Tell search engines and link previews what each page is

- /sitemap.xml, from the API's /slugs plus the landing and listings:
  every address in every language, with hreflang and x-default. A 503
  when the API is down, rather than a sitemap missing half the site.
- /robots.txt, naming the sitemap on whatever domain this is served on.
- Slugs renamed in the panel now redirect. The API has always kept the
  old address the moment one changed; middleware/redirects.ts answers
  it with a 301, from a list kept for a minute and through outages.
- og:url, og:image (the record's own picture, the logo otherwise) and
  schema.org JSON-LD: Organization on the landing; WebPage,
  CreativeWork, Person or Service elsewhere. Shared through core/head.
- The proxy-derived origin moves to core/origin.ts, used by the page
  shell, the landing and both server routes.

In development only /api/* reaches Nitro, so the two files 404 under
`npm start`; they were checked against a production build.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
pigui
2026-09-29 15:21:59 +02:00
parent 399b8d5450
commit 86b7485b32
17 changed files with 639 additions and 22 deletions

View File

@@ -185,7 +185,28 @@ async function list<T>(path: string, options: GetOptions = {}): Promise<ApiResul
return { ...result, data: result.data ?? [] };
}
/** One published record in `/slugs`: its address in every language it has. */
export interface SlugEntry {
id: number;
updated_at?: string;
alternates: Record<string, string>;
}
export interface Redirect {
from: string;
to: string;
status: 301 | 302;
/** Null: applies whatever language the visitor is browsing in. */
locale?: string | null;
}
export const pireonApi = {
/** Every published address, grouped by type. What the sitemap is built from. */
slugs: () => get<Record<string, SlugEntry[]>>('/slugs'),
/** Addresses that used to work and where they go now. Whole: there are never many. */
redirects: () => list<Redirect>('/redirects', { timeout: 3000 }),
site: (lang: Locale) => get<SitePayload>('/site', { lang }),
services: (lang: Locale, query?: GetOptions['query']) =>

View File

@@ -0,0 +1,42 @@
import { describe, expect, it } from 'vitest';
import { absoluteUrl, setStructuredData } from './head';
describe('absoluteUrl', () => {
it('puts a path on the origin and leaves a URL alone', () => {
expect(absoluteUrl('https://pireon.dev', '/logo.png')).toBe(
'https://pireon.dev/logo.png',
);
expect(absoluteUrl('https://pireon.dev', 'https://cdn.example/a.png')).toBe(
'https://cdn.example/a.png',
);
expect(absoluteUrl(null, '/logo.png')).toBe('/logo.png');
});
});
describe('setStructuredData', () => {
it("writes one script, replacing the previous page's", () => {
setStructuredData(document, { '@type': 'Person', name: 'Ada' });
setStructuredData(document, { '@type': 'Service', name: 'Formación' });
const scripts = document.head.querySelectorAll(
'script[type="application/ld+json"]',
);
expect(scripts).toHaveLength(1);
expect(JSON.parse(scripts[0].textContent ?? '')).toEqual({
'@context': 'https://schema.org',
'@type': 'Service',
name: 'Formación',
});
});
it('cannot be closed early by what it carries', () => {
setStructuredData(document, { name: '</script><script>alert(1)</script>' });
expect(
document.head.querySelector('script[type="application/ld+json"]')
?.textContent,
).not.toContain('</script>');
});
});

View File

@@ -0,0 +1,56 @@
/**
* The parts of `<head>` that describe a page to machines rather than people:
* the preview a chat app or social network draws, and schema.org data a
* search engine reads. Shared by the landing and every page reached by its
* address, so both say it the same way.
*/
import type { Meta } from '@angular/platform-browser';
const JSON_LD_ID = 'page-structured-data';
/** Relative paths become absolute on `origin`; absolute URLs pass through. */
export function absoluteUrl(origin: string | null, href: string): string {
return /^https?:\/\//.test(href) || !origin ? href : `${origin}${href}`;
}
/**
* `og:image` and the card size that goes with it. Without an image of the
* record's own, the site's logo, as a small card rather than a stretched one.
*/
export function setShareImage(
meta: Meta,
origin: string | null,
image?: string | null,
): void {
meta.updateTag({
property: 'og:image',
content: absoluteUrl(origin, image || '/logo.png'),
});
meta.updateTag({
name: 'twitter:card',
content: image ? 'summary_large_image' : 'summary',
});
}
/** One `<script type="application/ld+json">`, replaced rather than stacked. */
export function setStructuredData(
document: Document,
data: Record<string, unknown> | null,
): void {
document.getElementById(JSON_LD_ID)?.remove();
if (!data) {
return;
}
const script = document.createElement('script');
script.id = JSON_LD_ID;
script.setAttribute('type', 'application/ld+json');
// `<` escaped so no string in the data can close the script element.
script.textContent = JSON.stringify({
'@context': 'https://schema.org',
...data,
}).replace(/</g, '\\u003c');
document.head.appendChild(script);
}

View File

@@ -0,0 +1,38 @@
import { describe, expect, it } from 'vitest';
import { originFromHeaders } from './origin';
describe('originFromHeaders', () => {
it('believes the proxy over the host it connected to', () => {
expect(
originFromHeaders({
host: 'landing:3000',
'x-forwarded-host': 'pireon.dev',
'x-forwarded-proto': 'https',
}),
).toBe('https://pireon.dev');
});
it('takes the first hop of a list', () => {
expect(
originFromHeaders({
'x-forwarded-host': 'pireon.dev, internal',
'x-forwarded-proto': 'https,http',
}),
).toBe('https://pireon.dev');
});
it('assumes plain HTTP only on a developer machine', () => {
expect(originFromHeaders({ host: 'localhost:4200' })).toBe(
'http://localhost:4200',
);
expect(originFromHeaders({ host: 'pireon.dev' })).toBe(
'https://pireon.dev',
);
});
it('has nothing to say without headers', () => {
expect(originFromHeaders(undefined)).toBeNull();
expect(originFromHeaders({})).toBeNull();
});
});

View File

@@ -0,0 +1,27 @@
/**
* The origin a visitor used, as the reverse proxy in front of this app reports
* it. Absolute URLs — `canonical`, `hreflang`, the sitemap, `og:image` — need
* one, and this app has no configured address of its own: it answers on
* whichever domain the proxy sends it.
*/
export type HeaderBag = Record<string, string | string[] | undefined>;
export function originFromHeaders(
headers: HeaderBag | undefined,
): string | null {
if (!headers) {
return null;
}
const first = (value: string | string[] | undefined) =>
(Array.isArray(value) ? value[0] : value)?.split(',')[0]?.trim();
const host = first(headers['x-forwarded-host']) ?? first(headers['host']);
// Behind the proxy the header is always there; without one this is a
// developer's own machine, which does not speak TLS.
const protocol =
first(headers['x-forwarded-proto']) ??
(/^(localhost|127\.)/.test(host ?? '') ? 'http' : 'https');
return host ? `${protocol}://${host}` : null;
}

View File

@@ -0,0 +1,40 @@
import { describe, expect, it } from 'vitest';
import { findRedirect } from './redirects';
const redirects = [
{ from: '/proyectos/viejo', to: '/proyectos/nuevo', status: 301 as const },
{
from: '/ca/projectes/antic/',
to: '/ca/projectes/nou',
status: 302 as const,
},
{ from: '/bucle', to: '/bucle/', status: 301 as const },
];
describe('findRedirect', () => {
it('sends an old address to the new one, with its status', () => {
expect(findRedirect('/proyectos/viejo', redirects)).toEqual({
to: '/proyectos/nuevo',
status: 301,
});
expect(findRedirect('/ca/projectes/antic', redirects)).toEqual({
to: '/ca/projectes/nou',
status: 302,
});
});
it('treats a trailing slash as the same address', () => {
expect(findRedirect('/proyectos/viejo/', redirects)?.to).toBe(
'/proyectos/nuevo',
);
});
it('leaves every other address alone', () => {
expect(findRedirect('/proyectos/nuevo', redirects)).toBeNull();
});
it('never answers with a loop', () => {
expect(findRedirect('/bucle', redirects)).toBeNull();
});
});

View File

@@ -0,0 +1,28 @@
/**
* Which redirect, if any, an address answers with.
*
* The panel writes one the moment somebody renames a slug — the only moment
* the old address is still known — so a link shared before the rename lands
* on the record instead of a 404.
*/
import type { Redirect } from './api';
/** Trailing slashes are the same address; everything else must match exactly. */
function normalise(path: string): string {
return path.length > 1 ? path.replace(/\/+$/, '') : path;
}
export function findRedirect(
pathname: string,
redirects: Redirect[],
): Pick<Redirect, 'to' | 'status'> | null {
const path = normalise(pathname);
const match = redirects.find((redirect) => normalise(redirect.from) === path);
// One pointing at itself would loop forever in the visitor's browser.
if (!match || normalise(match.to) === path) {
return null;
}
return { to: match.to, status: match.status === 302 ? 302 : 301 };
}

View File

@@ -0,0 +1,49 @@
import { describe, expect, it } from 'vitest';
import { buildSitemap } from './sitemap';
const xml = buildSitemap('https://pireon.dev', {
pages: [
{
id: 1,
updated_at: '2026-09-27T20:36:14+00:00',
alternates: {
es: '/sobre-nosotros',
en: '/en/about-us',
ca: '/ca/qui-som',
},
},
],
});
describe('buildSitemap', () => {
it('lists every language of a record as its own url', () => {
for (const path of ['/sobre-nosotros', '/en/about-us', '/ca/qui-som']) {
expect(xml).toContain(`<loc>https://pireon.dev${path}</loc>`);
}
});
it('states every language on each url, with the default as x-default', () => {
expect(xml).toContain(
'hreflang="en" href="https://pireon.dev/en/about-us"',
);
expect(xml).toContain(
'hreflang="x-default" href="https://pireon.dev/sobre-nosotros"',
);
expect(xml).toContain('<lastmod>2026-09-27T20:36:14+00:00</lastmod>');
});
it('includes the landing and the listings this site composes', () => {
expect(xml).toContain('<loc>https://pireon.dev/</loc>');
// `&` escaped, as XML requires.
expect(xml).toContain('<loc>https://pireon.dev/?lang=en</loc>');
expect(xml).toContain('<loc>https://pireon.dev/radar</loc>');
expect(xml).toContain('<loc>https://pireon.dev/ca/col-labora</loc>');
});
it('is a well-formed document', () => {
const document = new DOMParser().parseFromString(xml, 'application/xml');
expect(document.getElementsByTagName('parsererror')).toHaveLength(0);
});
});

View File

@@ -0,0 +1,85 @@
/**
* `sitemap.xml`, built from the API's `/slugs` and the addresses this site
* composes itself.
*
* Each record is one `<url>` per language, and each carries every language as
* an `xhtml:link` — how a sitemap states `hreflang` for pages whose own HTML
* is not fetched first.
*/
import { DEFAULT_LOCALE, SUPPORTED_LOCALES, type SlugEntry } from './api';
import {
homeHref,
listingAlternates,
LISTINGS,
type ListingType,
} from './routes';
interface SitemapEntry {
alternates: Record<string, string>;
updated_at?: string;
}
export function buildSitemap(
origin: string,
slugs: Record<string, SlugEntry[]>,
): string {
const entries: SitemapEntry[] = [
// The landing answers on `?lang=`, not a prefix.
{
alternates: Object.fromEntries(
SUPPORTED_LOCALES.map((locale) => [locale, homeHref(locale)]),
),
},
...(Object.keys(LISTINGS) as ListingType[]).map((listing) => ({
alternates: listingAlternates(listing),
})),
...Object.values(slugs).flat(),
];
const urls = entries.flatMap((entry) =>
Object.values(entry.alternates).map((path) => url(origin, path, entry)),
);
return [
'<?xml version="1.0" encoding="UTF-8"?>',
'<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9" xmlns:xhtml="http://www.w3.org/1999/xhtml">',
...urls,
'</urlset>',
'',
].join('\n');
}
function url(origin: string, path: string, entry: SitemapEntry): string {
const links = Object.entries(entry.alternates).map(
([locale, href]) =>
` <xhtml:link rel="alternate" hreflang="${locale}" href="${escape(origin + href)}"/>`,
);
// The default language doubles as the fallback for everyone else.
const fallback = entry.alternates[DEFAULT_LOCALE];
if (fallback) {
links.push(
` <xhtml:link rel="alternate" hreflang="x-default" href="${escape(origin + fallback)}"/>`,
);
}
return [
' <url>',
` <loc>${escape(origin + path)}</loc>`,
...(entry.updated_at
? [` <lastmod>${escape(entry.updated_at)}</lastmod>`]
: []),
...links,
' </url>',
].join('\n');
}
function escape(value: string): string {
return value
.replace(/&/g, '&amp;')
.replace(/</g, '&lt;')
.replace(/>/g, '&gt;')
.replace(/"/g, '&quot;')
.replace(/'/g, '&apos;');
}