Merge pull request #17 from dergigi/social-preview

Add dynamic social preview for article deep-links
This commit is contained in:
Gigi
2025-10-16 15:58:52 +02:00
committed by GitHub
3 changed files with 284 additions and 0 deletions

282
api/article-og.ts Normal file
View File

@@ -0,0 +1,282 @@
import type { VercelRequest, VercelResponse } from '@vercel/node'
import { RelayPool } from 'applesauce-relay'
import { nip19 } from 'nostr-tools'
import { AddressPointer } from 'nostr-tools/nip19'
import { NostrEvent, Filter } from 'nostr-tools'
import { Helpers } from 'applesauce-core'
const { getArticleTitle, getArticleImage, getArticleSummary } = Helpers
// Relay configuration (from src/config/relays.ts)
const RELAYS = [
'wss://relay.damus.io',
'wss://nos.lol',
'wss://relay.nostr.band',
'wss://relay.dergigi.com',
'wss://wot.dergigi.com',
'wss://relay.snort.social',
'wss://relay.current.fyi',
'wss://nostr-pub.wellorder.net',
'wss://purplepag.es',
'wss://relay.primal.net'
]
type CacheEntry = {
html: string
expires: number
}
const WEEK_MS = 7 * 24 * 60 * 60 * 1000
const memoryCache = new Map<string, CacheEntry>()
function escapeHtml(text: string): string {
return text
.replace(/&/g, '&amp;')
.replace(/</g, '&lt;')
.replace(/>/g, '&gt;')
.replace(/"/g, '&quot;')
.replace(/'/g, '&#039;')
}
function setCacheHeaders(res: VercelResponse, maxAge: number = 86400): void {
res.setHeader('Cache-Control', `public, max-age=${maxAge}, s-maxage=604800`)
res.setHeader('Content-Type', 'text/html; charset=utf-8')
}
interface ArticleMetadata {
title: string
summary: string
image: string
author: string
published?: number
}
async function fetchEventsFromRelays(
relayPool: RelayPool,
relayUrls: string[],
filter: Filter,
timeoutMs: number
): Promise<NostrEvent[]> {
const events: NostrEvent[] = []
await new Promise<void>((resolve) => {
const timeout = setTimeout(() => resolve(), timeoutMs)
relayPool.req(relayUrls, filter).subscribe({
next: (msg) => {
if (msg.type === 'EVENT') {
events.push(msg.event)
}
},
error: () => resolve(),
complete: () => {
clearTimeout(timeout)
resolve()
}
})
})
// Sort by created_at and return most recent first
return events.sort((a, b) => b.created_at - a.created_at)
}
async function fetchArticleMetadata(naddr: string): Promise<ArticleMetadata | null> {
const relayPool = new RelayPool()
try {
// Decode naddr
const decoded = nip19.decode(naddr)
if (decoded.type !== 'naddr') {
return null
}
const pointer = decoded.data as AddressPointer
// Connect to relays
const relayUrls = pointer.relays && pointer.relays.length > 0 ? pointer.relays : RELAYS
relayUrls.forEach(url => relayPool.open(url))
// Fetch article and profile in parallel
const [articleEvents, profileEvents] = await Promise.all([
fetchEventsFromRelays(relayPool, relayUrls, {
kinds: [pointer.kind],
authors: [pointer.pubkey],
'#d': [pointer.identifier || '']
}, 5000),
fetchEventsFromRelays(relayPool, relayUrls, {
kinds: [0],
authors: [pointer.pubkey]
}, 3000)
])
if (articleEvents.length === 0) {
return null
}
const article = articleEvents[0]
// Extract article metadata
const title = getArticleTitle(article) || 'Untitled Article'
const summary = getArticleSummary(article) || 'Read this article on Boris'
const image = getArticleImage(article) || '/boris-social-1200.png'
// Extract author name from profile
let authorName = pointer.pubkey.slice(0, 8) + '...'
if (profileEvents.length > 0) {
try {
const profileData = JSON.parse(profileEvents[0].content)
authorName = profileData.display_name || profileData.name || authorName
} catch {
// Use fallback
}
}
return {
title,
summary,
image,
author: authorName,
published: article.created_at
}
} catch (err) {
console.error('Failed to fetch article metadata:', err)
return null
} finally {
relayPool.close()
}
}
function generateHtml(naddr: string, meta: ArticleMetadata | null): string {
const baseUrl = 'https://read.withboris.com'
const articleUrl = `${baseUrl}/a/${naddr}`
const title = meta?.title || 'Boris Nostr Bookmarks'
const description = meta?.summary || 'Your reading list for the Nostr world. A minimal nostr client for bookmark management with highlights.'
const image = meta?.image?.startsWith('http') ? meta.image : `${baseUrl}${meta?.image || '/boris-social-1200.png'}`
const author = meta?.author || 'Boris'
return `<!doctype html>
<html lang="en">
<head>
<meta charset="UTF-8" />
<link rel="icon" type="image/x-icon" href="/favicon.ico" />
<link rel="icon" type="image/png" sizes="32x32" href="/favicon-32x32.png" />
<link rel="icon" type="image/png" sizes="16x16" href="/favicon-16x16.png" />
<link rel="apple-touch-icon" sizes="180x180" href="/apple-touch-icon.png" />
<meta name="viewport" content="width=device-width, initial-scale=1, viewport-fit=cover" />
<meta name="theme-color" content="#0f172a" />
<link rel="manifest" href="/manifest.webmanifest" />
<title>${escapeHtml(title)}</title>
<meta name="description" content="${escapeHtml(description)}" />
<link rel="canonical" href="${articleUrl}" />
<!-- Open Graph / Social Media -->
<meta property="og:type" content="article" />
<meta property="og:url" content="${articleUrl}" />
<meta property="og:title" content="${escapeHtml(title)}" />
<meta property="og:description" content="${escapeHtml(description)}" />
<meta property="og:image" content="${escapeHtml(image)}" />
<meta property="og:site_name" content="Boris" />
${meta?.published ? `<meta property="article:published_time" content="${new Date(meta.published * 1000).toISOString()}" />` : ''}
<meta property="article:author" content="${escapeHtml(author)}" />
<!-- Twitter Card -->
<meta name="twitter:card" content="summary_large_image" />
<meta name="twitter:url" content="${articleUrl}" />
<meta name="twitter:title" content="${escapeHtml(title)}" />
<meta name="twitter:description" content="${escapeHtml(description)}" />
<meta name="twitter:image" content="${escapeHtml(image)}" />
</head>
<body>
<noscript>
<p>Redirecting to <a href="/">Boris</a>...</p>
</noscript>
</body>
</html>`
}
function isCrawler(userAgent: string | undefined): boolean {
if (!userAgent) return false
const crawlers = [
'bot', 'crawl', 'spider', 'slurp', 'facebook', 'twitter', 'linkedin',
'whatsapp', 'telegram', 'slack', 'discord', 'preview'
]
const ua = userAgent.toLowerCase()
return crawlers.some(crawler => ua.includes(crawler))
}
export default async function handler(req: VercelRequest, res: VercelResponse) {
const naddr = (req.query.naddr as string | undefined)?.trim()
if (!naddr) {
return res.status(400).json({ error: 'Missing naddr parameter' })
}
const userAgent = req.headers['user-agent'] as string | undefined
const isCrawlerRequest = isCrawler(userAgent)
// If it's a regular browser (not a bot), serve HTML that loads SPA
// Use history.replaceState to set the URL before the SPA boots
if (!isCrawlerRequest) {
const articlePath = `/a/${naddr}`
// Serve a minimal HTML that sets up the URL and loads the SPA
const html = `<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<link rel="icon" type="image/x-icon" href="/favicon.ico">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Boris - Loading Article...</title>
<script>
// Set the URL to the article path before SPA loads
if (window.location.pathname !== '${articlePath}') {
history.replaceState(null, '', '${articlePath}');
}
</script>
<script>
// Redirect to index.html which will load the SPA
// The history state is already set, so SPA will see the correct URL
window.location.replace('/');
</script>
</head>
<body>
<div id="root"></div>
</body>
</html>`
res.setHeader('Content-Type', 'text/html; charset=utf-8')
res.setHeader('Cache-Control', 'no-cache, no-store, must-revalidate')
return res.status(200).send(html)
}
// Check cache for bots/crawlers
const now = Date.now()
const cached = memoryCache.get(naddr)
if (cached && cached.expires > now) {
setCacheHeaders(res)
return res.status(200).send(cached.html)
}
try {
// Fetch metadata
const meta = await fetchArticleMetadata(naddr)
// Generate HTML
const html = generateHtml(naddr, meta)
// Cache the result
memoryCache.set(naddr, { html, expires: now + WEEK_MS })
// Send response
setCacheHeaders(res)
return res.status(200).send(html)
} catch (err) {
console.error('Error generating article OG HTML:', err)
// Fallback to basic HTML with SPA boot
const html = generateHtml(naddr, null)
setCacheHeaders(res, 3600)
return res.status(200).send(html)
}
}

View File

@@ -18,6 +18,7 @@
<meta property="og:url" content="https://read.withboris.com/" />
<meta property="og:title" content="Boris - Nostr Bookmarks" />
<meta property="og:description" content="Your reading list for the Nostr world. A minimal nostr client for bookmark management with highlights." />
<meta property="og:image" content="https://read.withboris.com/boris-social-1200.png" />
<meta property="og:site_name" content="Boris" />
<!-- Twitter Card -->
@@ -25,6 +26,7 @@
<meta name="twitter:url" content="https://read.withboris.com/" />
<meta name="twitter:title" content="Boris - Nostr Bookmarks" />
<meta name="twitter:description" content="Your reading list for the Nostr world. A minimal nostr client for bookmark management with highlights." />
<meta name="twitter:image" content="https://read.withboris.com/boris-social-1200.png" />
<!-- Default to system theme until settings load from Nostr -->
<script>

Binary file not shown.

After

Width:  |  Height:  |  Size: 819 KiB