diff --git a/src/pages/api/rss.test.ts b/src/pages/api/rss.test.ts
index d2ce4da..4cd11ee 100644
--- a/src/pages/api/rss.test.ts
+++ b/src/pages/api/rss.test.ts
@@ -174,6 +174,28 @@ describe('parseRssOrAtom', () => {
},
]);
});
+
+ it('decodes numeric HTML entities in descriptions', () => {
+ const xml = `
+
+ -
+ Encoded Description
+ https://example.com/encoded
+ Cloudflare’s edge roadmap…
+
+ `;
+
+ const items = parseRssOrAtom(xml, 5);
+
+ expect(items).toEqual([
+ {
+ title: 'Encoded Description',
+ link: 'https://example.com/encoded',
+ pubDate: '',
+ description: 'Cloudflare’s edge roadmap…',
+ },
+ ]);
+ });
});
describe('GET /api/rss', () => {
@@ -319,4 +341,65 @@ describe('GET /api/rss', () => {
expect(response.status).toBe(200);
expect(payload).toEqual({ items: [] });
});
+
+ it('falls back to WordPress posts API when feed fetch returns html', async () => {
+ const fetchMock = vi
+ .spyOn(globalThis, 'fetch')
+ .mockResolvedValueOnce(
+ new Response('challenge', {
+ status: 200,
+ headers: { 'Content-Type': 'text/html; charset=UTF-8' },
+ })
+ )
+ .mockResolvedValueOnce(
+ new Response(
+ JSON.stringify([
+ {
+ link: 'https://blog.jaysonknight.com/2026/03/26/edge-first-architecture-why-cloudflare-workers-changes-how-you-think-about-design/',
+ title: { rendered: 'Edge-First Architecture' },
+ excerpt: { rendered: 'Cloudflare’s distributed model...
' },
+ date_gmt: '2026-03-27T02:43:10',
+ date: '2026-03-26T22:43:10',
+ },
+ ]),
+ {
+ status: 200,
+ headers: { 'Content-Type': 'application/json; charset=utf-8' },
+ }
+ )
+ );
+
+ const response = await GET({
+ request: new Request('https://example.com/api/rss?url=https%3A%2F%2Fblog.jaysonknight.com%2Ffeed%2F&max=5'),
+ } as Parameters[0]);
+ const payload = (await response.json()) as { items: Array<{ title: string; link: string; description: string }> };
+
+ expect(response.status).toBe(200);
+ expect(payload.items).toEqual([
+ {
+ title: 'Edge-First Architecture',
+ link: 'https://blog.jaysonknight.com/2026/03/26/edge-first-architecture-why-cloudflare-workers-changes-how-you-think-about-design/',
+ pubDate: 'Fri, 27 Mar 2026 02:43:10 GMT',
+ description: 'Cloudflare’s distributed model...',
+ },
+ ]);
+
+ const fallbackCall = fetchMock.mock.calls[1];
+ expect(fallbackCall).toBeDefined();
+ const [fallbackUrl, fallbackOptions] = fallbackCall;
+ expect(fallbackUrl).toBeInstanceOf(URL);
+ expect((fallbackUrl as URL).toString()).toContain('https://blog.jaysonknight.com/wp-json/wp/v2/posts?');
+ expect((fallbackUrl as URL).searchParams.get('per_page')).toBe('5');
+ expect((fallbackUrl as URL).searchParams.get('_fields')).toBe('link,title.rendered,excerpt.rendered,date,date_gmt');
+ expect(fallbackOptions).toEqual(
+ expect.objectContaining({
+ cache: 'no-store',
+ signal: expect.any(AbortSignal),
+ headers: expect.objectContaining({
+ Accept: 'application/json',
+ 'User-Agent': 'Mozilla/5.0 (compatible; JKcom-RSSBot/1.0; +https://jaysonknight.com)',
+ }),
+ })
+ );
+ });
});
diff --git a/src/pages/api/rss.ts b/src/pages/api/rss.ts
index 3df9bdb..fc0c130 100644
--- a/src/pages/api/rss.ts
+++ b/src/pages/api/rss.ts
@@ -11,7 +11,7 @@ type FeedItem = {
const extractTag = (block: string, tags: string[]): string => {
for (const tag of tags) {
- const match = block.match(new RegExp(`<${tag}(?:\\s[^>]*)?>([\\s\\S]*?)<\\/${tag}>`, 'i'));
+ const match = block.match(new RegExp(`<${tag}(?:\s[^>]*)?>([\s\S]*?)<\/${tag}>`, 'i'));
if (match?.[1]) {
return match[1].trim();
}
@@ -22,20 +22,34 @@ const extractTag = (block: string, tags: string[]): string => {
const stripCdata = (value: string): string => value.replace(/^$/s, '$1').trim();
+const safeFromCodePoint = (codePoint: number): string => {
+ if (codePoint >= 0 && codePoint <= 0x10ffff) {
+ return String.fromCodePoint(codePoint);
+ }
+ return '\uFFFD';
+};
+
const decodeXmlEntities = (value: string): string =>
value
+ .replace(/&#x([0-9a-f]+);/gi, (_match, hex: string) => safeFromCodePoint(Number.parseInt(hex, 16)))
+ .replace(/&#(\d+);/g, (_match, dec: string) => safeFromCodePoint(Number.parseInt(dec, 10)))
.replace(/</g, '<')
.replace(/>/g, '>')
.replace(/"/g, '"')
.replace(/'/g, "'")
.replace(/&/g, '&');
-const cleanDescription = (value: string): string => {
- const text = decodeXmlEntities(stripCdata(value))
+/** Strips HTML tags, decodes entities, and normalises whitespace — without truncation. */
+const cleanText = (value: string): string =>
+ decodeXmlEntities(stripCdata(value))
.replace(/<[^>]+>/g, ' ')
.replace(/\s+/g, ' ')
.trim();
+/** Like cleanText but truncates to 200 characters with an ellipsis. */
+const cleanDescription = (value: string): string => {
+ const text = cleanText(value);
+
if (text.length <= 200) {
return text;
}
@@ -104,6 +118,74 @@ const getMax = (value: string | null): number => {
return Math.max(1, Math.min(parsed, 20));
};
+/**
+ * Wraps fetch with a timeout. Uses AbortSignal.timeout when available; otherwise
+ * falls back to AbortController + setTimeout and always clears the timer in a
+ * finally block so the event loop is never kept alive by a stale timer.
+ */
+const fetchWithTimeout = async (
+ input: RequestInfo | URL,
+ init: Omit = {},
+ timeoutMs = 8000
+): Promise => {
+ if (typeof AbortSignal.timeout === 'function') {
+ return fetch(input, { ...init, signal: AbortSignal.timeout(timeoutMs) });
+ }
+
+ const controller = new AbortController();
+ const timeoutId = setTimeout(() => controller.abort(), timeoutMs);
+ try {
+ return await fetch(input, { ...init, signal: controller.signal });
+ } finally {
+ clearTimeout(timeoutId);
+ }
+};
+
+const parseWordPressPosts = (
+ payload: unknown
+): Array<{ title: string; link: string; pubDate: string; description: string }> | null => {
+ if (!Array.isArray(payload)) {
+ return null;
+ }
+
+ const items = payload
+ .map((post) => {
+ const source = typeof post === 'object' && post ? (post as Record) : null;
+ const link = typeof source?.link === 'string' ? source.link.trim() : '';
+ const titleRendered =
+ source?.title && typeof source.title === 'object'
+ ? (source.title as Record).rendered
+ : undefined;
+ const excerptRendered =
+ source?.excerpt && typeof source.excerpt === 'object'
+ ? (source.excerpt as Record).rendered
+ : undefined;
+ const dateGmt = typeof source?.date_gmt === 'string' ? source.date_gmt.trim() : '';
+ const date = typeof source?.date === 'string' ? source.date.trim() : '';
+ // Use cleanText (no truncation) for titles so long titles are not silently cut off.
+ const title = cleanText(typeof titleRendered === 'string' ? titleRendered : '') || 'Untitled';
+ const pubDateSource = dateGmt ? `${dateGmt}Z` : date;
+ const parsedPubDate = pubDateSource ? new Date(pubDateSource) : null;
+ const pubDate = parsedPubDate && !Number.isNaN(parsedPubDate.getTime()) ? parsedPubDate.toUTCString() : pubDateSource;
+
+ return {
+ title,
+ link,
+ pubDate,
+ description: cleanDescription(typeof excerptRendered === 'string' ? excerptRendered : ''),
+ };
+ })
+ .filter((item) => item.link);
+
+ return items;
+};
+
+const respondWithItems = (items: FeedItem[]): Response =>
+ new Response(JSON.stringify({ items }), {
+ status: 200,
+ headers: { 'Content-Type': 'application/json' },
+ });
+
export const GET: APIRoute = async ({ request }) => {
const requestUrl = new URL(request.url);
const urlParam = requestUrl.searchParams.get('url')?.trim() ?? '';
@@ -127,18 +209,45 @@ export const GET: APIRoute = async ({ request }) => {
});
}
const sanitizedFeedUrl = sanitizeUrlForLog(parsedFeedUrl);
+ const tryWordPressFallback = async (): Promise => {
+ try {
+ const wpApiUrl = new URL('/wp-json/wp/v2/posts', `${parsedFeedUrl.origin}/`);
+ wpApiUrl.searchParams.set('per_page', String(max));
+ wpApiUrl.searchParams.set('_fields', 'link,title.rendered,excerpt.rendered,date,date_gmt');
+
+ const response = await fetchWithTimeout(wpApiUrl, {
+ headers: {
+ Accept: 'application/json',
+ 'User-Agent': 'Mozilla/5.0 (compatible; JKcom-RSSBot/1.0; +https://jaysonknight.com)',
+ },
+ cache: 'no-store',
+ });
+
+ if (!response.ok) {
+ return null;
+ }
+
+ const items = parseWordPressPosts(await response.json());
+ return items?.slice(0, max) ?? null;
+ } catch {
+ return null;
+ }
+ };
try {
- const response = await fetch(feedUrl, {
+ const response = await fetchWithTimeout(feedUrl, {
headers: {
Accept: 'application/rss+xml, application/atom+xml, application/xml, text/xml, */*',
'User-Agent': 'Mozilla/5.0 (compatible; JKcom-RSSBot/1.0; +https://jaysonknight.com)',
},
cache: 'no-store',
- signal: AbortSignal.timeout(8000),
});
if (!response.ok) {
+ const fallbackItems = await tryWordPressFallback();
+ if (fallbackItems) {
+ return respondWithItems(fallbackItems);
+ }
console.error('[api/rss] Failed to fetch feed with non-OK status:', response.status, 'for URL:', sanitizedFeedUrl);
return new Response(JSON.stringify({ error: `Failed to fetch feed (${response.status}).` }), {
status: 502,
@@ -149,6 +258,10 @@ export const GET: APIRoute = async ({ request }) => {
const contentType = response.headers.get('Content-Type')?.toLowerCase() ?? '';
const mimeType = contentType.split(';', 1)[0]?.trim() ?? '';
if (mimeType === 'text/html') {
+ const fallbackItems = await tryWordPressFallback();
+ if (fallbackItems) {
+ return respondWithItems(fallbackItems);
+ }
console.error('[api/rss] Feed returned text/html — likely bot challenge for URL:', sanitizedFeedUrl);
return new Response(
JSON.stringify({ error: 'Feed returned an HTML page instead of XML (possible bot challenge or redirect)' }),
@@ -161,6 +274,10 @@ export const GET: APIRoute = async ({ request }) => {
const xml = await response.text();
if (!isValidFeedDocument(xml)) {
+ const fallbackItems = await tryWordPressFallback();
+ if (fallbackItems) {
+ return respondWithItems(fallbackItems);
+ }
console.error('[api/rss] Response body is not a valid RSS/Atom document for URL:', sanitizedFeedUrl);
return new Response(JSON.stringify({ error: 'Feed URL did not return a valid RSS or Atom document' }), {
status: 502,
@@ -169,12 +286,19 @@ export const GET: APIRoute = async ({ request }) => {
}
const items = parseRssOrAtom(xml, max);
+ if (items.length === 0) {
+ const fallbackItems = await tryWordPressFallback();
+ if (fallbackItems) {
+ return respondWithItems(fallbackItems);
+ }
+ }
- return new Response(JSON.stringify({ items }), {
- status: 200,
- headers: { 'Content-Type': 'application/json' },
- });
+ return respondWithItems(items);
} catch (error) {
+ const fallbackItems = await tryWordPressFallback();
+ if (fallbackItems) {
+ return respondWithItems(fallbackItems);
+ }
const errorName = error instanceof Error ? error.name : 'UnknownError';
const errorMessage =
error instanceof Error
@@ -186,4 +310,4 @@ export const GET: APIRoute = async ({ request }) => {
headers: { 'Content-Type': 'application/json' },
});
}
-};
+};
\ No newline at end of file