@sovovs/bycli 2.1.13 → 2.1.15

package/clis/weixin/_wechat/markdown.js

@@ -22,8 +22,21 @@ export function wechatArticleToMarkdown({ html, title, accountName, author, publ
     .replace(/&/g, '&amp;').replace(/</g, '&lt;').replace(/>/g, '&gt;')
     .replace(/"/g, '&quot;').replace(/'/g, '&#39;')
     .replace(/([\\`*_[\]{}()#+.!|>~-])/g, '\\$1');
+  // A URL must not go through `safe()`: escaping `. _ - ( ) #` and entity-encoding
+  // `&` corrupts the link so it can no longer be copied or followed. Rendered as an
+  // autolink it only needs whitespace and control characters dropped, `<`/`>`
+  // percent-encoded so the autolink cannot be closed early, and an http(s) scheme
+  // allowlist to reject `javascript:` and other hostile schemes.
+  const safeUrl = value => {
+    const raw = [...String(value || '')]
+      .filter(char => char > ' ' && char.codePointAt(0) !== 127)
+      .join('');
+    if (!/^https?:\/\//i.test(raw)) return '';
+    return raw.replace(/</g, '%3C').replace(/>/g, '%3E');
+  };
+  const articleUrl = safeUrl(url);
   const metadata = [accountName && `> 公众号: ${safe(accountName)}`, author && `> 作者: ${safe(author)}`,
     publishedAt && `> 发布时间: ${safe(publishedAt)}`, digest && `> 摘要: ${safe(digest)}`,
-    url && `> 原文链接: ${safe(url)}`].filter(Boolean);
+    articleUrl && `> 原文链接: <${articleUrl}>`].filter(Boolean);
   return [`# ${safe(title || 'Untitled')}`, ...metadata, '', '---', '', markdown, ''].join('\n');
 }

package/dist/src/download/article-download.js

@@ -212,6 +212,13 @@ function convertToMarkdown(contentHtml, codeBlocks, configure, cleanSelectors, s
     md = md.replace(/^[ \t]*[-·][ \t]*$/gm, '');
     md = md.replace(/^[ \t]+$/gm, '');
     md = md.replace(/[ \t]+$/gm, '');
+    // Bare-text URLs in adjacent inline nodes (WeChat renders each on its own
+    // visual line but emits no separator) fuse into one unusable string like
+    // `https://a/Xhttps://b/Y`. Split consecutive bare URLs onto their own lines.
+    // A URL whose own query carries another URL (`?next=https://…`, common in
+    // redirect wrappers) must stay intact, so skip breaks that follow a query or
+    // path delimiter. Markdown destinations are already bounded by `()`/`<>`.
+    md = md.replace(/(https?:\/\/[^\s<>()[\]]*?)(?=https?:\/\/)/gi, (match) => (/[=?&/:,;-]$/.test(match) ? match : `${match}\n`));
     md = md.replace(/\n{3,}/g, '\n\n');
     return md;
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sovovs/bycli",
-  "version": "2.1.13",
+  "version": "2.1.15",
   "publishConfig": {
     "access": "public"
   },
@@ -90,7 +90,7 @@
   "dependencies": {
     "@mozilla/readability": "^0.6.0",
     "@sovovs/bycli-recorder-core": "^0.1.0",
-    "@sovovs/wechat-article-crawler": "^1.1.0",
+    "@sovovs/wechat-article-crawler": "^1.1.3",
     "cli-table3": "^0.6.5",
     "commander": "^14.0.3",
     "js-yaml": "^4.1.0",

package/scripts/check-package-install.mjs

@@ -59,14 +59,14 @@ try {
     project, 'node_modules/@sovovs/bycli/package.json',
   ), 'utf8'));
   assert.equal(mainManifest.dependencies?.['@sovovs/bycli-recorder-core'], '^0.1.0');
-  assert.equal(mainManifest.dependencies?.['@sovovs/wechat-article-crawler'], '^1.1.0');
+  assert.equal(mainManifest.dependencies?.['@sovovs/wechat-article-crawler'], '^1.1.3');
 
   const coreDirectory = join(project, 'node_modules/@sovovs/bycli-recorder-core');
   const crawlerDirectoryInstalled = join(project, 'node_modules/@sovovs/wechat-article-crawler');
   const crawlerManifest = JSON.parse(readFileSync(
     join(crawlerDirectoryInstalled, 'package.json'), 'utf8',
   ));
-  assert.equal(crawlerManifest.version, '1.1.2');
+  assert.equal(crawlerManifest.version, '1.1.3');
   const projectRequire = createRequire(join(project, 'package.json'));
   const crawlerEntry = projectRequire.resolve('@sovovs/wechat-article-crawler');
   const crawlerModule = await import(pathToFileURL(crawlerEntry).href);