Building URLs

A Scribe Site record carries two URL-related fields that together define the root of the site’s content:

Field	Example	Meaning
site.url	"anthonycregan.co.uk"	The site’s domain (no protocol)
site.urlPrefix	"blog"	Path prefix for content — empty string if content is at the root

An article’s own slug lives in article.url (on an ArticleRef) or article.url (on a full Article).

Article URL pattern

The canonical URL for a published article is:

https://{site.url}/{site.urlPrefix}/{group.slug}/{article.url}

When urlPrefix is empty, the pattern is:

https://{site.url}/{group.slug}/{article.url}

Helper function

Write a small utility function once and reuse it across your app:

import type { Site } from '@scribe-atp/core';

function articleUrl(site: Site, groupSlug: string, articleSlug: string): string {
  const base = site.urlPrefix
    ? `https://${site.url}/${site.urlPrefix}`
    : `https://${site.url}`;
  return `${base}/${groupSlug}/${articleSlug}`;
}

Usage:

const url = articleUrl(site, group.slug, article.url);
// e.g. "https://anthonycregan.co.uk/blog/2024/my-first-post"
//   or "https://norobots.blog/2024/my-first-post"

Tip

When building internal links within your own site, use relative paths (/blog/2024/my-first-post) rather than absolute URLs. The absolute form is for canonical <link> tags, RSS items, and sitemap entries.

Deriving site slugs

The site slug is the rkey used to fetch a site record. Derive it from the site’s domain using toSlug:

import { toSlug } from '@scribe-atp/core';

toSlug('norobots.blog')        // → "norobots-blog"
toSlug('anthonycregan.co.uk')  // → "anthonycregan-co-uk"
toSlug('alice.bsky.social')    // → "alice-bsky-social"

toSlug replaces . and other non-alphanumeric characters with -.

Deriving article slugs from AT URIs

If you have an AT URI and need the article slug (rkey), use slugFromUri:

import { slugFromUri } from '@scribe-atp/core';

slugFromUri('at://did:plc:abc123/app.scribe.article/my-first-post')
// → "my-first-post"

This is useful when working directly with AT Protocol records rather than the normalised Site object.