书文档到 Obsidian，AI 分析主题并整理
 */

const https = require('https');
const fs = require('fs');
const path = require('path');

// 配置
const CONFIG = {
  appId: 'cli_a75960290439500d',
  appSecret: 'tBd5CYrc3jFTa4yQ9aD3whBMyMvDfl2w',
  obsidianPath: '/Users/maygo/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档',
  domain: 'feishu.cn'
};

// 确保输出目录存在
function ensureDir(dir) {
  if (!fs.existsSync(dir)) {
    fs.mkdirSync(dir, { recursive: true });
  }
}

// 获取 tenant_access_token
async function getTenantToken() {
  return new Promise((resolve, reject) => {
    const data = JSON.stringify({
      app_id: CONFIG.appId,
      app_secret: CONFIG.appSecret
    });

    const options = {
      hostname: `open.${CONFIG.domain}`,
      path: '/open-apis/auth/v3/tenant_access_token/internal',
      method: 'POST',
      headers: {
        'Content-Type': 'application/json',
        'Content-Length': data.length
      }
    };

    const req = https.request(options, (res) => {
      let body = '';
      res.on('data', (chunk) => body += chunk);
      res.on('end', () => {
        const result = JSON.parse(body);
        if (result.code === 0) {
          resolve(result.tenant_access_token);
        } else {
          reject(new Error(`获取 token 失败: ${JSON.stringify(result)}`));
        }
      });
    });

    req.on('error', reject);
    req.write(data);
    req.end();
  });
}

// 通用 API 调用
async function callFeishuAPI(token, method, path, data = null) {
  return new Promise((resolve, reject) => {
    const options = {
      hostname: `open.${CONFIG.domain}`,
      path: path,
      method: method,
      headers: {
        'Authorization': `Bearer ${token}`,
        'Content-Type': 'application/json'
      }
    };

    if (data) {
      const jsonData = JSON.stringify(data);
      options.headers['Content-Length'] = jsonData.length;
    }

    const req = https.request(options, (res) => {
      let body = '';
      res.on('data', (chunk) => body += chunk);
      res.on('end', () => {
        try {
          const result = JSON.parse(body);
          resolve(result);
        } catch (e) {
          reject(new Error(`解析响应失败: ${body}`));
        }
      });
    });

    req.on('error', reject);
    if (data) {
      req.write(JSON.stringify(data));
    }
    req.end();
  });
}

module.exports = {
  CONFIG,
  ensureDir,
  getTenantToken,
  callFeishuAPI
};

#!/usr/bin/env node
/**
 * 自动批量导出飞书文档
 * Headless 模式，后台运行，每 5 分钟报告进度
 */

const fs = require('fs');
const path = require('path');

const CONFIG = {
  batchesFile: path.join(__dirname, 'batches.json'),
  progressFile: path.join(__dirname, 'progress.json'),
  outputDir: '/Users/maygo/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档',
  targetId: 'CC074F4EEB55CC58F90083D3AF5A7676',
  reportInterval: 5 * 60 * 1000 // 5 分钟
};

// 读取批次
const { batches } = JSON.parse(fs.readFileSync(CONFIG.batchesFile, 'utf8'));

// 读取或初始化进度
let progress = { completed: 0, failed: [], lastReport: Date.now() };
if (fs.existsSync(CONFIG.progressFile)) {
  progress = JSON.parse(fs.readFileSync(CONFIG.progressFile, 'utf8'));
}

// 保存进度
function saveProgress() {
  fs.writeFileSync(CONFIG.progressFile, JSON.stringify(progress, null, 2));
}

// 确保目录存在
function ensureDir(dir) {
  if (!fs.existsSync(dir)) {
    fs.mkdirSync(dir, { recursive: true });
  }
}

// 生成安全文件名
function sanitizeFilename(name) {
  return name.replace(/[\/\\:*?"<>|]/g, '_').substring(0, 100);
}

console.log('开始批量导出...');
console.log(`总文档数: ${batches.flat().length}`);
console.log(`已完成: ${progress.completed}`);
console.log(`剩余: ${batches.flat().length - progress.completed}\n`);

module.exports = { CONFIG, batches, progress, saveProgress, ensureDir, sanitizeFilename };

#!/usr/bin/env node
import { chromium } from 'playwright';
import fs from 'fs';
import path from 'path';
import { fileURLToPath } from 'url';

const __dirname = path.dirname(fileURLToPath(import.meta.url));

const CONFIG = {
  batchesFile: path.join(__dirname, 'batches.json'),
  progressFile: path.join(__dirname, 'progress.json'),
  outputDir: '/Users/maygo/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档',
  userDataDir: path.join(__dirname, 'browser-data'),
  reportInterval: 5 * 60 * 1000
};

const { batches } = JSON.parse(fs.readFileSync(CONFIG.batchesFile, 'utf8'));
const allDocs = batches.flat();

let progress = { completed: 0, failed: [], lastReport: Date.now(), results: [] };
if (fs.existsSync(CONFIG.progressFile)) {
  progress = JSON.parse(fs.readFileSync(CONFIG.progressFile, 'utf8'));
}

function saveProgress() {
  fs.writeFileSync(CONFIG.progressFile, JSON.stringify(progress, null, 2));
}

function ensureDir(dir) {
  if (!fs.existsSync(dir)) fs.mkdirSync(dir, { recursive: true });
}

function sanitize(name) {
  return name.replace(/[\/\\:*?"<>|]/g, '_').substring(0, 100);
}

async function extractDocContent(page) {
  await page.waitForTimeout(3000);
  const content = await page.evaluate(() => {
    const main = document.querySelector('[role="main"]') || document.body;
    return main.innerText || '';
  });
  return content;
}

async function processDoc(page, doc) {
  try {
    await page.goto(doc.url, { waitUntil: 'domcontentloaded', timeout: 60000 });
    const content = await extractDocContent(page);
    
    if (content.includes('扫码登录') || content.includes('账号登录')) {
      throw new Error('需要登录');
    }
    
    const categoryDir = path.join(CONFIG.outputDir, doc.category);
    ensureDir(categoryDir);
    
    const filename = `${sanitize(doc.name)}.md`;
    const frontmatter = `---
title: ${doc.name}
category: ${doc.category}
source: 工作飞书
url: ${doc.url}
exported_at: ${new Date().toISOString()}
---

`;
    
    fs.writeFileSync(path.join(categoryDir, filename), frontmatter + content, 'utf8');
    
    progress.completed++;
    progress.results.push({ id: doc.id, name: doc.name, status: 'success' });
    
    console.log(`✓ [${progress.completed}/${allDocs.length}] ${doc.name}`);
    
  } catch (err) {
    progress.failed.push({ id: doc.id, name: doc.name, error: err.message });
    console.error(`✗ ${doc.name}: ${err.message}`);
  } finally {
    saveProgress();
  }
}

async function main() {
  console.log('启动浏览器，等待 10 秒让你登录...');
  console.log(`总文档数: ${allDocs.length}\n`);
  
  const browser = await chromium.launchPersistentContext(CONFIG.userDataDir, {
    headless: false,
    args: ['--disable-blink-features=AutomationControlled']
  });
  
  const page = await browser.newPage();
  await page.goto('https://ottno-cd.feishu.cn/drive/home/');
  
  console.log('等待 10 秒...');
  await page.waitForTimeout(10000);
  
  console.log('\n开始导出...\n');
  
  const startFrom = progress.completed;
  
  for (let i = startFrom; i < allDocs.length; i++) {
    await processDoc(page, allDocs[i]);
    
    if (Date.now() - progress.lastReport > CONFIG.reportInterval) {
      console.log(`\n进度: ${progress.completed}/${allDocs.length} (${Math.round(progress.completed/allDocs.length*100)}%)`);
      console.log(`失败: ${progress.failed.length}\n`);
      progress.lastReport = Date.now();
    }
  }
  
  await browser.close();
  
  console.log(`\n✓ 完成！成功: ${progress.completed - progress.failed.length} 失败: ${progress.failed.length}`);
}

main().catch(console.error);

#!/usr/bin/env node
/**
 * 批量导出飞书工作文档到 Obsidian
 * 用法: node batch-export.js <doc-list.md>
 */

const fs = require('fs');
const path = require('path');

const OBSIDIAN_BASE = '/Users/maygo/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档';
const DOC_LIST_PATH = process.argv[2] || '/Users/maygo/.openclaw/media/inbound/梅杭俊_云文档列表---eb5002fd-4e2c-47d6-880a-e941538a424f';

// 解析文档列表
function parseDocList(content) {
  const lines = content.split('\n');
  const docs = [];
  
  for (const line of lines) {
    const match = line.match(/\|\s*(\d+)\s*\|\s*(.+?)\s*\|\s*(https:\/\/[^\s]+)/);
    if (match) {
      const [, id, name, url] = match;
      const tokenMatch = url.match(/\/([a-zA-Z0-9]+)$/);
      if (tokenMatch) {
        docs.push({
          id: parseInt(id),
          name: name.trim(),
          url: url.trim(),
          token: tokenMatch[1]
        });
      }
    }
  }
  
  return docs;
}

// 主题分类规则
function categorizeDoc(name) {
  if (name.includes('降本') || name.includes('预算') || name.includes('EBITDA') || name.includes('现金流')) {
    return '财务预算管理';
  }
  if (name.includes('智能纪要') || name.includes('文字记录') || name.includes('会议速递')) {
    return '会议纪要';
  }
  if (name.includes('周工作报告') || name.includes('工作汇报')) {
    return '工作汇报';
  }
  if (name.includes('Excel') || name.includes('PowerQuery') || name.includes('多维表格') || name.includes('技术文档')) {
    return '技术工具';
  }
  if (name.includes('门店') || name.includes('闭店') || name.includes('经营分析')) {
    return '运营管理';
  }
  return '其他文档';
}

// 生成安全文件名
function sanitizeFilename(name) {
  return name
    .replace(/[\/\\:*?"<>|]/g, '_')
    .replace(/\s+/g, '_')
    .substring(0, 100);
}

// 主函数
async function main() {
  console.log('读取文档列表...');
  const content = fs.readFileSync(DOC_LIST_PATH, 'utf8');
  const docs = parseDocList(content);
  
  console.log(`找到 ${docs.length} 个文档\n`);
  
  // 按主题分组
  const grouped = {};
  for (const doc of docs) {
    const category = categorizeDoc(doc.name);
    if (!grouped[category]) grouped[category] = [];
    grouped[category].push(doc);
  }
  
  // 输出分类统计
  console.log('文档分类统计:');
  for (const [category, items] of Object.entries(grouped)) {
    console.log(`  ${category}: ${items.length} 个`);
  }
  
  // 生成导出清单
  const manifest = {
    total: docs.length,
    categories: grouped,
    exportTime: new Date().toISOString(),
    obsidianPath: OBSIDIAN_BASE
  };
  
  fs.writeFileSync(
    path.join(__dirname, 'export-manifest.json'),
    JSON.stringify(manifest, null, 2)
  );
  
  console.log('\n✓ 导出清单已生成: export-manifest.json');
  console.log('\n下一步: 使用 OpenClaw feishu_doc 工具批量读取文档');
}

main().catch(console.error);

#!/usr/bin/env node
/**
 * 使用 OpenClaw browser 工具批量导出
 * 在同一个标签页中逐个访问
 */

const fs = require('fs');
const path = require('path');

const CONFIG = {
  batchesFile: path.join(__dirname, 'batches.json'),
  progressFile: path.join(__dirname, 'progress-browser.json'),
  outputDir: '/Users/maygo/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档'
};

const { batches } = JSON.parse(fs.readFileSync(CONFIG.batchesFile, 'utf8'));
const allDocs = batches.flat();

let progress = { completed: 0, failed: [], results: [] };
if (fs.existsSync(CONFIG.progressFile)) {
  progress = JSON.parse(fs.readFileSync(CONFIG.progressFile, 'utf8'));
}

// 输出下一个要处理的文档
const nextDoc = allDocs[progress.completed];
if (!nextDoc) {
  console.log('全部完成！');
  process.exit(0);
}

console.log(JSON.stringify({
  action: 'next',
  doc: nextDoc,
  progress: `${progress.completed}/${allDocs.length}`
}));

#!/usr/bin/env node
/**
 * 通过浏览器批量导出飞书文档
 */

const fs = require('fs');
const path = require('path');

const MANIFEST_PATH = path.join(__dirname, 'export-manifest.json');
const OUTPUT_DIR = '/Users/maygo/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档';
const PROGRESS_FILE = path.join(__dirname, 'export-progress.json');

// 读取清单
const manifest = JSON.parse(fs.readFileSync(MANIFEST_PATH, 'utf8'));
const allDocs = [];

for (const [category, docs] of Object.entries(manifest.categories)) {
  for (const doc of docs) {
    allDocs.push({ ...doc, category });
  }
}

console.log(`总共 ${allDocs.length} 个文档`);
console.log(`输出目录: ${OUTPUT_DIR}`);
console.log('\n需要通过 OpenClaw browser 工具逐个访问并提取内容');
console.log('建议分批执行，每批 20-30 个文档\n');

// 生成批次
const BATCH_SIZE = 20;
const batches = [];
for (let i = 0; i < allDocs.length; i += BATCH_SIZE) {
  batches.push(allDocs.slice(i, i + BATCH_SIZE));
}

console.log(`已分为 ${batches.length} 批，每批 ${BATCH_SIZE} 个文档`);
console.log('\n第一批文档:');
batches[0].slice(0, 5).forEach((doc, i) => {
  console.log(`  ${i+1}. ${doc.name} (${doc.category})`);
});

// 保存批次信息
fs.writeFileSync(
  path.join(__dirname, 'batches.json'),
  JSON.stringify({ batches, batchSize: BATCH_SIZE }, null, 2)
);

console.log('\n✓ 批次信息已保存到 batches.json');

#!/bin/bash
set -e

PROFILE="$HOME/.openclaw/workspace/skills/feishu-work-archive/profile"
OUTPUT_DIR="$HOME/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档"
BATCHES="$HOME/.openclaw/workspace/skills/feishu-work-archive/batches.json"
PROGRESS="$HOME/.openclaw/workspace/skills/feishu-work-archive/progress-final.json"

# 初始化进度
if [ ! -f "$PROGRESS" ]; then
  echo '{"completed":0,"failed":[]}' > "$PROGRESS"
fi

completed=$(jq -r '.completed' "$PROGRESS")
total=$(jq '[.batches[][]] | length' "$BATCHES")

echo "继续从第 $((completed + 1)) 个开始，共 $total 个"
echo ""

# 读取所有文档，跳过已完成的
docs=$(jq -r "[.batches[][]][$completed:] | .[] | @json" "$BATCHES")

count=$completed
while IFS= read -r doc; do
  count=$((count + 1))
  
  name=$(echo "$doc" | jq -r '.name')
  url=$(echo "$doc" | jq -r '.url')
  category=$(echo "$doc" | jq -r '.category')
  
  echo "[$count/$total] $name"
  
  # 访问文档
  if ! agent-browser --profile "$PROFILE" open "$url" > /dev/null 2>&1; then
    echo "  ✗ 访问失败"
    jq ".failed += [{\"name\": \"$name\", \"error\": \"访问失败\"}]" "$PROGRESS" > "$PROGRESS.tmp" && mv "$PROGRESS.tmp" "$PROGRESS"
    continue
  fi
  sleep 2
  
  # 提取内容
  content=$(agent-browser eval "document.querySelector('[role=\"main\"]')?.innerText || document.body.innerText" 2>/dev/null || echo "")
  
  if [ -z "$content" ]; then
    echo "  ✗ 内容为空"
    jq ".failed += [{\"name\": \"$name\", \"error\": \"内容为空\"}]" "$PROGRESS" > "$PROGRESS.tmp" && mv "$PROGRESS.tmp" "$PROGRESS"
    jq ".completed = $count" "$PROGRESS" > "$PROGRESS.tmp" && mv "$PROGRESS.tmp" "$PROGRESS"
    continue
  fi
  
  # 保存
  category_dir="$OUTPUT_DIR/$category"
  mkdir -p "$category_dir"
  
  filename=$(echo "$name" | sed 's/[\/\\:*?"<>|]/_/g' | cut -c1-100).md
  filepath="$category_dir/$filename"
  
  cat > "$filepath" <<EOF
---
title: $name
category: $category
source: 工作飞书
url: $url
exported_at: $(date -u +"%Y-%m-%dT%H:%M:%SZ")
---

$content
EOF
  
  echo "  ✓ 已保存"
  
  # 更新进度
  jq ".completed = $count" "$PROGRESS" > "$PROGRESS.tmp" && mv "$PROGRESS.tmp" "$PROGRESS"
  
  # 每 20 个报告一次
  if [ $((count % 20)) -eq 0 ]; then
    failed=$(jq '.failed | length' "$PROGRESS")
    echo ""
    echo "进度: $count/$total (失败: $failed)"
    echo ""
  fi
  
done <<< "$docs"

failed=$(jq '.failed | length' "$PROGRESS")
echo ""
echo "全部完成！成功: $((count - failed)) 失败: $failed"

#!/usr/bin/env node
const { getTenantToken, callFeishuAPI, CONFIG, ensureDir } = require('./archive.js');
const fs = require('fs');
const path = require('path');

// 列出云空间文件
async function listDriveFiles(token, folderToken = null) {
  const params = new URLSearchParams({
    page_size: 50,
    ...(folderToken && { folder_token: folderToken })
  });
  
  const result = await callFeishuAPI(token, 'GET', `/open-apis/drive/v1/files?${params}`);
  return result;
}

// 读取文档内容
async function readDocument(token, docToken) {
  const result = await callFeishuAPI(token, 'GET', `/open-apis/docx/v1/documents/${docToken}/raw_content`);
  return result;
}

// 保存文档到本地
function saveToObsidian(filename, content, metadata = {}) {
  ensureDir(CONFIG.obsidianPath);
  
  const frontmatter = `---
title: ${metadata.title || filename}
source: feishu
doc_token: ${metadata.doc_token || ''}
exported_at: ${new Date().toISOString()}
tags: [工作归档, 飞书]
---

`;
  
  const fullContent = frontmatter + content;
  const filepath = path.join(CONFIG.obsidianPath, `${filename}.md`);
  fs.writeFileSync(filepath, fullContent, 'utf8');
  console.log(`✓ 已保存: ${filepath}`);
  return filepath;
}

// 主函数
async function main() {
  const cmd = process.argv[2] || 'list';
  
  try {
    console.log('获取访问令牌...');
    const token = await getTenantToken();
    
    if (cmd === 'list') {
      console.log('\n列出云空间文件...');
      const files = await listDriveFiles(token);
      
      if (files.code === 0 && files.data?.files) {
        console.log(`\n找到 ${files.data.files.length} 个文件:\n`);
        files.data.files.forEach((file, i) => {
          console.log(`${i+1}. [${file.type}] ${file.name}`);
          console.log(`   Token: ${file.token}`);
        });
      } else {
        console.log('返回:', JSON.stringify(files, null, 2));
      }
    } else if (cmd === 'export') {
      const docToken = process.argv[3];
      if (!docToken) {
        console.error('用法: node list-export.js export <doc_token>');
        process.exit(1);
      }
      
      console.log(`\n读取文档 ${docToken}...`);
      const doc = await readDocument(token, docToken);
      
      if (doc.code === 0 && doc.data?.content) {
        const filename = `doc_${docToken}_${Date.now()}`;
        saveToObsidian(filename, doc.data.content, {
          doc_token: docToken,
          title: filename
        });
      } else {
        console.log('返回:', JSON.stringify(doc, null, 2));
      }
    }
  } catch (err) {
    console.error('错误:', err.message);
    process.exit(1);
  }
}

if (require.main === module) {
  main();
}

#!/usr/bin/env node
/**
 * 本地文档处理工具
 * 读取手动导出的文档，AI 分析主题，整理到 Obsidian
 * 
 * 用法:
 * 1. 从飞书手动导出文档到 ~/Downloads/工作文档/
 * 2. 运行: node local-process.js
 */

const fs = require('fs');
const path = require('path');

const CONFIG = {
  inputDir: path.join(process.env.HOME, 'Downloads/工作文档'),
  outputDir: '/Users/maygo/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档',
  indexFile: 'INDEX.md'
};

// 确保目录存在
function ensureDir(dir) {
  if (!fs.existsSync(dir)) {
    fs.mkdirSync(dir, { recursive: true });
  }
}

// 扫描输入目录
function scanInputFiles() {
  ensureDir(CONFIG.inputDir);
  const files = fs.readdirSync(CONFIG.inputDir);
  return files.filter(f => 
    f.endsWith('.md') || 
    f.endsWith('.txt') || 
    f.endsWith('.docx')
  );
}

// 读取文件内容
function readFile(filename) {
  const filepath = path.join(CONFIG.inputDir, filename);
  return fs.readFileSync(filepath, 'utf8');
}

// 保存到 Obsidian
function saveToObsidian(category, filename, content, metadata = {}) {
  const categoryDir = path.join(CONFIG.outputDir, category);
  ensureDir(categoryDir);
  
  const frontmatter = `---
title: ${metadata.title || filename}
category: ${category}
source: 工作飞书
exported_at: ${new Date().toISOString()}
tags: [工作归档, ${category}]
---

`;
  
  const outputPath = path.join(categoryDir, filename);
  fs.writeFileSync(outputPath, frontmatter + content, 'utf8');
  return outputPath;
}

// 生成索引
function generateIndex(processed) {
  const grouped = {};
  for (const item of processed) {
    if (!grouped[item.category]) grouped[item.category] = [];
    grouped[item.category].push(item);
  }
  
  let index = `# 工作文档归档索引\n\n`;
  index += `**生成时间:** ${new Date().toLocaleString('zh-CN')}\n`;
  index += `**文档总数:** ${processed.length}\n\n`;
  
  for (const [category, items] of Object.entries(grouped)) {
    index += `## ${category} (${items.length})\n\n`;
    for (const item of items) {
      index += `- [[${item.filename}|${item.title}]]\n`;
    }
    index += `\n`;
  }
  
  fs.writeFileSync(
    path.join(CONFIG.outputDir, CONFIG.indexFile),
    index,
    'utf8'
  );
}

// 主函数
async function main() {
  console.log('扫描输入目录:', CONFIG.inputDir);
  const files = scanInputFiles();
  
  if (files.length === 0) {
    console.log('\n未找到文件。请先手动导出文档到:', CONFIG.inputDir);
    console.log('\n操作步骤:');
    console.log('1. 在飞书打开文档');
    console.log('2. 点击右上角 "..." → 导出 → Markdown');
    console.log('3. 保存到', CONFIG.inputDir);
    return;
  }
  
  console.log(`找到 ${files.length} 个文件\n`);
  
  const processed = [];
  
  for (const file of files) {
    console.log(`处理: ${file}`);
    const content = readFile(file);
    
    // 简单分类（后续可接入 AI）
    let category = '其他文档';
    if (file.includes('预算') || file.includes('降本')) category = '财务预算管理';
    else if (file.includes('纪要') || file.includes('会议')) category = '会议纪要';
    else if (file.includes('周工作报告')) category = '工作汇报';
    
    const outputPath = saveToObsidian(category, file, content, {
      title: file.replace(/\.(md|txt|docx)$/, '')
    });
    
    processed.push({
      filename: file,
      title: file.replace(/\.(md|txt|docx)$/, ''),
      category,
      path: outputPath
    });
    
    console.log(`  ✓ 已保存到: ${category}/${file}`);
  }
  
  generateIndex(processed);
  console.log(`\n✓ 完成！已处理 ${processed.length} 个文档`);
  console.log(`✓ 索引文件: ${path.join(CONFIG.outputDir, CONFIG.indexFile)}`);
}

main().catch(console.error);

{
  "name": "feishu-doc-exporter",
  "version": "1.0.0",
  "lockfileVersion": 3,
  "requires": true,
  "packages": {
    "": {
      "name": "feishu-doc-exporter",
      "version": "1.0.0",
      "dependencies": {
        "playwright": "^1.40.0"
      }
    },
    "node_modules/fsevents": {
      "version": "2.3.2",
      "resolved": "https://registry.npmjs.org/fsevents/-/fsevents-2.3.2.tgz",
      "integrity": "sha512-xiqMQR4xAeHTuB9uWm+fFRcIOgKBMiOBP+eXiyT7jsgVCq1bkVygt00oASowB7EdtpOHaaPgKt812P9ab+DDKA==",
      "hasInstallScript": true,
      "license": "MIT",
      "optional": true,
      "os": [
        "darwin"
      ],
      "engines": {
        "node": "^8.16.0 || ^10.6.0 || >=11.0.0"
      }
    },
    "node_modules/playwright": {
      "version": "1.58.2",
      "resolved": "https://registry.npmjs.org/playwright/-/playwright-1.58.2.tgz",
      "integrity": "sha512-vA30H8Nvkq/cPBnNw4Q8TWz1EJyqgpuinBcHET0YVJVFldr8JDNiU9LaWAE1KqSkRYazuaBhTpB5ZzShOezQ6A==",
      "license": "Apache-2.0",
      "dependencies": {
        "playwright-core": "1.58.2"
      },
      "bin": {
        "playwright": "cli.js"
      },
      "engines": {
        "node": ">=18"
      },
      "optionalDependencies": {
        "fsevents": "2.3.2"
      }
    },
    "node_modules/playwright-core": {
      "version": "1.58.2",
      "resolved": "https://registry.npmjs.org/playwright-core/-/playwright-core-1.58.2.tgz",
      "integrity": "sha512-yZkEtftgwS8CsfYo7nm0KE8jsvm6i/PTgVtB8DL726wNf6H2IMsDuxCpJj59KDaxCtSnrWan2AeDqM7JBaultg==",
      "license": "Apache-2.0",
      "bin": {
        "playwright-core": "cli.js"
      },
      "engines": {
        "node": ">=18"
      }
    }
  }
}

{
  "name": "feishu-doc-exporter",
  "version": "1.0.0",
  "type": "module",
  "dependencies": {
    "playwright": "^1.40.0"
  }
}

#!/usr/bin/env node
import { chromium } from 'playwright';
import fs from 'fs';
import path from 'path';
import { fileURLToPath } from 'url';

const __dirname = path.dirname(fileURLToPath(import.meta.url));

const CONFIG = {
  batchesFile: path.join(__dirname, 'batches.json'),
  progressFile: path.join(__dirname, 'progress.json'),
  outputDir: '/Users/maygo/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档',
  userDataDir: path.join(__dirname, 'browser-data'),
  reportInterval: 5 * 60 * 1000
};

const { batches } = JSON.parse(fs.readFileSync(CONFIG.batchesFile, 'utf8'));
const allDocs = batches.flat();

let progress = { completed: 0, failed: [], lastReport: Date.now(), results: [] };
if (fs.existsSync(CONFIG.progressFile)) {
  progress = JSON.parse(fs.readFileSync(CONFIG.progressFile, 'utf8'));
}

function saveProgress() {
  fs.writeFileSync(CONFIG.progressFile, JSON.stringify(progress, null, 2));
}

function ensureDir(dir) {
  if (!fs.existsSync(dir)) fs.mkdirSync(dir, { recursive: true });
}

function sanitize(name) {
  return name.replace(/[\/\\:*?"<>|]/g, '_').substring(0, 100);
}

async function extractDocContent(page) {
  await page.waitForTimeout(3000);
  const content = await page.evaluate(() => {
    const main = document.querySelector('[role="main"]') || document.body;
    return main.innerText || '';
  });
  return content;
}

async function processDoc(page, doc) {
  try {
    await page.goto(doc.url, { waitUntil: 'domcontentloaded', timeout: 60000 });
    const content = await extractDocContent(page);
    
    if (content.includes('扫码登录') || content.includes('账号登录')) {
      throw new Error('需要登录');
    }
    
    const categoryDir = path.join(CONFIG.outputDir, doc.category);
    ensureDir(categoryDir);
    
    const filename = `${sanitize(doc.name)}.md`;
    const frontmatter = `---
title: ${doc.name}
category: ${doc.category}
source: 工作飞书
url: ${doc.url}
exported_at: ${new Date().toISOString()}
---

`;
    
    fs.writeFileSync(path.join(categoryDir, filename), frontmatter + content, 'utf8');
    
    progress.completed++;
    progress.results.push({ id: doc.id, name: doc.name, status: 'success' });
    
    console.log(`✓ [${progress.completed}/${allDocs.length}] ${doc.name}`);
    
  } catch (err) {
    progress.failed.push({ id: doc.id, name: doc.name, error: err.message });
    console.error(`✗ ${doc.name}: ${err.message}`);
  } finally {
    saveProgress();
  }
}

async function main() {
  console.log('首次运行需要手动登录飞书...');
  console.log(`总文档数: ${allDocs.length}\n`);
  
  const browser = await chromium.launchPersistentContext(CONFIG.userDataDir, {
    headless: false,
    args: ['--disable-blink-features=AutomationControlled']
  });
  
  const page = await browser.newPage();
  
  console.log('请在浏览器中登录飞书，登录完成后按回车继续...');
  await page.goto('https://ottno-cd.feishu.cn/drive/home/');
  
  await new Promise(resolve => {
    process.stdin.once('data', resolve);
  });
  
  console.log('\n开始导出...\n');
  
  const startFrom = progress.completed;
  
  for (let i = startFrom; i < allDocs.length; i++) {
    await processDoc(page, allDocs[i]);
    
    if (Date.now() - progress.lastReport > CONFIG.reportInterval) {
      console.log(`\n进度: ${progress.completed}/${allDocs.length} (${Math.round(progress.completed/allDocs.length*100)}%)`);
      console.log(`失败: ${progress.failed.length}\n`);
      progress.lastReport = Date.now();
    }
  }
  
  await browser.close();
  
  console.log(`\n✓ 完成！成功: ${progress.completed - progress.failed.length} 失败: ${progress.failed.length}`);
}

main().catch(console.error);

#!/usr/bin/env node
import { chromium } from 'playwright';
import fs from 'fs';
import path from 'path';
import { fileURLToPath } from 'url';

const __dirname = path.dirname(fileURLToPath(import.meta.url));

const CONFIG = {
  batchesFile: path.join(__dirname, 'batches.json'),
  cookiesFile: path.join(__dirname, 'cookies.txt'),
  progressFile: path.join(__dirname, 'progress.json'),
  outputDir: '/Users/maygo/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档',
  reportInterval: 5 * 60 * 1000
};

const { batches } = JSON.parse(fs.readFileSync(CONFIG.batchesFile, 'utf8'));
const allDocs = batches.flat();

const cookieStr = fs.readFileSync(CONFIG.cookiesFile, 'utf8').trim();
const cookies = cookieStr.split('; ').map(c => {
  const [name, ...valueParts] = c.split('=');
  return { name: name.trim(), value: valueParts.join('=').trim(), domain: '.feishu.cn', path: '/' };
});

let progress = { completed: 0, failed: [], lastReport: Date.now(), results: [] };
if (fs.existsSync(CONFIG.progressFile)) {
  progress = JSON.parse(fs.readFileSync(CONFIG.progressFile, 'utf8'));
}

function saveProgress() {
  fs.writeFileSync(CONFIG.progressFile, JSON.stringify(progress, null, 2));
}

function ensureDir(dir) {
  if (!fs.existsSync(dir)) fs.mkdirSync(dir, { recursive: true });
}

function sanitize(name) {
  return name.replace(/[\/\\:*?"<>|]/g, '_').substring(0, 100);
}

async function extractDocContent(page) {
  await page.waitForTimeout(3000);
  const content = await page.evaluate(() => {
    const main = document.querySelector('[role="main"]') || document.body;
    return main.innerText || '';
  });
  return content;
}

async function processDoc(browser, doc) {
  const page = await browser.newPage();
  try {
    await page.goto(doc.url, { waitUntil: 'domcontentloaded', timeout: 60000 });
    const content = await extractDocContent(page);
    
    const categoryDir = path.join(CONFIG.outputDir, doc.category);
    ensureDir(categoryDir);
    
    const filename = `${sanitize(doc.name)}.md`;
    const frontmatter = `---
title: ${doc.name}
category: ${doc.category}
source: 工作飞书
url: ${doc.url}
exported_at: ${new Date().toISOString()}
---

`;
    
    fs.writeFileSync(path.join(categoryDir, filename), frontmatter + content, 'utf8');
    
    progress.completed++;
    progress.results.push({ id: doc.id, name: doc.name, status: 'success' });
    
    console.log(`✓ [${progress.completed}/${allDocs.length}] ${doc.name}`);
    
  } catch (err) {
    progress.failed.push({ id: doc.id, name: doc.name, error: err.message });
    console.error(`✗ ${doc.name}: ${err.message}`);
  } finally {
    await page.close();
    saveProgress();
  }
}

async function main() {
  console.log(`开始导出 ${allDocs.length} 个文档...`);
  
  const browser = await chromium.launch({ 
    headless: false,
    args: ['--disable-blink-features=AutomationControlled']
  });
  const context = await browser.newContext({
    userAgent: 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36'
  });
  await context.addCookies(cookies);
  
  const startFrom = progress.completed;
  
  for (let i = startFrom; i < allDocs.length; i++) {
    await processDoc(browser, allDocs[i]);
    
    if (Date.now() - progress.lastReport > CONFIG.reportInterval) {
      console.log(`\n进度: ${progress.completed}/${allDocs.length} (${Math.round(progress.completed/allDocs.length*100)}%)`);
      console.log(`失败: ${progress.failed.length}\n`);
      progress.lastReport = Date.now();
    }
  }
  
  await browser.close();
  
  console.log(`\n✓ 完成！成功: ${progress.completed - progress.failed.length} 失败: ${progress.failed.length}`);
}

main().catch(console.error);

#!/bin/bash
set -e

PROFILE="$HOME/.openclaw/workspace/skills/feishu-work-archive/profile"
OUTPUT_DIR="$HOME/Library/CloudStorage/OneDrive-个人/obsidian-vault/Projects/工作归档"
BATCHES="$HOME/.openclaw/workspace/skills/feishu-work-archive/batches.json"

# 读取前 10 个文档
docs=$(jq -r '[.batches[][]][0:10] | .[] | @json' "$BATCHES")

count=0
while IFS= read -r doc; do
  count=$((count + 1))
  
  name=$(echo "$doc" | jq -r '.name')
  url=$(echo "$doc" | jq -r '.url')
  category=$(echo "$doc" | jq -r '.category')
  
  echo "[$count/10] $name"
  
  # 访问文档
  agent-browser --profile "$PROFILE" open "$url" > /dev/null 2>&1
  sleep 2
  
  # 提取内容
  content=$(agent-browser eval "document.querySelector('[role=\"main\"]')?.innerText || document.body.innerText" 2>/dev/null || echo "")
  
  if [ -z "$content" ]; then
    echo "  ✗ 内容为空"
    continue
  fi
  
  # 保存
  category_dir="$OUTPUT_DIR/$category"
  mkdir -p "$category_dir"
  
  filename=$(echo "$name" | sed 's/[\/\\:*?"<>|]/_/g' | cut -c1-100).md
  filepath="$category_dir/$filename"
  
  cat > "$filepath" <<EOF
---
title: $name
category: $category
source: 工作飞书
url: $url
exported_at: $(date -u +"%Y-%m-%dT%H:%M:%SZ")
---

$content
EOF
  
  echo "  ✓ 已保存"
  
done <<< "$docs"

echo ""
echo "完成！处理了 $count 个文档"