revert(seo): drop the dist/blog origin rebase

Per maintainer review: the blog is BentoPDF's own content, so self-hosted
builds keep its canonicals on www.bentopdf.com.

scripts/enhance-seo-pages.mjs

@@ -3,14 +3,15 @@ import path from 'path';
 import { JSDOM } from 'jsdom';
 import { fileURLToPath } from 'url';
 import { serializeJsonLd } from './json-ld.mjs';
-import { DEFAULT_SITE_URL, rebaseDefaultOrigin } from './site-origin.mjs';
 
 const __filename = fileURLToPath(import.meta.url);
 const __dirname = path.dirname(__filename);
 
 const DIST_DIR = path.resolve(__dirname, '../dist');
-const SITE_URL = (process.env.SITE_URL || DEFAULT_SITE_URL).replace(/\/+$/, '');
-const BLOG_DIST_DIR = path.join(DIST_DIR, 'blog');
+const SITE_URL = (process.env.SITE_URL || 'https://www.bentopdf.com').replace(
+  /\/+$/,
+  ''
+);
 
 const FALLBACK_SOCIAL_IMAGE = '/images/og-tools.png';
 
@@ -21,26 +22,6 @@ function listTopLevelHtml() {
     .map((f) => path.join(DIST_DIR, f));
 }
 
-// generate-blog.mjs writes the blog index and posts with SITE_URL, but
-// hand-written blog pages (blog/author-alam.html) carry the official origin
-// literally, and generate-i18n-pages.mjs only rewrites top-level pages. Rebase
-// them here so a custom-SITE_URL build passes seo-audit's canonical check.
-function rebaseBlogPages() {
-  if (!fs.existsSync(BLOG_DIST_DIR)) return 0;
-  let rebased = 0;
-  for (const f of fs.readdirSync(BLOG_DIST_DIR)) {
-    if (!f.endsWith('.html')) continue;
-    const file = path.join(BLOG_DIST_DIR, f);
-    const html = fs.readFileSync(file, 'utf-8');
-    const out = rebaseDefaultOrigin(html, SITE_URL);
-    if (out !== html) {
-      fs.writeFileSync(file, out);
-      rebased++;
-    }
-  }
-  return rebased;
-}
-
 function imageExists(url) {
   if (!url) return true;
   let pathname;
@@ -178,9 +159,8 @@ function enhance() {
     fs.writeFileSync(file, dom.serialize());
     dom.window.close();
   }
-  const rebasedBlogPages = rebaseBlogPages();
   console.log(
-    `SEO enhance: ${files.length} pages processed, ${faqSchemaCount} FAQPage schemas added, ${rebasedBlogPages} blog pages rebased to ${SITE_URL}.`
+    `SEO enhance: ${files.length} pages processed, ${faqSchemaCount} FAQPage schemas added.`
   );
 }
 

scripts/generate-blog.mjs

@@ -3,14 +3,13 @@ import path from 'path';
 import { marked } from 'marked';
 import { fileURLToPath } from 'url';
 import { serializeJsonLd } from './json-ld.mjs';
-import { DEFAULT_SITE_URL } from './site-origin.mjs';
 
 const __filename = fileURLToPath(import.meta.url);
 const __dirname = path.dirname(__filename);
 
 const POSTS_DIR = path.resolve(__dirname, '../blog/posts');
 const BLOG_DIR = path.resolve(__dirname, '../blog');
-const SITE_URL = (process.env.SITE_URL || DEFAULT_SITE_URL).replace(/\/+$/, '');
+const SITE_URL = 'https://www.bentopdf.com';
 const AUTHOR = {
   name: 'Alam',
   url: `${SITE_URL}/blog/author-alam`,

scripts/site-origin.mjs

@@ -1,23 +0,0 @@
-/**
- * The public origin the build writes into canonical / og:url / JSON-LD URLs.
- * Self-hosters set SITE_URL at build time; the default is the official site.
- */
-export const DEFAULT_SITE_URL = 'https://www.bentopdf.com';
-
-// The exact authority only: not a longer host (`.com.example`), a port
-// (`:8443`, a different origin) or userinfo (`@other.example`, another host).
-const DEFAULT_ORIGIN_RE = /https:\/\/www\.bentopdf\.com(?![\w.:@-])/g;
-
-/**
- * Rewrites absolute URLs on the default origin to `siteUrl` (no trailing
- * slash). Used for hand-written pages that carry the official origin
- * literally, so a build with a custom SITE_URL emits canonicals on its own
- * host. Other origins are left untouched: look-alike hosts such as
- * `https://www.bentopdf.com.example`, other ports and userinfo URLs.
- */
-export function rebaseDefaultOrigin(html, siteUrl) {
-  if (siteUrl === DEFAULT_SITE_URL) return html;
-  // A callback keeps `siteUrl` literal: a string replacement would expand
-  // `$&`, `$1`, … inside it.
-  return html.replace(DEFAULT_ORIGIN_RE, () => siteUrl);
-}

src/tests/site-origin.test.ts

@@ -1,56 +0,0 @@
-import { describe, it, expect } from 'vitest';
-import {
-  DEFAULT_SITE_URL,
-  rebaseDefaultOrigin,
-} from '../../scripts/site-origin.mjs';
-
-const PAGE = `<link rel="canonical" href="https://www.bentopdf.com/blog/author-alam" />
-<meta property="og:image" content="https://www.bentopdf.com/images/author-alam.jpg" />
-<script type="application/ld+json">{"url":"https://www.bentopdf.com","sameAs":["https://github.com/alam00000"]}</script>`;
-
-describe('rebaseDefaultOrigin', () => {
-  it('rewrites every default-origin URL to the custom site URL', () => {
-    const out = rebaseDefaultOrigin(PAGE, 'https://pdf.example.org');
-    expect(out).not.toContain('www.bentopdf.com');
-    expect(out).toContain('href="https://pdf.example.org/blog/author-alam"');
-    expect(out).toContain(
-      'content="https://pdf.example.org/images/author-alam.jpg"'
-    );
-    expect(out).toContain('"url":"https://pdf.example.org"');
-    expect(out).toContain('https://github.com/alam00000');
-  });
-
-  it('keeps a subpath site URL intact', () => {
-    expect(
-      rebaseDefaultOrigin(
-        'href="https://www.bentopdf.com/blog/"',
-        'https://example.org/pdf'
-      )
-    ).toBe('href="https://example.org/pdf/blog/"');
-  });
-
-  it('leaves look-alike hosts alone', () => {
-    const html =
-      'href="https://www.bentopdf.com.example/x" href="https://www.bentopdf.community/"';
-    expect(rebaseDefaultOrigin(html, 'https://pdf.example.org')).toBe(html);
-  });
-
-  it('leaves other ports and userinfo URLs alone', () => {
-    const html =
-      'href="https://www.bentopdf.com:8443/blog/" href="https://[email protected]/"';
-    expect(rebaseDefaultOrigin(html, 'https://pdf.example.org')).toBe(html);
-  });
-
-  it('inserts the site URL literally', () => {
-    expect(
-      rebaseDefaultOrigin(
-        'href="https://www.bentopdf.com/blog/"',
-        'https://example.org/$&'
-      )
-    ).toBe('href="https://example.org/$&/blog/"');
-  });
-
-  it('is a no-op for the default site URL', () => {
-    expect(rebaseDefaultOrigin(PAGE, DEFAULT_SITE_URL)).toBe(PAGE);
-  });
-});