From f571acbb683bc5fb34a50ebc0763ef980b014f08 Mon Sep 17 00:00:00 2001
From: Dennis Reimann <mail@dennisreimann.de>
Date: Thu, 16 Apr 2026 16:57:26 +0200
Subject: [PATCH] Paginate feed

---
 tasks/generate_feed.js | 131 ++++++++++++++++++++++++++++++++++++-----
 1 file changed, 117 insertions(+), 14 deletions(-)

diff --git a/tasks/generate_feed.js b/tasks/generate_feed.js
index 798a5e66fce..ee31936cf6a 100644
--- a/tasks/generate_feed.js
+++ b/tasks/generate_feed.js
@@ -11,19 +11,122 @@ const json2xmlOpts = {
   cdataPropName: '__cdata',
   indentBy: '  '
 }
-const builder = new XMLBuilder(json2xmlOpts)
 
-// Load and adapt feed
-const outputXML = builder.build(feed)
-const validation = XMLValidator.validate(outputXML)
-if (validation) {
-  write(
-    'dist/feed.xml',
-    xmlFormat(outputXML, {
-      indentation: json2xmlOpts.indentBy,
-      collapseContent: true
-    })
-  )
-} else {
-  console.error(validation.err)
+const PAGE_SIZE = 21
+
+function writePage(feedJson, itemsSlice, pageIndex, totalPages) {
+  const builder = new XMLBuilder(json2xmlOpts)
+  const feedCopy = JSON.parse(JSON.stringify(feedJson))
+
+  // Ensure channel.item is the page slice
+  feedCopy.rss = feedCopy.rss || {}
+  feedCopy.rss.channel = feedCopy.rss.channel || {}
+  feedCopy.rss.channel.item = itemsSlice
+
+  // Build atom links: adapt self, add prev/next where appropriate
+  const origAtom = (feedJson.rss && feedJson.rss.channel && feedJson.rss.channel['atom:link']) || []
+  const origSelf = Array.isArray(origAtom)
+    ? (origAtom.find(l => l.__attr && l.__attr.rel === 'self') || {}).__attr
+    : (origAtom && origAtom.__attr) || {}
+  const originalHref = (origSelf && origSelf.href) || ''
+  const base = originalHref.replace(/\/feed(?:-page-\d+)?\.xml$/, '')
+
+  const pageNumber = pageIndex + 1
+  const selfHref = pageIndex === 0 ? `${base}/feed.xml` : `${base}/feed-page-${pageNumber}.xml`
+
+  const newAtom = []
+  newAtom.push({ __attr: { href: selfHref, rel: 'self', type: 'application/rss+xml' } })
+  // preserve hub if present
+  const hub = Array.isArray(origAtom) && origAtom.find(l => l.__attr && l.__attr.rel === 'hub')
+  if (hub) newAtom.push(hub)
+
+  if (pageIndex > 0) {
+    const prevHref = pageIndex === 1 ? `${base}/feed.xml` : `${base}/feed-page-${pageIndex}.xml`
+    newAtom.push({ __attr: { href: prevHref, rel: 'prev', type: 'application/rss+xml' } })
+  }
+
+  if (pageIndex < totalPages - 1) {
+    const nextHref = `${base}/feed-page-${pageNumber + 1}.xml`
+    newAtom.push({ __attr: { href: nextHref, rel: 'next', type: 'application/rss+xml' } })
+  }
+
+  feedCopy.rss.channel['atom:link'] = newAtom
+
+  const outputXML = builder.build(feedCopy)
+  const validation = XMLValidator.validate(outputXML)
+  if (validation) {
+    const filename = pageIndex === 0 ? 'dist/feed.xml' : `dist/feed-page-${pageNumber}.xml`
+    write(
+      filename,
+      xmlFormat(outputXML, {
+        indentation: json2xmlOpts.indentBy,
+        collapseContent: true
+      })
+    )
+    return filename
+  } else {
+    console.error('XML validation error:', validation.err)
+    return null
+  }
 }
+
+// Main: split items and write pages
+const items = (feed.rss && feed.rss.channel && feed.rss.channel.item) || []
+const totalPages = Math.max(1, Math.ceil(items.length / PAGE_SIZE))
+const written = []
+for (let i = 0; i < totalPages; i++) {
+  const start = i * PAGE_SIZE
+  const slice = items.slice(start, start + PAGE_SIZE)
+  const out = writePage(feed, slice, i, totalPages)
+  if (out) written.push(out)
+}
+
+if (written.length) {
+  console.log('Wrote paginated feeds:', written.join(', '))
+} else {
+  console.error('No feed files written')
+}
+
+// Additionally write a full archive feed with all items
+function writeFullFeed(feedJson, allItems) {
+  const builder = new XMLBuilder(json2xmlOpts)
+  const feedCopy = JSON.parse(JSON.stringify(feedJson))
+
+  feedCopy.rss = feedCopy.rss || {}
+  feedCopy.rss.channel = feedCopy.rss.channel || {}
+  feedCopy.rss.channel.item = allItems
+
+  // preserve and adapt atom links
+  const origAtom = (feedJson.rss && feedJson.rss.channel && feedJson.rss.channel['atom:link']) || []
+  const origSelf = Array.isArray(origAtom)
+    ? (origAtom.find(l => l.__attr && l.__attr.rel === 'self') || {}).__attr
+    : (origAtom && origAtom.__attr) || {}
+  const originalHref = (origSelf && origSelf.href) || ''
+  const base = originalHref.replace(/\/feed(?:-page-\d+)?\.xml$/, '')
+
+  const newAtom = []
+  newAtom.push({ __attr: { href: `${base}/feed-all.xml`, rel: 'self', type: 'application/rss+xml' } })
+  const hub = Array.isArray(origAtom) && origAtom.find(l => l.__attr && l.__attr.rel === 'hub')
+  if (hub) newAtom.push(hub)
+  feedCopy.rss.channel['atom:link'] = newAtom
+
+  const outputXML = builder.build(feedCopy)
+  const validation = XMLValidator.validate(outputXML)
+  if (validation) {
+    const filename = 'dist/feed-all.xml'
+    write(
+      filename,
+      xmlFormat(outputXML, {
+        indentation: json2xmlOpts.indentBy,
+        collapseContent: true
+      })
+    )
+    return filename
+  } else {
+    console.error('XML validation error (feed-all):', validation.err)
+    return null
+  }
+}
+
+const allOut = writeFullFeed(feed, items)
+if (allOut) console.log('Wrote full archive feed:', allOut)