@sovovs/bycli 2.1.39 → 2.1.41

package/cli-manifest.json

@@ -13570,7 +13570,30 @@
       "title",
       "url",
       "contentType",
-      "addedDate"
+      "addedDate",
+      "mediaId",
+      "mediaType",
+      "mediaState",
+      "mediaAuditStatus",
+      "mediaTypeInfo",
+      "sourcePath",
+      "jumpUrl",
+      "createTime",
+      "updateTime",
+      "lastModifyTime",
+      "lastOpenTime",
+      "fileSize",
+      "abstract",
+      "introduction",
+      "tags",
+      "isTop",
+      "accessStatus",
+      "accessStatusUpdateTs",
+      "parseProgress",
+      "parseErrInfo",
+      "summaryState",
+      "coverUrls",
+      "logo"
     ],
     "defaultFormat": "json",
     "type": "js",
@@ -13578,6 +13601,28 @@
     "sourceFile": "ima/knowledge.js",
     "navigateBefore": false
   },
+  {
+    "site": "ima",
+    "name": "knowledge-list",
+    "description": "获取全部 ima 知识库的常用字段与原始元数据",
+    "access": "read",
+    "domain": "ima.qq.com",
+    "strategy": "cookie",
+    "browser": true,
+    "args": [],
+    "columns": [
+      "id",
+      "name",
+      "type",
+      "typeName",
+      "raw"
+    ],
+    "defaultFormat": "json",
+    "type": "js",
+    "modulePath": "ima/knowledge-list.js",
+    "sourceFile": "ima/knowledge-list.js",
+    "navigateBefore": false
+  },
   {
     "site": "imdb",
     "name": "person",
@@ -29035,7 +29080,11 @@
     "type": "js",
     "modulePath": "weixin/download.js",
     "sourceFile": "weixin/download.js",
-    "navigateBefore": "https://mp.weixin.qq.com"
+    "navigateBefore": "https://mp.weixin.qq.com",
+    "adapterConcurrency": {
+      "isolatedTabs": true,
+      "maxParallel": 3
+    }
   },
   {
     "site": "weixin",
@@ -29114,7 +29163,11 @@
     "type": "js",
     "modulePath": "weixin/download-publish-data.js",
     "sourceFile": "weixin/download-publish-data.js",
-    "navigateBefore": false
+    "navigateBefore": false,
+    "adapterConcurrency": {
+      "isolatedTabs": true,
+      "maxParallel": 3
+    }
   },
   {
     "site": "weixin",

package/clis/ima/knowledge-list.js

@@ -0,0 +1,48 @@
+import { cli, Strategy } from '@sovovs/bycli/registry';
+import { CommandExecutionError, ConfigError } from '@sovovs/bycli/errors';
+
+import { readKnowledgeBasesFromChrome } from './native-client.js';
+
+const COMMAND = 'ima knowledge-list';
+
+function throwDriverError(error) {
+    if (error?.code === 'IMA_CHROME_AUTH_REQUIRED') {
+        throw new ConfigError(
+            error.message || 'ima reader authentication is required',
+            'Open https://ima.qq.com/wikis in Chrome, sign in, then retry with the latest bycli Browser Bridge extension.',
+        );
+    }
+    throw new CommandExecutionError(
+        error?.message || `${COMMAND} failed to read knowledge bases`,
+    );
+}
+
+export async function runKnowledgeListCommand(read) {
+    try {
+        const items = await read();
+        if (!Array.isArray(items)) {
+            throw new CommandExecutionError(`${COMMAND} returned malformed knowledge-base rows`);
+        }
+        return items;
+    } catch (error) {
+        if (error instanceof ConfigError || error instanceof CommandExecutionError) throw error;
+        throwDriverError(error);
+    }
+}
+
+export const knowledgeListCommand = cli({
+    site: 'ima',
+    name: 'knowledge-list',
+    access: 'read',
+    description: '获取全部 ima 知识库的常用字段与原始元数据',
+    domain: 'ima.qq.com',
+    defaultFormat: 'json',
+    args: [],
+    columns: ['id', 'name', 'type', 'typeName', 'raw'],
+    strategy: Strategy.COOKIE,
+    browser: true,
+    navigateBefore: false,
+    func: async (page) => runKnowledgeListCommand(
+        () => readKnowledgeBasesFromChrome(page),
+    ),
+});

package/clis/ima/knowledge.js

@@ -84,6 +84,29 @@ export const knowledgeCommand = cli({
         'url',
         'contentType',
         'addedDate',
+        'mediaId',
+        'mediaType',
+        'mediaState',
+        'mediaAuditStatus',
+        'mediaTypeInfo',
+        'sourcePath',
+        'jumpUrl',
+        'createTime',
+        'updateTime',
+        'lastModifyTime',
+        'lastOpenTime',
+        'fileSize',
+        'abstract',
+        'introduction',
+        'tags',
+        'isTop',
+        'accessStatus',
+        'accessStatusUpdateTs',
+        'parseProgress',
+        'parseErrInfo',
+        'summaryState',
+        'coverUrls',
+        'logo',
     ],
     strategy: Strategy.COOKIE,
     browser: true,

package/clis/ima/native-api.js

@@ -4,6 +4,10 @@ const MEDIA_TYPE_NAMES = new Map([
     [12, '问答'], [13, 'TXT'], [14, 'XMIND'], [15, '音频'], [16, '视频网站'],
     [19, '播客'], [20, 'HTML'], [21, 'EPUB'], [98, '源代码'], [99, '文件夹'],
 ]);
+const KNOWLEDGE_BASE_TYPE_NAMES = new Map([
+    [1001, '我的知识库'], [1002, '共享知识库'],
+    [1004, '团队知识库'], [1005, '其他知识库'],
+]);
 function field(value, camelName, snakeName) {
     return value?.[camelName] ?? value?.[snakeName];
 }
@@ -20,6 +24,15 @@ function knowledgeBaseFromRaw(raw) {
     };
 }
 
+function knowledgeBaseRow(raw, type) {
+    return {
+        ...knowledgeBaseFromRaw(raw),
+        type,
+        typeName: KNOWLEDGE_BASE_TYPE_NAMES.get(type) ?? String(type),
+        raw,
+    };
+}
+
 function basesFromGroups(response) {
     const groups = field(response, 'results', 'results');
     if (!Array.isArray(groups)) throw new Error('ima API returned malformed knowledge-base groups');
@@ -29,7 +42,17 @@ function basesFromGroups(response) {
     });
 }
 
-export async function findKnowledgeBase(query, request) {
+function entriesFromGroups(response) {
+    const groups = field(response, 'results', 'results');
+    if (!Array.isArray(groups)) throw new Error('ima API returned malformed knowledge-base groups');
+    return groups.flatMap((group) => {
+        const list = field(group, 'knowledgeBaseList', 'knowledge_base_list');
+        const type = Number(field(group, 'type', 'type'));
+        return Array.isArray(list) ? list.map((raw) => ({ raw, type })) : [];
+    });
+}
+
+async function collectKnowledgeBaseEntries(request) {
     const initialGroups = [
         { type: 1001, limit: 20 },
         { type: 1002, limit: 20 },
@@ -47,7 +70,7 @@ export async function findKnowledgeBase(query, request) {
         if (Number(response?.code) !== 0) {
             throw new Error(response?.msg || `ima API error ${response?.code ?? 'unknown'}`);
         }
-        all.push(...basesFromGroups(response));
+        all.push(...entriesFromGroups(response));
         const groups = field(response, 'results', 'results');
         for (const group of groups) {
             const cursor = field(group, 'nextCursor', 'next_cursor');
@@ -61,19 +84,29 @@ export async function findKnowledgeBase(query, request) {
                     throw new Error(`ima API returned a repeated cursor for knowledge-base group ${type}`);
                 }
                 queuedPages.add(pageKey);
-                pendingPages.push({
-                    type,
-                    cursor: String(cursor),
-                    limit: 10,
-                });
+                pendingPages.push({ type, cursor: String(cursor), limit: 10 });
             }
         }
         const next = pendingPages.shift();
         if (!next) break;
-        response = await request('/get_knowledge_base_list', {
-            params: [next],
-        });
+        response = await request('/get_knowledge_base_list', { params: [next] });
     }
+    return all;
+}
+
+export async function listKnowledgeBases(request) {
+    const entries = await collectKnowledgeBaseEntries(request);
+    const unique = new Map();
+    for (const { raw, type } of entries) {
+        const row = knowledgeBaseRow(raw, type);
+        if (!unique.has(row.id)) unique.set(row.id, row);
+    }
+    return [...unique.values()];
+}
+
+export async function findKnowledgeBase(query, request) {
+    const all = (await collectKnowledgeBaseEntries(request))
+        .map(({ raw }) => knowledgeBaseFromRaw(raw));
 
     const matches = all.filter((base) => base.id === query || base.name === query);
     if (matches.length === 0) {
@@ -100,6 +133,30 @@ function articleFromRaw(raw, knowledgeBaseId, knowledgeBaseName, folderPath) {
         addedDate: field(raw, 'timeWording', 'time_wording')
             || field(raw, 'createTime', 'create_time')
             || null,
+        mediaId: field(raw, 'mediaId', 'media_id') ?? null,
+        mediaType,
+        mediaState: field(raw, 'mediaState', 'media_state') ?? null,
+        mediaAuditStatus: field(raw, 'mediaAuditStatus', 'media_audit_status') ?? null,
+        mediaTypeInfo: field(raw, 'mediaTypeInfo', 'media_type_info') ?? null,
+        sourcePath: sourcePath ?? null,
+        jumpUrl: jumpUrl ?? null,
+        createTime: field(raw, 'createTime', 'create_time') ?? null,
+        updateTime: field(raw, 'updateTime', 'update_time') ?? null,
+        lastModifyTime: field(raw, 'lastModifyTime', 'last_modify_time') ?? null,
+        lastOpenTime: field(raw, 'lastOpenTime', 'last_open_time') ?? null,
+        fileSize: field(raw, 'fileSize', 'file_size') ?? null,
+        abstract: field(raw, 'abstract', 'abstract') ?? null,
+        introduction: field(raw, 'introduction', 'introduction') ?? null,
+        tags: Array.isArray(field(raw, 'tags', 'tags')) ? field(raw, 'tags', 'tags') : [],
+        isTop: field(raw, 'isTop', 'is_top') ?? null,
+        accessStatus: field(raw, 'accessStatus', 'access_status') ?? null,
+        accessStatusUpdateTs: field(raw, 'accessStatusUpdateTs', 'access_status_update_ts') ?? null,
+        parseProgress: field(raw, 'parseProgress', 'parse_progress') ?? null,
+        parseErrInfo: field(raw, 'parseErrInfo', 'parse_err_info') ?? null,
+        summaryState: field(raw, 'summaryState', 'summary_state') ?? null,
+        coverUrls: Array.isArray(field(raw, 'coverUrls', 'cover_urls'))
+            ? field(raw, 'coverUrls', 'cover_urls') : [],
+        logo: field(raw, 'logo', 'logo') ?? null,
     };
 }
 

package/clis/ima/native-client.js

@@ -1,4 +1,4 @@
-import { readKnowledgeBaseFromApi } from './native-api.js';
+import { listKnowledgeBases, readKnowledgeBaseFromApi } from './native-api.js';
 
 const IMA_WIKIS_URL = 'https://ima.qq.com/wikis';
 
@@ -33,7 +33,7 @@ async function triggerImaAuthRequest(page, query) {
     }, query);
 }
 
-export async function readKnowledgeBaseFromChrome(page, query, dependencies = {}) {
+async function acquireImaChromeAuth(page, query, dependencies = {}) {
     if (!page || typeof page.startImaAuthCapture !== 'function'
         || typeof page.readImaAuth !== 'function' || typeof page.requestImaReader !== 'function'
         || typeof page.evaluate !== 'function') {
@@ -44,12 +44,11 @@ export async function readKnowledgeBaseFromChrome(page, query, dependencies = {}
     }
     const timeoutMs = dependencies.timeoutMs ?? 30_000;
     const sleep = dependencies.sleep ?? wait;
-    let authId;
     try {
         await page.startImaAuthCapture();
         await page.goto(IMA_WIKIS_URL);
         await triggerImaAuthRequest(page, query);
-        authId = await waitForImaAuth(page, timeoutMs, sleep);
+        return await waitForImaAuth(page, timeoutMs, sleep);
     } catch (error) {
         if (error?.code === 'IMA_CHROME_AUTH_REQUIRED') throw error;
         throw codedError(
@@ -57,14 +56,31 @@ export async function readKnowledgeBaseFromChrome(page, query, dependencies = {}
             `Chrome Browser Bridge could not acquire ima reader authentication: ${error instanceof Error ? error.message : String(error)}`,
         );
     }
+}
+
+async function releaseImaAuth(page, authId) {
+    if (typeof page.releaseImaAuth === 'function') {
+        await page.releaseImaAuth(authId).catch(() => {});
+    }
+}
+
+export async function readKnowledgeBasesFromChrome(page, dependencies = {}) {
+    const authId = await acquireImaChromeAuth(page, undefined, dependencies);
+    try {
+        return await listKnowledgeBases((path, body) => page.requestImaReader(authId, path, body));
+    } finally {
+        await releaseImaAuth(page, authId);
+    }
+}
+
+export async function readKnowledgeBaseFromChrome(page, query, dependencies = {}) {
+    const authId = await acquireImaChromeAuth(page, query, dependencies);
     try {
         return await readKnowledgeBaseFromApi(
             query,
             (path, body) => page.requestImaReader(authId, path, body),
         );
     } finally {
-        if (typeof page.releaseImaAuth === 'function') {
-            await page.releaseImaAuth(authId).catch(() => {});
-        }
+        await releaseImaAuth(page, authId);
     }
 }

package/clis/ima/utils.js

@@ -37,6 +37,7 @@ export function normalizeArticleUrl(value) {
 
 export function toKnowledgeRow(item) {
     return {
+        ...item,
         knowledgeBaseId: item.knowledgeBaseId || null,
         knowledgeBase: item.knowledgeBase,
         folderPath: Array.isArray(item.folderPath) ? item.folderPath : [],

package/clis/weixin/_wechat/article-artifact.js

@@ -0,0 +1,55 @@
+import { constants } from 'node:fs';
+import { access, readFile, realpath, stat } from 'node:fs/promises';
+import { dirname, extname, relative, resolve } from 'node:path';
+import { CommandExecutionError } from '@sovovs/bycli/errors';
+
+function isContained(parent, child) {
+  const rel = relative(parent, child);
+  return rel === '' || (!rel.startsWith('..') && !rel.startsWith('/') && !rel.startsWith('\\'));
+}
+
+async function validateRegularNonEmptyFile(path, label) {
+  await access(path, constants.R_OK);
+  const info = await stat(path);
+  if (!info.isFile() || info.size <= 0) throw new Error(`${label} is empty or not a regular file`);
+  return realpath(path);
+}
+
+function localMarkdownImageTargets(markdown) {
+  const targets = [];
+  for (const match of markdown.matchAll(/!\[[^\]]*\]\(([^)]+)\)/g)) {
+    let target = match[1].trim();
+    if (target.startsWith('<') && target.endsWith('>')) target = target.slice(1, -1);
+    target = target.split(/\s+["']/u, 1)[0];
+    if (!target || /^(?:https?:|data:|#|\/\/)/iu.test(target)) continue;
+    targets.push(decodeURI(target));
+  }
+  return targets;
+}
+
+export async function validateDownloadedArticleRows(rows, outputDir) {
+  try {
+    const successfulRows = rows.filter(row => row && String(row.status).toLowerCase() === 'success');
+    if (successfulRows.length === 0) return rows;
+    const resolvedOutput = await realpath(resolve(outputDir));
+    for (const row of successfulRows) {
+      if (typeof row.saved !== 'string' || extname(row.saved).toLowerCase() !== '.md') {
+        throw new Error('successful article row returned no Markdown path');
+      }
+      const saved = await validateRegularNonEmptyFile(resolve(row.saved), 'saved Markdown');
+      if (!isContained(resolvedOutput, saved)) throw new Error('saved Markdown escaped the output directory');
+      const articleDir = dirname(saved);
+      const markdown = await readFile(saved, 'utf8');
+      for (const target of localMarkdownImageTargets(markdown)) {
+        const image = await validateRegularNonEmptyFile(resolve(articleDir, target), 'local Markdown image');
+        if (!isContained(articleDir, image)) throw new Error('local Markdown image escaped the article directory');
+      }
+    }
+    return rows;
+  } catch (error) {
+    if (error instanceof CommandExecutionError) throw error;
+    throw new CommandExecutionError(
+      `Downloaded Weixin article artifact validation failed: ${error instanceof Error ? error.message : String(error)}`,
+    );
+  }
+}

package/clis/weixin/_wechat/article-identity.js

@@ -0,0 +1,27 @@
+import { createHash } from 'node:crypto';
+
+export function hashResourceValue(value) {
+  return createHash('sha256').update(String(value)).digest('hex');
+}
+
+export function canonicalWechatArticleIdentity(rawUrl) {
+  const url = new URL(rawUrl);
+  url.hash = '';
+  const host = url.hostname.toLowerCase();
+  let canonical;
+  if (url.pathname.startsWith('/s/')) {
+    canonical = `${host}${url.pathname}`;
+  } else {
+    const tupleNames = ['__biz', 'mid', 'idx', 'sn'];
+    const hasTuple = tupleNames.every(name => url.searchParams.has(name));
+    if (url.pathname === '/s' && hasTuple) {
+      canonical = `${host}/s?${tupleNames.map(name => `${name}=${url.searchParams.get(name)}`).join('&')}`;
+    } else {
+      const sorted = [...url.searchParams.entries()]
+        .sort(([leftKey, leftValue], [rightKey, rightValue]) => leftKey.localeCompare(rightKey) || leftValue.localeCompare(rightValue));
+      const query = new URLSearchParams(sorted).toString();
+      canonical = `${host}${url.pathname}${query ? `?${query}` : ''}`;
+    }
+  }
+  return hashResourceValue(canonical);
+}

package/clis/weixin/_wechat/publish-analysis.js

@@ -427,7 +427,7 @@ function safeFilename(title) {
   return `${name}.md`;
 }
 
-async function publishMarkdown(outputDir, filename, content) {
+async function publishMarkdown(outputDir, filename, content, beforePublish) {
   await mkdir(outputDir, { recursive: true });
   const temporary = resolve(outputDir, `.bycli-publish-analysis-${randomUUID()}.tmp`);
   await writeFile(temporary, content, { encoding: 'utf8', flag: 'wx' });
@@ -436,13 +436,17 @@ async function publishMarkdown(outputDir, filename, content) {
     const stem = filename.slice(0, -extension.length);
     for (let index = 0; index <= 9999; index += 1) {
       const path = resolve(outputDir, index === 0 ? filename : `${stem}-${index}${extension}`);
-      try { await link(temporary, path); return path; } catch (error) { if (error?.code !== 'EEXIST') throw error; }
+      try {
+        await beforePublish?.();
+        await link(temporary, path);
+        return path;
+      } catch (error) { if (error?.code !== 'EEXIST') throw error; }
     }
     throw new CommandExecutionError('WeChat publish analysis could not allocate a report filename');
   } finally { await unlink(temporary).catch(() => {}); }
 }
 
-export async function collectPublishAnalysis(page, { detailUrl, title, publishedAt, outputDir }) {
+export async function collectPublishAnalysis(page, { detailUrl, title, publishedAt, outputDir, beforePublish }) {
   if (typeof page?.goto !== 'function' || typeof page?.readNetworkCapture !== 'function') {
     throw new CommandExecutionError('WeChat publish analysis requires browser network capture support');
   }
@@ -480,7 +484,7 @@ export async function collectPublishAnalysis(page, { detailUrl, title, published
   }
   if (Object.keys(data).length === 0) throw new CommandExecutionError('WeChat publish analysis returned no readable analysis data');
   const content = formatAnalysisMarkdown({ title, publishedAt, data });
-  const path = await publishMarkdown(resolve(outputDir), safeFilename(title), content);
+  const path = await publishMarkdown(resolve(outputDir), safeFilename(title), content, beforePublish);
   const info = await stat(path);
   return { status: 'saved', path, size: info.size, metrics };
 }

package/clis/weixin/_wechat/publish-download.js

@@ -37,7 +37,7 @@ function safeFilename(title) {
   return name;
 }
 
-async function publishExclusively(source, outputDir, filename) {
+async function publishExclusively(source, outputDir, filename, beforePublish) {
   const extension = extname(filename);
   const stem = filename.slice(0, -extension.length);
   const temporary = resolve(outputDir, `.bycli-publish-data-${randomUUID()}.tmp`);
@@ -61,6 +61,7 @@ async function publishExclusively(source, outputDir, filename) {
     for (let index = 0; index <= 9999; index += 1) {
       const candidate = resolve(outputDir, index === 0 ? filename : `${stem}-${index}${extension}`);
       try {
+        await beforePublish?.();
         await link(temporary, candidate);
         return candidate;
       } catch (error) {
@@ -151,6 +152,7 @@ export async function downloadPublishData(page, options) {
     downloaded.filename,
     outputDir,
     safeFilename(options.title),
+    options.beforePublish,
   );
   try {
     await unlink(downloaded.filename);

package/clis/weixin/download-publish-data.js

@@ -3,6 +3,7 @@ import { access, stat } from 'node:fs/promises';
 import { extname, resolve } from 'node:path';
 import { ArgumentError, CommandExecutionError } from '@sovovs/bycli/errors';
 import { cli, Strategy } from '@sovovs/bycli/registry';
+import { assertCurrentAdapterLease, withAdapterResourceLocks } from '@sovovs/bycli/adapter-coordination';
 import { resolveBrowserCredentials } from './_wechat/auth-session.js';
 import { buildSecretSet, redactText } from './_wechat/redact.js';
 import { collectPublishAnalysis } from './_wechat/publish-analysis.js';
@@ -15,6 +16,7 @@ import {
   validatePublishDate,
   validatePublishedQuery,
 } from './_wechat/publish-records.js';
+import { canonicalWechatArticleIdentity, hashResourceValue } from './_wechat/article-identity.js';
 
 const METRIC_COLUMNS = [
   'readUsers', 'avgReadMinutes', 'finishedReadRatio', 'newFollowers', 'listenUsers',
@@ -35,6 +37,13 @@ function sanitizedError(error, secrets, fallback) {
     .replace(/https?:\/\/mp\.weixin\.qq\.com\/\S*/giu, '[REDACTED]');
 }
 
+function rethrowStopError(error) {
+  const code = error && typeof error === 'object' ? error.code : undefined;
+  if (code === 'AUTH_REQUIRED' || code === 'RATE_LIMITED' || code === 'CAPTCHA' || code === 'MFA_REQUIRED' || code === 'ADAPTER_LEASE_LOST') {
+    throw error;
+  }
+}
+
 async function validateArtifact(result, { label, expectedStatus, expectedExtension }) {
   if (!result || result.status !== expectedStatus) {
     throw new CommandExecutionError(`${label} returned an invalid status`);
@@ -71,6 +80,7 @@ export const downloadPublishDataCommand = cli({
   strategy: Strategy.INTERCEPT,
   browser: true,
   navigateBefore: false,
+  adapterConcurrency: { isolatedTabs: true, maxParallel: 3 },
   args: [
     { name: 'query', positional: true, required: true, help: 'Exact article URL or title text' },
     { name: 'date', help: 'Optional publication date in YYYY-MM-DD' },
@@ -104,9 +114,16 @@ export const downloadPublishDataCommand = cli({
       title: record.title,
       outputDir,
       timeoutSeconds,
+      beforePublish: assertCurrentAdapterLease,
     };
     const secrets = buildSecretSet({ token, cookie });
 
+    return withAdapterResourceLocks([
+      `article:${canonicalWechatArticleIdentity(record.url)}`,
+      `data:${hashResourceValue(`${record.msgid}:${record.publishDate ?? record.publishedAt}`)}`,
+      `output:${hashResourceValue(resolve(outputDir))}`,
+    ], async () => {
+
     let dataResult = null;
     let markdownResult = null;
     const errors = [];
@@ -118,6 +135,7 @@ export const downloadPublishDataCommand = cli({
         expectedExtension: '.xls',
       });
     } catch (error) {
+      rethrowStopError(error);
       errors.push(`Excel download failed: ${sanitizedError(error, secrets, 'Excel download failed')}`);
     }
     try {
@@ -131,6 +149,7 @@ export const downloadPublishDataCommand = cli({
         expectedExtension: '.md',
       });
     } catch (error) {
+      rethrowStopError(error);
       errors.push(`Markdown analysis failed: ${sanitizedError(error, secrets, 'Markdown analysis failed')}`);
     }
 
@@ -175,5 +194,6 @@ export const downloadPublishDataCommand = cli({
       detailLikes: metrics?.likes ?? null,
       detailComments: metrics?.comments ?? null,
     }];
+    });
   },
 });

package/clis/weixin/download.js

@@ -9,6 +9,8 @@
 import { cli, Strategy } from '@sovovs/bycli/registry';
 import { downloadArticle } from '@sovovs/bycli/download/article-download';
 import { AuthRequiredError } from '@sovovs/bycli/errors';
+import { assertCurrentAdapterLease, withAdapterResourceLocks } from '@sovovs/bycli/adapter-coordination';
+import { resolve } from 'node:path';
 import { buildExtractWechatArticleContentJs } from './_wechat/article-content.js';
 import {
     isTrustedSogouRedirectUrl,
@@ -16,6 +18,8 @@ import {
     normalizeWechatUrl,
     resolveWechatArticleUrl,
 } from './_wechat/article-link.js';
+import { canonicalWechatArticleIdentity, hashResourceValue } from './_wechat/article-identity.js';
+import { validateDownloadedArticleRows } from './_wechat/article-artifact.js';
 export { extractWechatArticleContent } from './_wechat/article-content.js';
 export {
     isTrustedSogouRedirectUrl,
@@ -152,6 +156,7 @@ cli({
     description: '下载微信公众号文章为 Markdown 格式',
     domain: 'mp.weixin.qq.com',
     strategy: Strategy.COOKIE,
+    adapterConcurrency: { isolatedTabs: true, maxParallel: 3 },
     args: [
         { name: 'url', required: true, help: 'WeChat article URL (mp.weixin.qq.com/s/xxx)' },
         { name: 'output', default: './weixin-articles', help: 'Output directory' },
@@ -160,6 +165,11 @@ cli({
     columns: ['title', 'author', 'publish_time', 'status', 'size', 'saved', 'source_url', 'resolved_url'],
     func: async (page, kwargs) => {
         const { sourceUrl, resolvedUrl, alreadyNavigated } = await resolveWechatArticleUrl(page, kwargs.url);
+        const outputDir = resolve(kwargs.output ?? './weixin-articles');
+        return withAdapterResourceLocks([
+            `article:${canonicalWechatArticleIdentity(resolvedUrl)}`,
+            `output:${hashResourceValue(outputDir)}`,
+        ], async () => {
         // Navigate and wait for content to load. Sogou resolution already lands on the article.
         if (!alreadyNavigated)
             await page.goto(resolvedUrl);
@@ -235,7 +245,7 @@ cli({
             codeBlocks: data?.codeBlocks,
             imageUrls: data?.imageUrls,
         }, {
-            output: kwargs.output,
+            output: outputDir,
             downloadImages: kwargs['download-images'],
             imageHeaders: { Referer: 'https://mp.weixin.qq.com/' },
             frontmatterLabels: { author: '公众号' },
@@ -244,7 +254,10 @@ cli({
                 return m ? m[1] : 'png';
             },
             secureMarkdown: true,
+            beforePublish: assertCurrentAdapterLease,
+        });
+        const validatedRows = await validateDownloadedArticleRows(rows, outputDir);
+        return validatedRows.map(row => ({ ...row, source_url: sourceUrl, resolved_url: resolvedUrl }));
         });
-        return rows.map(row => ({ ...row, source_url: sourceUrl, resolved_url: resolvedUrl }));
     },
 });

package/dist/src/adapter-coordination.d.ts

@@ -0,0 +1,26 @@
+import type { AdapterLease, AdapterLeaseRelease, AdapterLeaseRequest, AdapterResourceGrant } from './adapter-scheduler.js';
+export interface AdapterCoordinationDependencies {
+    acquire?: (request: AdapterLeaseRequest) => Promise<AdapterLease>;
+    heartbeat?: (lease: AdapterLease) => Promise<AdapterLease>;
+    release?: (release: AdapterLeaseRelease) => Promise<boolean>;
+    heartbeatIntervalMs?: number;
+    onLeaseLost?: () => Promise<void>;
+    warn?: (message: string) => void;
+}
+export declare function getCurrentAdapterLease(): AdapterLease | undefined;
+/**
+ * Renew the active lease immediately before an irreversible local publication.
+ * A restarted daemon or reclaimed lease rejects this fencing check.
+ */
+export declare function assertCurrentAdapterLease(dependencies?: {
+    heartbeat?: (lease: AdapterLease) => Promise<AdapterLease>;
+}): Promise<void>;
+export interface AdapterResourceDependencies {
+    acquire?: (lease: AdapterLease, keys: string[], timeoutMs: number) => Promise<AdapterResourceGrant>;
+    release?: (lease: AdapterLease, grantId: string) => Promise<boolean>;
+    timeoutMs?: number;
+    warn?: (message: string) => void;
+}
+export declare function settleAdapterOperationAfterTimeout<T>(operation: Promise<T>, timeoutMs: number, timeoutError: Error, stop: () => Promise<void>): Promise<T>;
+export declare function withAdapterResourceLocks<T>(keys: string[], operation: () => Promise<T>, dependencies?: AdapterResourceDependencies): Promise<T>;
+export declare function withAdapterCommandLease<T>(request: AdapterLeaseRequest, operation: () => Promise<T>, dependencies?: AdapterCoordinationDependencies): Promise<T>;