import { chromium } from 'playwright'; import cron from 'node-cron'; import fs from 'node:fs'; import path from 'node:path'; import readline from 'node:readline'; import { execSync } from 'node:child_process'; import { config, datasets } from './config.js'; import { sendLoginAlert } from './notify.js'; import { diffRecords, loadCurrentState, nowStamp, saveCheckpoint, saveDatasetRun, saveDelta, saveRunSummary, withHash, } from './storage.js'; const sleep = (ms) => new Promise((resolve) => setTimeout(resolve, ms)); let _context = null; let _runtimeController = null; const AUTH_PAGE_KEYWORDS = [ 'RAM 用户登录', '主账号登录', '钉钉扫码登录', '用户名', '下一步', '登录并使用 RAM', ]; async function closeContextIfNeeded() { if (!_context) return; await _context.close(); _context = null; } function getRuntimeController() { if (_runtimeController) return _runtimeController; let paused = false; let terminated = false; let keypressBound = false; const onKeypress = (_str, key = {}) => { if (key.name === 'f7') { if (!paused) { paused = true; console.log('[控制] 已暂停(F7)。按 F8 继续,按 F9 终止。'); } return; } if (key.name === 'f8') { if (paused) { paused = false; console.log('[控制] 已继续(F8)。'); } return; } if (key.name === 'f9') { terminated = true; paused = false; console.log('[控制] 已请求终止(F9),将在安全检查点停止。'); } }; const bind = () => { if (keypressBound || !process.stdin.isTTY) return; readline.emitKeypressEvents(process.stdin); if (typeof process.stdin.setRawMode === 'function') { process.stdin.setRawMode(true); } process.stdin.resume(); process.stdin.on('keypress', onKeypress); keypressBound = true; console.log('[控制] 热键已启用:F7 暂停 / F8 继续 / F9 终止'); }; const unbind = () => { if (!keypressBound) return; process.stdin.off('keypress', onKeypress); if (process.stdin.isTTY && typeof process.stdin.setRawMode === 'function') { process.stdin.setRawMode(false); } keypressBound = false; }; const waitIfPaused = async (label = '任务') => { if (terminated) { throw new Error(`[控制] 已终止:${label}`); } while (paused) { await sleep(300); if (terminated) { throw new Error(`[控制] 已终止:${label}`); } } }; const throwIfTerminated = (label = '任务') => { if (terminated) { throw new Error(`[控制] 已终止:${label}`); } }; _runtimeController = { bind, unbind, waitIfPaused, throwIfTerminated, }; return _runtimeController; } async function runtimeCheckpoint(label) { const controller = getRuntimeController(); controller.throwIfTerminated(label); await controller.waitIfPaused(label); } async function getContext() { if (_context) return _context; _context = await chromium.launchPersistentContext(config.userDataDir, { channel: 'chrome', headless: config.headless, acceptDownloads: true, downloadsPath: config.downloadDir, }); await restoreStorageState(_context); return _context; } async function restoreStorageState(context) { if (!fs.existsSync(config.storageStateFile)) { return; } try { const state = JSON.parse(fs.readFileSync(config.storageStateFile, 'utf-8')); if (Array.isArray(state.cookies) && state.cookies.length > 0) { await context.addCookies(state.cookies); console.log(`[storageState] 已恢复 ${state.cookies.length} 个 cookie`); } } catch (error) { console.warn(`[storageState] 恢复失败,继续使用 .browser profile: ${error.message}`); } } async function saveStorageState(context) { await context.storageState({ path: config.storageStateFile }); console.log(`[storageState] 已保存登录态快照: ${config.storageStateFile}`); } function loadLatestBillsCheckpoint() { const checkpointDir = path.join(config.dataDir, 'checkpoints', 'bills'); if (!fs.existsSync(checkpointDir)) { return null; } const candidates = fs.readdirSync(checkpointDir) .filter((fileName) => fileName.endsWith('.json')) .map((fileName) => { const filePath = path.join(checkpointDir, fileName); const stat = fs.statSync(filePath); return { fileName, filePath, mtimeMs: stat.mtimeMs }; }) .sort((a, b) => b.mtimeMs - a.mtimeMs); if (candidates.length === 0) { return null; } try { const latest = JSON.parse(fs.readFileSync(candidates[0].filePath, 'utf-8')); if (!latest || typeof latest !== 'object') { return null; } return latest; } catch (error) { console.warn(`[账单检查点] 读取失败,忽略断点续爬: ${error.message}`); return null; } } async function getPageBodyPreview(page) { return page .evaluate(() => document.body?.innerText?.substring(0, 500) || '(空)') .catch(() => '(无法获取)'); } function isAuthUrl(url) { return /account\.aliyun\.com|signin\.aliyun\.com/.test(url) || url.includes('login.htm') || url.includes('/#/signin'); } function hasAuthKeywords(text) { return AUTH_PAGE_KEYWORDS.some((keyword) => text.includes(keyword)); } async function detectAuthRedirect(page) { const currentUrl = page.url(); const bodyText = await getPageBodyPreview(page); return { currentUrl, bodyText, isAuthPage: isAuthUrl(currentUrl) || hasAuthKeywords(bodyText), }; } async function ensureDatasetAccessible(page, dataset, timeout = 120000, options = {}) { await page.goto(dataset.url, { waitUntil: 'domcontentloaded' }); await waitUntilReady(page, dataset.heading, timeout, options); } export async function login() { const runtimeController = getRuntimeController(); runtimeController.bind(); const context = await getContext(); const cleanupAndExit = async (signal) => { console.log(`[login] 收到 ${signal},正在保存登录态并关闭浏览器...`); await closeContextIfNeeded(); process.exit(130); }; const onSigint = () => { void cleanupAndExit('SIGINT'); }; const onSigterm = () => { void cleanupAndExit('SIGTERM'); }; process.once('SIGINT', onSigint); process.once('SIGTERM', onSigterm); try { const page = context.pages()[0] || (await context.newPage()); await page.goto(datasets.customers.url, { waitUntil: 'domcontentloaded' }); console.log('请在打开的浏览器里完成阿里云伙伴中心登录。检测到进入“我的客户”和“账单查询”页面后,脚本会自动保存登录态并关闭浏览器。'); await waitUntilReady(page, datasets.customers.heading, 10 * 60 * 1000, { allowInteractiveAuth: true }); console.log('[login] 我的客户页验证通过,继续验证账单页登录态...'); await ensureDatasetAccessible(page, datasets.bills, 60 * 1000, { allowInteractiveAuth: true }); await sleep(1000); await saveStorageState(context); console.log('登录态已写入 .browser 目录,且已验证“我的客户”和“账单查询”页面可访问,后续可直接执行 npm run sync 或 npm run bills。'); } finally { process.off('SIGINT', onSigint); process.off('SIGTERM', onSigterm); await closeContextIfNeeded(); runtimeController.unbind(); } } export async function syncAll() { const runtimeController = getRuntimeController(); runtimeController.bind(); const context = await getContext(); try { const summary = { startedAt: new Date().toISOString(), datasets: {} }; const page = context.pages()[0] || (await context.newPage()); summary.datasets.customers = await syncCustomers(page); summary.datasets.customerDetails = await syncCustomerDetails(page); summary.datasets.orders = await syncOrders(page); // syncOrders 完成后,从最新的 orders.json 读取 orderId 列表 const latestOrders = loadCurrentState('orders', datasets.orders.uniqueKey); const orderIdsForDetail = collectValidOrderIds(latestOrders.records || []); summary.datasets.orderDetails = await syncOrderDetails(page, orderIdsForDetail); summary.datasets.bills = await syncBills(page); summary.finishedAt = new Date().toISOString(); const stamp = nowStamp(); saveRunSummary(stamp, summary); return summary; } finally { if (config.closeBrowser) { await closeContextIfNeeded(); } else { console.log('浏览器保持运行'); } runtimeController.unbind(); } } export async function syncBillsOnly(options = {}) { const runtimeController = getRuntimeController(); runtimeController.bind(); const context = await getContext(); try { const summary = { startedAt: new Date().toISOString(), datasets: {} }; const page = context.pages()[0] || (await context.newPage()); summary.datasets.bills = await syncBills(page, options); summary.finishedAt = new Date().toISOString(); const stamp = nowStamp(); saveRunSummary(stamp, summary); return summary; } finally { if (config.closeBrowser) { await closeContextIfNeeded(); } else { console.log('浏览器保持运行'); } runtimeController.unbind(); } } export async function scheduleSync() { console.log(`定时任务已启动: ${config.cron} (${config.timezone})`); cron.schedule( config.cron, async () => { try { console.log(`[${new Date().toISOString()}] 开始执行同步`); const summary = await syncAll(); console.log(`[${new Date().toISOString()}] 同步完成`, JSON.stringify(summary, null, 2)); try { const scriptPath = path.resolve(config.rootDir, config.dbSyncScript); const incrementalFlag = config.fullSync ? '' : ' --incremental'; console.log(`[入库] 执行 ${scriptPath}${incrementalFlag ? ' (增量模式)' : ''}`); const output = execSync(`python "${scriptPath}"${incrementalFlag}`, { cwd: path.dirname(scriptPath), encoding: 'utf-8', timeout: 120000, }); console.log(output); } catch (e) { console.error('[入库] 失败:', e.message); } } catch (error) { console.error(`[${new Date().toISOString()}] 同步失败`, error); } }, { timezone: config.timezone }, ); } async function syncCustomers(page) { await runtimeCheckpoint('同步客户'); const dataset = datasets.customers; await page.goto(dataset.url, { waitUntil: 'domcontentloaded' }); await waitUntilReady(page, dataset.heading); await trySetPageSize(page, dataset.pageSize); const records = await scrapePagedTable(page, dataset, {}); return persistDataset(dataset, records, {}); } async function syncCustomerDetails(page) { await runtimeCheckpoint('同步客户详情'); const dataset = datasets.customerDetails; const customersState = loadCurrentState('customers', datasets.customers.uniqueKey); const allAccountIds = collectValidAccountIds(customersState.records || []); if (allAccountIds.length === 0) { console.log('[客户详情] 本地无有效客户 accountId,跳过'); return persistDataset(dataset, [], {}); } console.log(`[客户详情] 共 ${allAccountIds.length} 个客户需要获取详情`); const allDetails = []; const detailBaseUrl = 'https://aps.aliyun.com/?spm=5176.12818093.top-nav.ditem-fx.785716d0LKDpKT#/detail/my_customer/~/customer/'; for (let index = 0; index < allAccountIds.length; index += 1) { await runtimeCheckpoint(`客户详情 ${index + 1}/${allAccountIds.length}`); const accountId = allAccountIds[index]; console.log(`[客户详情] ${index + 1}/${allAccountIds.length} accountId=${accountId}`); // 先跳 about:blank 再跳详情URL(强制 SPA 完整重新加载) await page.goto('about:blank'); await sleep(300); await page.goto(`${detailBaseUrl}${accountId}`, { waitUntil: 'domcontentloaded' }); try { await page.waitForFunction( (text) => document.body && document.body.innerText.includes(text), '详情', { timeout: 15000 }, ); await sleep(1000); } catch { console.warn(`[客户详情] ${accountId} 详情页加载超时,跳过`); continue; } const detail = await extractCustomerDetail(page); allDetails.push({ ...detail, __context: { accountId } }); } return persistDataset(dataset, dedupeByHash(allDetails), {}); } async function syncOrders(page) { await runtimeCheckpoint('同步订单'); const dataset = datasets.orders; let windows; if (config.fullSync) { windows = buildMonthlyDateWindows(config.orderStartDate); } else { windows = buildIncrementalOrderWindows(); } const allRecords = []; for (const window of windows) { await runtimeCheckpoint(`订单窗口 ${window.start} ~ ${window.end}`); await page.goto(dataset.url, { waitUntil: 'domcontentloaded' }); await waitUntilReady(page, dataset.heading); await setDateRange(page, window.start, window.end); await clickQuery(page); await trySetPageSize(page, dataset.pageSize); const records = await scrapePagedTable(page, dataset, window); allRecords.push(...records); } return persistDataset(dataset, dedupeByHash(allRecords), {}); } function buildIncrementalOrderWindows() { const configuredStartDate = normalizeConfiguredDate(config.incrementalOrderStartDate); if (configuredStartDate) { const windows = buildMonthlyDateWindows(configuredStartDate); console.log(`[增量模式] 订单从指定日期开始查询: ${configuredStartDate}`); return windows; } const yesterday = new Date(); yesterday.setDate(yesterday.getDate() - 1); const dateStr = formatDate(yesterday); console.log(`[增量模式] 订单仅查询: ${dateStr}`); return [{ windowStart: dateStr, windowEnd: dateStr, start: dateStr, end: dateStr }]; } function normalizeConfiguredDate(value) { const normalized = String(value || '').trim(); if (!normalized) { return ''; } if (!/^\d{4}-\d{2}-\d{2}$/.test(normalized)) { throw new Error(`ALIYUN_APS_INCREMENTAL_ORDER_START_DATE 格式无效: ${normalized},期望 YYYY-MM-DD`); } const parsed = new Date(`${normalized}T00:00:00+08:00`); if (Number.isNaN(parsed.getTime())) { throw new Error(`ALIYUN_APS_INCREMENTAL_ORDER_START_DATE 不是有效日期: ${normalized}`); } return normalized; } async function syncBills(page, options = {}) { await runtimeCheckpoint('同步账单'); const dataset = datasets.bills; const { resume = false } = options; let months; let latestConsumptionDate = null; if (config.fullSync) { months = buildMonthList(config.billStartMonth); } else { latestConsumptionDate = getLatestBillConsumptionDate(); const incrementalMonth = latestConsumptionDate?.slice(0, 7) || `${new Date().getFullYear()}-${String(new Date().getMonth() + 1).padStart(2, '0')}`; months = [incrementalMonth]; console.log(`[增量模式] 账单仅查询: ${incrementalMonth}${latestConsumptionDate ? `, 数据库最新消费时间: ${latestConsumptionDate}` : ''}`); } const resumeCheckpoint = resume ? loadLatestBillsCheckpoint() : null; if (resumeCheckpoint?.month) { const resumeIndex = months.indexOf(resumeCheckpoint.month); if (resumeIndex >= 0) { months = months.slice(resumeIndex); console.log(`[账单续爬] 从 checkpoint 恢复: month=${resumeCheckpoint.month}, page=${resumeCheckpoint.pageNum || 1}, records=${(resumeCheckpoint.records || []).length}`); } } const allNormalizedRecords = []; for (const month of months) { await runtimeCheckpoint(`账单月份 ${month}`); await page.goto(dataset.url, { waitUntil: 'domcontentloaded' }); await waitUntilReady(page, dataset.heading); await setMonthValue(page, month); await clickQuery(page); await trySetPageSize(page, dataset.pageSize); let monthNormalizedRecords = []; let resumeFromPage = 0; let shouldContinueScrape = true; if (resumeCheckpoint?.month === month) { monthNormalizedRecords = Array.isArray(resumeCheckpoint.records) ? resumeCheckpoint.records : []; resumeFromPage = Number.parseInt(String(resumeCheckpoint.pageNum || 0), 10) || 0; if (resumeFromPage > 0) { const moved = await moveBillsToResumeStart(page, resumeFromPage); if (!moved) { console.log(`[账单续爬] checkpoint 已在最后一页,无需继续抓取 month=${month}`); shouldContinueScrape = false; } } } let rawRecords = []; if (shouldContinueScrape) { rawRecords = await scrapePagedTable(page, dataset, { month }, { onPage: async ({ pageNum, pageRows }) => { const normalizedPageRows = normalizeDatasetRecords(dataset, pageRows, { month }); monthNormalizedRecords.push(...normalizedPageRows); let checkpointRecords = monthNormalizedRecords; if (latestConsumptionDate) { checkpointRecords = monthNormalizedRecords.filter((record) => isAfterLatestConsumptionDate(record, latestConsumptionDate)); } await saveBillsCheckpoint(dataset, month, pageNum, checkpointRecords); }, }); } if (resumeFromPage === 0) { monthNormalizedRecords = normalizeDatasetRecords(dataset, rawRecords, { month }); } if (latestConsumptionDate) { const before = monthNormalizedRecords.length; monthNormalizedRecords = monthNormalizedRecords.filter((record) => isAfterLatestConsumptionDate(record, latestConsumptionDate)); console.log(`[增量模式] 账单按消费时间过滤: ${before} -> ${monthNormalizedRecords.length}`); } allNormalizedRecords.push(...monthNormalizedRecords); } return persistNormalizedDataset(dataset, dedupeByHash(allNormalizedRecords)); } async function saveBillsCheckpoint(dataset, month, pageNum, normalizedRecords) { const normalized = dedupeByHash(normalizedRecords); const checkpointName = `${month}-latest`; saveCheckpoint(dataset.name, checkpointName, { month, pageNum, savedAt: new Date().toISOString(), stats: { total: normalized.length, }, records: normalized, }); console.log(`[账单检查点] 已落盘: month=${month}, page=${pageNum}, records=${normalized.length}`); } function normalizeDatasetRecords(dataset, records, context) { return records.map((record) => withHash(dataset.normalize(record, record.__context || context))); } async function moveBillsToResumeStart(page, resumeFromPage) { if (resumeFromPage <= 0) { return true; } const reached = await jumpToPage(page, resumeFromPage); if (!reached) { throw new Error(`账单续爬失败:无法定位到 checkpoint 页码 ${resumeFromPage}`); } const moved = await gotoNextPage(page); return moved; } function getLatestBillConsumptionDate() { const scriptPath = path.resolve(config.rootDir, config.dbSyncScript); try { const output = execSync(`python "${scriptPath}" --latest-bill-consumption-time`, { cwd: path.dirname(scriptPath), encoding: 'utf-8', timeout: 120000, }).trim(); const latest = output.split(/\r?\n/).map((line) => line.trim()).filter(Boolean).at(-1) || ''; return /^\d{4}-\d{2}-\d{2}/.test(latest) ? latest.slice(0, 10) : null; } catch (error) { console.error('[增量模式] 查询数据库最新账单消费时间失败:', error.message); return null; } } function isAfterLatestConsumptionDate(record, latestConsumptionDate) { const consumeDate = String(record['消费时间'] || record.consumeDate || '').trim().slice(0, 10); if (!/^\d{4}-\d{2}-\d{2}$/.test(consumeDate)) { return false; } return consumeDate > latestConsumptionDate; } async function syncOrderDetails(page, cachedOrderIds) { await runtimeCheckpoint('同步订单详情'); const dataset = datasets.orderDetails; // 使用传入的 orderId 列表(在 syncOrders 覆盖 orders.json 之前缓存的) const allOrderIds = cachedOrderIds || []; if (allOrderIds.length === 0) { console.log('[订单详情] 本地无订单数据,跳过'); return persistDataset(dataset, [], {}); } console.log(`[订单详情] 共 ${allOrderIds.length} 个订单需要获取详情`); const allDetails = []; const detailBaseUrl = 'https://aps.aliyun.com/?spm=5176.12818093.top-nav.ditem-fx.785716d0LKDpKT#/detail/order/~/costCenter/order/detail/'; for (let index = 0; index < allOrderIds.length; index += 1) { await runtimeCheckpoint(`订单详情 ${index + 1}/${allOrderIds.length}`); const orderId = allOrderIds[index]; console.log(`[订单详情] ${index + 1}/${allOrderIds.length} orderId=${orderId}`); // 先跳 about:blank 再跳详情URL(强制 SPA 完整重新加载) await page.goto('about:blank'); await sleep(300); await page.goto(`${detailBaseUrl}${orderId}?projectId=`, { waitUntil: 'domcontentloaded' }); try { await page.waitForFunction( (text) => document.body && document.body.innerText.includes(text), '订单详情', { timeout: 15000 }, ); await sleep(1000); } catch { console.warn(`[订单详情] ${orderId} 详情页加载超时,跳过`); continue; } const detail = await extractOrderDetail(page); if (!isValidOrderId(detail.orderId)) { detail.orderId = orderId; } allDetails.push({ ...detail, __context: {} }); } return persistDataset(dataset, dedupeByHash(allDetails), {}); } function persistDataset(dataset, records, context) { const normalized = normalizeDatasetRecords(dataset, records, context); return persistNormalizedDataset(dataset, normalized); } function persistNormalizedDataset(dataset, normalizedRecords) { const previousState = loadCurrentState(dataset.name, dataset.uniqueKey); const nextState = diffRecords(previousState, normalizedRecords, dataset.uniqueKey); const stamp = saveDatasetRun(dataset.name, nextState); saveDelta(dataset.name, stamp, nextState.delta); return { stamp, stats: nextState.stats, }; } async function waitUntilReady(page, heading, timeout = 120000, options = {}) { await runtimeCheckpoint(`等待页面 ${heading}`); const { allowInteractiveAuth = false } = options; await page.waitForLoadState('domcontentloaded'); console.log(`[waitUntilReady] 当前URL: ${page.url()}`); console.log(`[waitUntilReady] 等待页面出现: "${heading}"`); const initialState = await detectAuthRedirect(page); if (initialState.isAuthPage) { console.error(`[waitUntilReady] 检测到登录页/鉴权页: ${initialState.currentUrl}`); console.error(`[waitUntilReady] 页面内容前500字: ${initialState.bodyText}`); if (!allowInteractiveAuth && isAuthUrl(initialState.currentUrl)) { try { await sendLoginAlert(initialState.currentUrl); } catch (notifyErr) { console.error('[通知] 发送登录提醒失败:', notifyErr.message); } } if (!allowInteractiveAuth) { throw new Error(`当前页面仍处于登录/鉴权页,无法进入「${heading}」。请重新执行 npm run login,并确认该账号对该页面有访问权限。`); } console.log(`[waitUntilReady] 允许交互式登录,等待用户完成认证后进入「${heading}」...`); } try { await page.waitForFunction( (text) => document.body && document.body.innerText.includes(text), heading, { timeout }, ); } catch (err) { // 超时时打印诊断信息 const { currentUrl, bodyText, isAuthPage } = await detectAuthRedirect(page); console.error(`[waitUntilReady] 超时!当前URL: ${currentUrl}`); console.error(`[waitUntilReady] 页面内容前500字: ${bodyText}`); if (isAuthPage && !allowInteractiveAuth) { try { await sendLoginAlert(currentUrl); } catch (notifyErr) { console.error('[通知] 发送登录提醒失败:', notifyErr.message); } throw new Error(`当前页面停留在登录/鉴权页,未能进入「${heading}」。请重新执行 npm run login,并确认该账号对该页面有访问权限。`); } if (isAuthPage && allowInteractiveAuth) { throw new Error(`交互式登录超时,仍未进入「${heading}」。请确认已在浏览器中完成 RAM/阿里云登录,并且当前账号有访问该页面的权限。`); } throw err; } const finalState = await detectAuthRedirect(page); if (finalState.isAuthPage && !allowInteractiveAuth) { throw new Error(`当前页面仍处于登录/鉴权页,未成功进入「${heading}」。请重新执行 npm run login,并确认该账号对该页面有访问权限。`); } await sleep(1500); } async function scrapePagedTable(page, dataset, context, options = {}) { const { onPage, skipInitialPage = false } = options; const pages = []; const visited = new Set(); let shouldSkipCurrentPage = skipInitialPage; while (true) { await runtimeCheckpoint(`抓取 ${dataset.name} 分页`); await waitForTableRows(page); const pageData = await extractTable(page); const pageNum = await currentPageNumber(page); const pageKey = `${pageNum}-${pageData.rows.length}`; console.log(`[抓取] 第${pageNum}页, ${pageData.rows.length}行, key="${pageKey}"`); if (shouldSkipCurrentPage) { console.log(`[抓取] 跳过 checkpoint 已保存页: ${pageNum}`); shouldSkipCurrentPage = false; const moved = await gotoNextPage(page); if (!moved) { console.log(`[抓取] checkpoint 已位于最后一页,停止`); break; } continue; } if (visited.has(pageKey)) { console.log(`[抓取] 重复页面key,停止翻页`); break; } visited.add(pageKey); const pageRows = pageData.rows.map((row) => ({ ...row, __context: context })); pages.push(...pageRows); if (onPage) { await onPage({ pageData, pageNum, pageRows }); } const moved = await gotoNextPage(page); if (!moved) { console.log(`[抓取] 翻页失败或已到最后一页,停止`); break; } } console.log(`[抓取] 共采集 ${pages.length} 条记录`); return pages; } async function raiseIfSessionExpired(page, label) { const { currentUrl, bodyText, isAuthPage } = await detectAuthRedirect(page); if (!isAuthPage) { return; } console.error(`[鉴权] ${label} 时检测到登录页/鉴权页: ${currentUrl}`); console.error(`[鉴权] 页面内容前500字: ${bodyText}`); try { await sendLoginAlert(currentUrl); } catch (notifyErr) { console.error('[通知] 发送登录提醒失败:', notifyErr.message); } throw new Error(`运行过程中登录态失效(${label})。请重新执行 npm run login 后再继续同步。`); } async function extractTable(page) { return page.evaluate(() => { const normalize = (value) => String(value || '') .replace(/\u00a0/g, ' ') .replace(/\s+\n/g, '\n') .replace(/\n\s+/g, '\n') .trim(); const headerTables = Array.from(document.querySelectorAll('table')).filter((table) => table.querySelectorAll('thead th').length > 1); const headerTable = headerTables.sort((a, b) => b.querySelectorAll('thead th').length - a.querySelectorAll('thead th').length)[0]; if (!headerTable) return { headers: [], rows: [] }; const headers = Array.from(headerTable.querySelectorAll('thead th')).map((cell) => normalize(cell.textContent)); const bodyTables = Array.from(document.querySelectorAll('table')).filter((table) => table.querySelectorAll('tbody tr').length > 0); const bodyTable = bodyTables.sort((a, b) => { const aSize = Math.max(...Array.from(a.querySelectorAll('tbody tr')).map((row) => row.querySelectorAll('td').length), 0); const bSize = Math.max(...Array.from(b.querySelectorAll('tbody tr')).map((row) => row.querySelectorAll('td').length), 0); return bSize - aSize; })[0]; if (!bodyTable) return { headers, rows: [] }; const rows = Array.from(bodyTable.querySelectorAll('tbody tr')) .map((row) => Array.from(row.querySelectorAll('td')).map((cell) => normalize(cell.innerText || cell.textContent))) .filter((cells) => cells.some(Boolean)) .map((cells) => { const record = {}; headers.forEach((header, index) => { record[header || `column_${index + 1}`] = cells[index] || ''; }); return record; }); return { headers, rows }; }); } async function waitForTableRows(page) { await runtimeCheckpoint('等待表格加载'); try { await page.waitForFunction(() => document.querySelectorAll('table tbody tr').length > 0, null, { timeout: 120000 }); } catch (error) { await raiseIfSessionExpired(page, '等待表格加载'); throw error; } await sleep(800); } async function currentPageNumber(page) { const active = page.locator('.next-pagination-item.next-current'); if ((await active.count()) === 0) return 1; return Number.parseInt((await active.first().innerText()).trim(), 10) || 1; } async function jumpToPage(page, targetPage) { if (targetPage <= 1) { return true; } const current = await currentPageNumber(page); if (current === targetPage) { return true; } const jumpInputCandidates = [ '.next-pagination-jump-input input', 'input[aria-label*="页码"]', 'input[aria-label*="页"]', ]; for (const selector of jumpInputCandidates) { const input = page.locator(selector).first(); if ((await input.count()) === 0) { continue; } await input.click().catch(() => null); await sleep(100); await page.keyboard.press('Control+A').catch(() => null); await page.keyboard.type(String(targetPage), { delay: 20 }).catch(() => null); await page.keyboard.press('Enter').catch(() => null); await sleep(1500); const afterJump = await currentPageNumber(page); if (afterJump === targetPage) { console.log(`[账单续爬] 已跳转到第 ${targetPage} 页`); return true; } } console.warn(`[账单续爬] 未找到可用跳页输入框,尝试顺序翻到第 ${targetPage} 页`); let guard = 0; while (guard < targetPage + 5) { const currentPage = await currentPageNumber(page); if (currentPage >= targetPage) { return currentPage === targetPage; } const moved = await gotoNextPage(page); if (!moved) { return false; } guard += 1; } return false; } async function gotoNextPage(page) { await runtimeCheckpoint('翻页'); const before = await currentPageNumber(page); // 用 Playwright locator 定位"下一页"按钮 const nextBtn = page.locator('button.next-pagination-item.next-next'); if ((await nextBtn.count()) === 0) { console.log('[翻页] 未找到下一页按钮'); return false; } const disabled = (await nextBtn.getAttribute('disabled')) != null; if (disabled) { console.log('[翻页] 下一页按钮已禁用'); return false; } // 用 Playwright click(而非 DOM click),确保 React 事件正常触发 await nextBtn.click(); await sleep(2000); await raiseIfSessionExpired(page, `翻页 ${before} -> next`); const after = await currentPageNumber(page); console.log(`[翻页] ${before} -> ${after}`); if (before > 1 && after === 1) { throw new Error(`分页从第 ${before} 页异常回退到第 1 页,疑似登录态失效或页面会话已重置。请重新执行 npm run login 后再继续同步。`); } if (after < before) { throw new Error(`分页从第 ${before} 页异常回退到第 ${after} 页,疑似登录态失效或页面状态被重置。请重新执行 npm run login 后再继续同步。`); } return before !== after; } async function trySetPageSize(page, pageSize) { await runtimeCheckpoint(`设置每页 ${pageSize}`); const input = page.locator('input[aria-label="请选择每页显示几条"]').first(); if ((await input.count()) === 0) return; await input.click().catch(() => null); await sleep(300); const option = page.locator(`text=${pageSize}`).last(); if ((await option.count()) === 0) { await page.keyboard.press('Escape').catch(() => null); return; } await option.click().catch(() => null); await sleep(1200); } async function setDateRange(page, start, end) { await runtimeCheckpoint(`设置订单日期 ${start} ~ ${end}`); console.log(`[订单日期] 设置: ${start} ~ ${end}`); await _fillDateRange(page, start, end); // 验证 const startActual = await page.locator('input[placeholder="起始日期"]').inputValue().catch(() => ''); const endActual = await page.locator('input[placeholder="结束日期"]').inputValue().catch(() => ''); // 如果结果不对,用反向顺序重试(先填开始再填结束) if (startActual !== start || endActual !== end) { console.log(`[订单日期] 首次结果不对: "${startActual}" ~ "${endActual}",反向重试`); await _fillDateRange(page, start, end, true); const s2 = await page.locator('input[placeholder="起始日期"]').inputValue().catch(() => ''); const e2 = await page.locator('input[placeholder="结束日期"]').inputValue().catch(() => ''); console.log(`[订单日期] 重试结果: "${s2}" ~ "${e2}"`); } else { console.log(`[订单日期] 结果: "${startActual}" ~ "${endActual}"`); } } async function _fillDateRange(page, start, end, startFirst = false) { await runtimeCheckpoint('填写订单日期'); const trigger = page.locator('input[placeholder="结束日期"]'); await trigger.click(); await sleep(1000); const panelStartInput = page.locator('.next-range-picker-panel-input-start-date input'); const panelEndInput = page.locator('.next-range-picker-panel-input-end-date input'); if (startFirst) { // 先填开始日期 await panelStartInput.click(); await sleep(100); await page.keyboard.press('Control+A'); await page.keyboard.type(start, { delay: 30 }); await sleep(300); // 再填结束日期 await panelEndInput.click(); await sleep(100); await page.keyboard.press('Control+A'); await page.keyboard.type(end, { delay: 30 }); await sleep(300); } else { // 先填结束日期(默认) await panelEndInput.click(); await sleep(100); await page.keyboard.press('Control+A'); await page.keyboard.type(end, { delay: 30 }); await sleep(300); // 再填开始日期 await panelStartInput.click(); await sleep(100); await page.keyboard.press('Control+A'); await page.keyboard.type(start, { delay: 30 }); await sleep(300); } await page.keyboard.press('Enter'); await sleep(500); await page.mouse.click(0, 0); await sleep(300); await page.keyboard.press('Escape'); await sleep(300); await page.locator('.next-overlay-wrapper.opened').waitFor({ state: 'hidden', timeout: 3000 }).catch(() => null); await sleep(300); } async function setMonthValue(page, month) { await runtimeCheckpoint(`设置账单月份 ${month}`); // 先尝试按 inputValue 匹配 YYYY-MM 格式 const inputs = page.locator('input'); const total = await inputs.count(); const allValues = []; for (let index = 0; index < total; index += 1) { const input = inputs.nth(index); const value = await input.inputValue().catch(() => ''); const placeholder = await input.getAttribute('placeholder').catch(() => ''); allValues.push({ index, value, placeholder }); if (/^\d{4}-\d{2}$/.test(value)) { console.log(`[账单月份] 通过 value 匹配到 input[${index}], 设置: ${month}`); await typeIntoDateInput(input, month, page); return; } } // 如果 value 为空,尝试按 placeholder 匹配月份选择器 for (const item of allValues) { if (item.placeholder && /月/.test(item.placeholder)) { console.log(`[账单月份] 通过 placeholder 匹配到 input[${item.index}], 设置: ${month}`); await typeIntoDateInput(inputs.nth(item.index), month, page); return; } } // 兜底:找任何看起来像日期/月份选择器的 input(排除搜索框等) for (const item of allValues) { const input = inputs.nth(item.index); const cls = await input.evaluate((el) => el.closest('[class*="date-picker"], [class*="month-picker"], [class*="range-picker"]')?.className || '').catch(() => ''); if (cls) { console.log(`[账单月份] 通过父级 class 匹配到 input[${item.index}] (${cls}), 设置: ${month}`); await typeIntoDateInput(input, month, page); return; } } console.error('[DEBUG] 账单页面所有 input:', JSON.stringify(allValues, null, 2)); throw new Error('未识别到账单佣金月份输入框,请打开页面确认结构是否变化。'); } /** * 用键盘输入日期值。 * 策略:focus → 全选 → 快速键入 → Tab 移开焦点(触发 blur 提交,但不会像 click 那样打开面板)。 * 即使面板弹出,快速键入 + Tab 也能在面板滚动前完成提交并关闭。 */ async function typeIntoDateInput(locator, value, page) { await runtimeCheckpoint(`填写日期输入 ${value}`); // 移除 readonly await locator.evaluate((node) => node.removeAttribute('readonly')); // focus 并全选当前内容 await locator.focus(); await sleep(100); await page.keyboard.press('Control+A'); await sleep(100); // 快速逐字符输入新值 await page.keyboard.type(value, { delay: 30 }); await sleep(200); // Tab 移开焦点 → 触发 onBlur 提交值 + 关闭面板 await page.keyboard.press('Tab'); await sleep(300); // 如果面板还在,Escape 兜底关闭 await page.keyboard.press('Escape'); await sleep(300); // 验证 const actual = await locator.inputValue().catch(() => ''); if (actual !== value) { console.warn(`[WARN] typeIntoDateInput: 期望 "${value}",实际 "${actual}"`); } else { console.log(`[日期设置] 成功: "${value}"`); } } async function clickQuery(page) { await runtimeCheckpoint('点击查询'); const button = page.locator('button:has-text("查询")').first(); await button.click(); await sleep(1800); } function buildMonthlyDateWindows(startDate) { const start = new Date(`${startDate}T00:00:00+08:00`); const end = new Date(); const windows = []; const cursor = new Date(start.getFullYear(), start.getMonth(), 1); while (cursor <= end) { const windowStart = new Date(cursor); const windowEnd = new Date(cursor.getFullYear(), cursor.getMonth() + 1, 0); const actualEnd = windowEnd > end ? end : windowEnd; windows.push({ windowStart: formatDate(windowStart), windowEnd: formatDate(actualEnd), start: formatDate(windowStart), end: formatDate(actualEnd), }); cursor.setMonth(cursor.getMonth() + 1); } return windows; } function buildMonthList(startMonth) { const [year, month] = startMonth.split('-').map(Number); const cursor = new Date(year, month - 1, 1); const end = new Date(); const months = []; while (cursor <= end) { months.push(`${cursor.getFullYear()}-${String(cursor.getMonth() + 1).padStart(2, '0')}`); cursor.setMonth(cursor.getMonth() + 1); } return months; } function formatDate(date) { return `${date.getFullYear()}-${String(date.getMonth() + 1).padStart(2, '0')}-${String(date.getDate()).padStart(2, '0')}`; } function dedupeByHash(records) { const seen = new Set(); return records.filter((record) => { const key = JSON.stringify(record); if (seen.has(key)) return false; seen.add(key); return true; }); } function collectValidOrderIds(records) { const ids = []; const seen = new Set(); for (const record of records) { // 支持两种字段名:normalized 后的 orderId 和原始的 订单号 const rawOrderId = String(record.orderId || record['订单号'] || '').trim(); if (!rawOrderId || rawOrderId.includes('没有数据')) { continue; } if (!isValidOrderId(rawOrderId)) { console.log(`[订单详情] 跳过无效订单号: ${rawOrderId}`); continue; } if (seen.has(rawOrderId)) { continue; } seen.add(rawOrderId); ids.push(rawOrderId); } return ids; } function collectValidAccountIds(records) { const ids = []; const seen = new Set(); for (const record of records) { const rawAccountId = String(record.accountId || '').trim(); if (!rawAccountId || rawAccountId.includes('没有数据')) { continue; } if (!isValidAccountId(rawAccountId)) { console.log(`[客户详情] 跳过无效 accountId: ${rawAccountId}`); continue; } if (seen.has(rawAccountId)) { continue; } seen.add(rawAccountId); ids.push(rawAccountId); } return ids; } function isValidOrderId(orderId) { const value = String(orderId || '').trim(); if (!value) return false; if (value.includes('�')) return false; return /^\d+$/.test(value); } function isValidAccountId(accountId) { const value = String(accountId || '').trim(); if (!value) return false; if (value.includes('�')) return false; return /^\d+$/.test(value); } async function extractOrderDetail(page) { return page.evaluate(() => { const text = document.body?.innerText || ''; const extract = (label) => { const lineBreakPattern = new RegExp(`${label}\\s*(?:\\r?\\n)+\\s*([^\\r\\n]+)`); const lineBreakMatch = text.match(lineBreakPattern); if (lineBreakMatch) return lineBreakMatch[1].trim(); const inlinePattern = new RegExp(`${label}\\s*[::]?\\s*([^\\r\\n]+)`); const inlineMatch = text.match(inlinePattern); return inlineMatch ? inlineMatch[1].trim() : ''; }; return { orderId: extract('订单号'), orderType: extract('订单类型'), status: extract('状态'), tradeType: extract('交易类型'), customerCategory: extract('客户分类'), dealerName: extract('二级经销商名称'), dealerUid: extract('二级经销商UID'), customerType: extract('客户类型'), opportunityId: extract('商机ID'), paymentTime: extract('支付时间'), orderTime: extract('下单时间'), productName: extract('产品名称'), productCode: extract('产品code'), originalPriceCny: extract('订单原价\\(CNY\\)'), paidAmountCny: extract('实付金额\\(CNY\\)'), discount: extract('订单折扣'), payableAmountCny: extract('应付金额(实付\\+代金券)\\(CNY\\)'), couponAmountCny: extract('代金券金额\\(CNY\\)'), }; }); } async function extractCustomerDetail(page) { return page.evaluate(() => { const normalize = (value) => String(value || '') .replace(/\u00a0/g, ' ') .trim(); const text = normalize(document.body?.innerText || '').replace(/\r/g, ''); const extract = (label, sourceText = text) => { const lineBreakPattern = new RegExp(`${label}\\s*(?:\\n)+\\s*([^\\n]+)`); const lineBreakMatch = sourceText.match(lineBreakPattern); if (lineBreakMatch) return normalize(lineBreakMatch[1]); const inlinePattern = new RegExp(`${label}\\s*[::]?\\s*([^\\n]+)`); const inlineMatch = sourceText.match(inlinePattern); return inlineMatch ? normalize(inlineMatch[1]) : ''; }; const normalizeAmount = (value) => normalize(value).replace(/[¥,]/g, '').trim(); const buildSection = (startLabel, endLabel = '') => { const start = text.indexOf(startLabel); if (start < 0) return ''; const end = endLabel ? text.indexOf(endLabel, start + startLabel.length) : -1; if (end > start) return text.slice(start, end); return text.slice(start); }; const lastMonthSection = buildSection('上月应付总金额(CNY)', '本月应付总金额(CNY)'); const currentMonthSection = buildSection('本月应付总金额(CNY)'); const extractAmountFromSection = (sectionText, label) => normalizeAmount(extract(label, sectionText)); let department = ''; const table = Array.from(document.querySelectorAll('table')).find((node) => (node.innerText || '').includes('所属部门'), ); if (table) { const rows = table.querySelectorAll('tbody tr'); for (const row of rows) { const cells = row.querySelectorAll('td'); if (cells.length >= 2) { const value = normalize(cells[1]?.innerText || cells[1]?.textContent || ''); if (value) { department = value; break; } } } } return { customerAccount: extract('客户账号'), customerName: extract('客户名称'), customerType: extract('客户类型'), tradeMode: extract('交易模式'), customerSource: extract('客户来源'), realNameStatus: extract('实名状态'), email: extract('邮箱'), relationDate: extract('关联日期'), phone: extract('手机号'), remark: extract('备注'), paymentNoticeStatus: extract('代为支付告知状态'), department, lastMonthPayableTotalCny: extractAmountFromSection(lastMonthSection, '上月应付总金额(CNY)'), lastMonthPrepayCny: extractAmountFromSection(lastMonthSection, '预付费金额'), lastMonthPostpayCny: extractAmountFromSection(lastMonthSection, '后付费金额'), currentMonthPayableTotalCny: extractAmountFromSection(currentMonthSection, '本月应付总金额(CNY)'), currentMonthPrepayCny: extractAmountFromSection(currentMonthSection, '预付费金额'), currentMonthPostpayCny: extractAmountFromSection(currentMonthSection, '后付费金额'), }; }); }