Merge branch 'lotus' into fix/link-safety-bypasses

Two easy bypasses of the "this link goes somewhere else" check:

- Any whitespace in the link text disabled it, so "Log in at
  chat.lotusguild.org" → evil.example wasn't flagged. hostsFromText now
  scans each word for host-shaped tokens; a link is flagged if any named
  host is a different site. Bare file names (main.py, report.pdf) and
  version numbers are skipped inside sentences so ordinary links aren't
  flagged.
- Shared-hosting suffixes counted as one site (alice.github.io vs
  evil.github.io, *.pages.dev, netlify.app, vercel.app, …).
  registrableDomain now treats a curated list of them as public suffixes
  (longest match, incl. s3.amazonaws.com). Curated rather than the full
  PSL to stay inside the bundle budget.

Verified with unit tests and headless on real messages: the four bypass
cases are flagged; the honest sentence, file/version text and "click
here" are not.

Fixes #335

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017PCb59uY3UKqDqt7B6hzd5

src/app/utils/linkSafety.test.ts

@@ -38,3 +38,38 @@ test('hostFromText only accepts URL/host-shaped text; IDN becomes punycode', ()
   assert.equal(analyzeLink('paypal.com', 'https://xn--pypal-4ve.com/')?.punycode, true);
   assert.equal(analyzeLink('paypal.com', 'https://xn--pypal-4ve.com/')?.mismatch, true);
 });
+
+test('a trusted name inside a sentence still exposes a different destination (#335)', () => {
+  const a = analyzeLink('Log in at chat.lotusguild.org', 'https://evil.example/');
+  assert.equal(a?.mismatch, true);
+  assert.equal(a?.shownHost, 'chat.lotusguild.org');
+  assert.equal(
+    analyzeLink('chat.lotusguild.org (official)', 'https://evil.example/')?.mismatch,
+    true,
+  );
+  assert.equal(
+    analyzeLink('Open https://paypal.com/login now', 'https://evil.example/')?.mismatch,
+    true,
+  );
+  // honest sentences are not flagged
+  assert.equal(
+    analyzeLink('Log in at chat.lotusguild.org', 'https://chat.lotusguild.org/login')?.mismatch,
+    false,
+  );
+  assert.equal(analyzeLink('read the docs', 'https://evil.example/')?.mismatch, false);
+  assert.equal(analyzeLink('see main.py and report.pdf', 'https://github.com/x')?.mismatch, false);
+  assert.equal(analyzeLink('release v1.2.3 notes', 'https://github.com/x')?.mismatch, false);
+});
+
+test('shared hosting suffixes are separate sites (#335)', () => {
+  assert.equal(registrableDomain('alice.github.io'), 'alice.github.io');
+  assert.equal(registrableDomain('www.alice.github.io'), 'alice.github.io');
+  assert.equal(registrableDomain('bucket.s3.amazonaws.com'), 'bucket.s3.amazonaws.com');
+  assert.equal(analyzeLink('alice.github.io', 'https://evil.github.io/')?.mismatch, true);
+  assert.equal(analyzeLink('lotusguild.pages.dev', 'https://evil.pages.dev/')?.mismatch, true);
+  assert.equal(
+    analyzeLink('alice.github.io/blog', 'https://alice.github.io/blog/1')?.mismatch,
+    false,
+  );
+  assert.equal(registrableDomain('news.bbc.co.uk'), 'bbc.co.uk');
+});

src/app/utils/linkSafety.ts

@@ -48,11 +48,53 @@ const TWO_LABEL_SUFFIXES = new Set([
   'com.tw',
 ]);
 
+// [Gitea #335] Shared-hosting suffixes (the Public Suffix List's "private"
+// section, the ones phishing actually uses): every customer gets their own
+// subdomain, so `alice.github.io` and `evil.github.io` are different sites.
+// A curated list rather than the full PSL keeps this out of the bundle budget.
+const SHARED_HOSTING_SUFFIXES = new Set([
+  'github.io',
+  'gitlab.io',
+  'codeberg.page',
+  'pages.dev',
+  'workers.dev',
+  'trycloudflare.com',
+  'netlify.app',
+  'vercel.app',
+  'web.app',
+  'firebaseapp.com',
+  'appspot.com',
+  'herokuapp.com',
+  'onrender.com',
+  'fly.dev',
+  'glitch.me',
+  'replit.app',
+  'repl.co',
+  'surge.sh',
+  'blogspot.com',
+  'wordpress.com',
+  'wixsite.com',
+  'weebly.com',
+  'azurewebsites.net',
+  'cloudfront.net',
+  'ngrok.io',
+  'ngrok-free.app',
+  'duckdns.org',
+  's3.amazonaws.com',
+]);
+
+const isPublicSuffix = (suffix: string): boolean =>
+  TWO_LABEL_SUFFIXES.has(suffix) || SHARED_HOSTING_SUFFIXES.has(suffix);
+
 export const registrableDomain = (host: string): string => {
   const labels = host.toLowerCase().replace(/\.$/, '').split('.');
-  if (labels.length <= 2) return labels.join('.');
-  const lastTwo = labels.slice(-2).join('.');
-  return TWO_LABEL_SUFFIXES.has(lastTwo) ? labels.slice(-3).join('.') : lastTwo;
+  // Longest matching public suffix wins (3 labels, e.g. s3.amazonaws.com, then 2).
+  for (const n of [3, 2]) {
+    if (labels.length > n && isPublicSuffix(labels.slice(-n).join('.'))) {
+      return labels.slice(-(n + 1)).join('.');
+    }
+  }
+  return labels.slice(-2).join('.');
 };
 
 const HOST_LIKE =
@@ -72,6 +114,85 @@ export const hostFromText = (text: string): string | null => {
   }
 };
 
+// Words that look like hosts but are usually files or versions in chat
+// ("see main.py", "report.pdf", "v1.2.3"): skipped inside sentences unless
+// written as a URL (scheme, www., or a path) so ordinary links aren't flagged.
+const FILE_EXTENSIONS = new Set([
+  'py',
+  'sh',
+  'md',
+  'rs',
+  'js',
+  'ts',
+  'tsx',
+  'jsx',
+  'json',
+  'yaml',
+  'yml',
+  'toml',
+  'txt',
+  'log',
+  'csv',
+  'pdf',
+  'zip',
+  'gz',
+  'tar',
+  'png',
+  'jpg',
+  'jpeg',
+  'gif',
+  'webp',
+  'svg',
+  'mp3',
+  'mp4',
+  'mov',
+  'html',
+  'htm',
+  'css',
+  'exe',
+  'dll',
+  'lock',
+  'cfg',
+  'conf',
+  'ini',
+  'xml',
+  'doc',
+  'docx',
+  'xls',
+  'xlsx',
+  'ppt',
+  'pptx',
+  'apk',
+  'iso',
+  'jar',
+  'java',
+  'go',
+  'rb',
+  'php',
+  'c',
+  'h',
+  'cpp',
+]);
+const TRIM_PUNCT = /^[\s"'“”‘’«»([{<]+|[\s"'“”‘’«»)\]}>.,;:!?]+$/g;
+
+/** Every host a sentence names (Gitea #335), e.g. "Log in at chat.lotusguild.org". */
+export const hostsFromText = (text: string): string[] => {
+  const hosts: string[] = [];
+  text.split(/\s+/).forEach((raw) => {
+    const word = raw.replace(TRIM_PUNCT, '');
+    const host = word && hostFromText(word);
+    if (!host) return;
+    const tld = host.split('.').pop() ?? '';
+    if (!/[a-z]/i.test(tld)) return; // version numbers, IPs-as-words
+    const urlShaped =
+      /^[a-z][a-z0-9+.-]*:\/\//i.test(word) || /^www\./i.test(word) || word.includes('/');
+    if (!urlShaped && host.split('.').length === 2 && FILE_EXTENSIONS.has(tld.toLowerCase()))
+      return;
+    if (!hosts.includes(host)) hosts.push(host);
+  });
+  return hosts;
+};
+
 export type LinkSafety = {
   /** The visible text names a different site than the destination. */
   mismatch: boolean;
@@ -92,8 +213,11 @@ export const analyzeLink = (text: string, href: string): LinkSafety | null => {
   if (url.hostname === 'matrix.to') return null;
   const realHost = url.hostname;
   const punycode = realHost.split('.').some((l) => l.startsWith('xn--'));
-  const shownHost = hostFromText(text) ?? undefined;
-  const mismatch =
-    shownHost !== undefined && registrableDomain(shownHost) !== registrableDomain(realHost);
-  return { mismatch, shownHost, realHost, punycode };
+  // A single URL/host-shaped text, or (Gitea #335) host-shaped words inside a
+  // sentence: "Log in at chat.lotusguild.org" must not hide a different site.
+  const single = hostFromText(text);
+  const shown = single ? [single] : hostsFromText(text);
+  const realSite = registrableDomain(realHost);
+  const other = shown.find((h) => registrableDomain(h) !== realSite);
+  return { mismatch: other !== undefined, shownHost: other ?? shown[0], realHost, punycode };
 };