@sovovs/bycli 2.1.56 → 2.1.58

package/cli-manifest.json

@@ -13546,6 +13546,62 @@
     "sourceFile": "hupu/unlike.js",
     "navigateBefore": false
   },
+  {
+    "site": "ima",
+    "name": "download",
+    "description": "下载 ima 知识库中的原始文件",
+    "access": "read",
+    "domain": "ima.qq.com",
+    "strategy": "cookie",
+    "browser": true,
+    "args": [
+      {
+        "name": "knowledgeBase",
+        "type": "str",
+        "required": true,
+        "positional": true,
+        "help": "知识库名称或 ID"
+      },
+      {
+        "name": "file",
+        "type": "str",
+        "required": true,
+        "positional": true,
+        "help": "文件标题"
+      },
+      {
+        "name": "output",
+        "type": "str",
+        "default": ".",
+        "required": false,
+        "help": "输出文件或目录"
+      },
+      {
+        "name": "mediaId",
+        "type": "str",
+        "required": false,
+        "help": "IMA mediaId，用于精确定位"
+      },
+      {
+        "name": "viewerUrl",
+        "type": "str",
+        "required": false,
+        "help": "已打开的 IMA 查看器 URL（含 originUrl 签名参数）"
+      }
+    ],
+    "columns": [
+      "path",
+      "title",
+      "contentType",
+      "size",
+      "sha256",
+      "pages"
+    ],
+    "type": "js",
+    "modulePath": "ima/download.js",
+    "sourceFile": "ima/download.js",
+    "navigateBefore": false
+  },
   {
     "site": "ima",
     "name": "knowledge",

package/clis/ima/download-utils.js

@@ -0,0 +1,32 @@
+import crypto from 'node:crypto';
+import fs from 'node:fs';
+import path from 'node:path';
+
+const IMA_RESOURCE_HOSTS = new Set(['res-skb.ima.qq.com']);
+
+export function validateImaOriginUrl(value) {
+  let url;
+  try { url = new URL(String(value)); } catch { throw new Error('IMA originUrl is not a valid URL'); }
+  if (url.protocol !== 'https:' || !IMA_RESOURCE_HOSTS.has(url.hostname)) {
+    throw new Error(`IMA resource URL is not allowed: ${url.hostname}`);
+  }
+  return url.toString();
+}
+
+export function extractOriginUrl(viewerUrl) {
+  let url;
+  try { url = new URL(String(viewerUrl)); } catch { throw new Error('viewer URL is invalid'); }
+  const origin = url.searchParams.get('originUrl');
+  if (!origin) throw new Error('viewer URL does not contain originUrl');
+  return validateImaOriginUrl(origin);
+}
+
+export function resolveOutputPath(output, title) {
+  const target = String(output || '.');
+  if (path.extname(target)) return target;
+  return path.join(target, path.basename(String(title || 'ima-download')));
+}
+
+export function sha256File(filePath) {
+  return crypto.createHash('sha256').update(fs.readFileSync(filePath)).digest('hex');
+}

package/clis/ima/download.js

@@ -0,0 +1,67 @@
+import fs from 'node:fs';
+import path from 'node:path';
+import { httpDownload } from '@sovovs/bycli/download';
+import { cli, Strategy } from '@sovovs/bycli/registry';
+import { CommandExecutionError, EmptyResultError } from '@sovovs/bycli/errors';
+import { readImaMediaUrl, readKnowledgeBaseFromChrome } from './native-client.js';
+import { extractOriginUrl, resolveOutputPath, sha256File } from './download-utils.js';
+
+function countPdfPages(filePath) {
+  const bytes = fs.readFileSync(filePath);
+  if (bytes.subarray(0, 5).toString() !== '%PDF-') return null;
+  return (bytes.toString('latin1').match(/\/Type\s*\/Page\b/g) || []).length || null;
+}
+
+export async function runDownloadCommand(page, kwargs, deps = {}) {
+  const kb = String(kwargs.knowledgeBase ?? '').trim();
+  const title = String(kwargs.file ?? '').trim();
+  if (!kb || !title) throw new CommandExecutionError('knowledge base and file title are required');
+  const reader = deps.read ?? ((query) => readKnowledgeBaseFromChrome(page, query));
+  const envelope = await reader(kb);
+  const items = (envelope?.items ?? []).filter((item) => item.title === title
+    && (!kwargs.mediaId || item.mediaId === kwargs.mediaId));
+  if (items.length === 0) throw new EmptyResultError('ima download', `File "${title}" was not found`);
+  if (items.length > 1) throw new CommandExecutionError('Multiple files matched; specify --media-id');
+  const item = items[0];
+  let originUrl;
+  if (item.url) originUrl = item.url;
+  if (!originUrl && kwargs.viewerUrl) originUrl = extractOriginUrl(kwargs.viewerUrl);
+  if (!originUrl && page.getCurrentUrl) {
+    const currentUrl = await page.getCurrentUrl();
+    if (currentUrl?.includes('originUrl=')) originUrl = extractOriginUrl(currentUrl);
+  }
+  if (!originUrl && deps.readOriginalUrl) originUrl = await deps.readOriginalUrl(item);
+  if (!originUrl && page.fetchJson) originUrl = await readImaMediaUrl(page, { ...item, knowledgeBaseId: item.knowledgeBaseId || envelope.knowledgeBaseId });
+  if (!originUrl && process.platform === 'darwin') {
+    const { readKnowledgeBase: readKnowledgeBaseFromDesktop } = await import('./ax.js');
+    const desktop = await Promise.resolve(readKnowledgeBaseFromDesktop(kb));
+    const desktopItem = desktop?.items?.find((entry) => entry.title === title);
+    if (desktopItem?.url) originUrl = desktopItem.url;
+  }
+  if (!originUrl) throw new CommandExecutionError('IMA original file URL is unavailable; open the file in the IMA client first');
+  const outputPath = resolveOutputPath(kwargs.output || '.', item.title);
+  fs.mkdirSync(path.dirname(outputPath), { recursive: true });
+  const download = deps.httpDownload ?? httpDownload;
+  const result = await download(originUrl, outputPath, { timeout: 120000 });
+  if (!result.success) throw new CommandExecutionError(result.error || 'IMA file download failed');
+  const stat = fs.statSync(outputPath);
+  const sha256 = await sha256File(outputPath);
+  const pages = item.contentType === 'PDF' ? countPdfPages(outputPath) : null;
+  return [{ path: outputPath, title: item.title, contentType: item.contentType, size: stat.size, sha256, pages }];
+}
+
+cli({
+  site: 'ima', name: 'download', access: 'read', description: '下载 ima 知识库中的原始文件',
+  domain: 'ima.qq.com', strategy: Strategy.COOKIE, browser: true, navigateBefore: false,
+  args: [
+    { name: 'knowledgeBase', positional: true, required: true, help: '知识库名称或 ID' },
+    { name: 'file', positional: true, required: true, help: '文件标题' },
+    { name: 'output', flags: '--output <path>', default: '.', help: '输出文件或目录' },
+    { name: 'mediaId', flags: '--media-id <id>', help: 'IMA mediaId，用于精确定位' },
+    { name: 'viewerUrl', flags: '--viewer-url <url>', help: '已打开的 IMA 查看器 URL（含 originUrl 签名参数）' },
+  ],
+  columns: ['path', 'title', 'contentType', 'size', 'sha256', 'pages'],
+  func: (page, kwargs) => runDownloadCommand(page, kwargs),
+});
+
+export const __test__ = { countPdfPages, runDownloadCommand };

package/clis/ima/native-client.js

@@ -1,6 +1,7 @@
 import { listKnowledgeBases, readKnowledgeBaseFromApi } from './native-api.js';
 
 const IMA_WIKIS_URL = 'https://ima.qq.com/wikis';
+const IMA_GET_MEDIA_URL = 'https://ima.qq.com/cgi-bin/file_manager/get_media';
 
 function codedError(code, message) {
     return Object.assign(new Error(message), { code });
@@ -84,3 +85,23 @@ export async function readKnowledgeBaseFromChrome(page, query, dependencies = {}
         await releaseImaAuth(page, authId);
     }
 }
+
+export async function readImaMediaUrl(page, item, dependencies = {}) {
+    if (!page || typeof page.fetchJson !== 'function') {
+        throw codedError('IMA_ORIGINAL_URL_UNAVAILABLE', 'IMA browser page cannot call get_media');
+    }
+    const response = await page.fetchJson(IMA_GET_MEDIA_URL, {
+        method: 'POST',
+        body: {
+            knowledgeBaseId: String(item?.knowledgeBaseId ?? item?.knowledge_base_id ?? ''),
+            mediaId: String(item?.mediaId ?? item?.media_id ?? ''),
+            scene: dependencies.scene ?? 4,
+        },
+    });
+    const data = response && typeof response === 'object' ? response : {};
+    const url = data.jumpUrlInfo?.url ?? data.jump_url_info?.url;
+    if (Number(data.action) !== 1 || typeof url !== 'string' || !url) {
+        throw codedError('IMA_ORIGINAL_URL_UNAVAILABLE', data.toastText || data.toast_text || 'IMA did not return an exportable file URL');
+    }
+    return url;
+}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sovovs/bycli",
-  "version": "2.1.56",
+  "version": "2.1.58",
   "publishConfig": {
     "access": "public"
   },