@opentechevents/export-rss

Converts a valid OTE Feed (v0.4) into an RSS 2.0 document, one <item> per event.

import { feedToRss } from "@opentechevents/export-rss";

const rss = feedToRss(feed); // string, ready to serve as application/rss+xml

feedToRss is a pure function: no network, no filesystem, no clock. Output is deterministic — the same feed always produces byte-identical XML. It assumes the feed is valid; validate first with @opentechevents/validate.

CLI

ote-export-rss <feed.json> [output.xml]

Reads the feed, validates it, writes the RSS to output.xml (or stdout when omitted). Exit codes: 0 exported · 1 invalid JSON or invalid feed · 2 usage or I/O error.

Mapping (OTE v0.4 → RSS 2.0)

| OTE | RSS | | --- | --- | | feed title / description / url | channel title / description / link | | feed license (+ licenseUrl) | channel copyright, omitted when absent (see below) | | feed textLanguage | channel language | | feed updatedAt | channel lastBuildDate | | id | guid isPermaLink="false" | | name | title (prefixed [Tentative] / [Cancelled] / [Postponed] / [Rescheduled] / [Moved online] when applicable) | | url | link | | tags | one category per tag | | organizers | author (email (name)) for the first organizer with an email; otherwise dc:creator (name only) for the first organizer | | image | media:content/media:description (first image only) — chosen over <enclosure> specifically because it can carry image[].alt without needing to infer a MIME type or byte length (both required by <enclosure>, neither modeled by OTE) | | offers / cfp / eligibility / partOf | item description body, as unlabeled paragraphs (see below) | | everything else | item description body (see below) |

Emitting media:/dc: content declares both namespaces (xmlns:media, xmlns:dc) on the root <rss> element unconditionally, whether or not a given feed actually uses them — simpler and harmless.

Decisions worth knowing:

• RSS does not model events. Dates, venue, online URL, attendance mode and non-scheduled status go in the item body as entity-encoded HTML, followed by the event's own description. offers/cfp/eligibility/ partOf have no RSS structure either (accepted total loss, per the spec's own mapping tables) — folded into the same body as unlabeled paragraphs, not the labeled <strong>Field:</strong> style used for When/Where/Online/Attendance/Status, so this package's own reverse-parser (parse.ts, used by the preview app) — which only recognizes those five labels — reads them as part of the free-text description instead of silently dropping an unrecognized label.
• description is plain text or Markdown (OTE spec) and is rendered to HTML before being embedded (entity-encoded, per the point above) — so a feed reader that treats <description> as markup shows real formatting (bold, links, lists) instead of literal Markdown syntax. Raw inline/block HTML found inside the Markdown source is escaped rather than passed through live, so it can't smuggle real markup into whatever renders the item body. A lone newline is treated as a line break (breaks: true), since a short plain-text description is the common case and a single \n in one reads as an intended break, not two words meant to run together.
• feed.license is optional (D029). When every event declares its own license instead of a shared feed-level one, there's no single value for channel copyright to state — RSS's channel model has no per-item license — so it's omitted rather than guessing.
• No pubDate. OTE has no publication instant and the exporter never invents data; updatedAt means last modified, which is not the same thing. The channel's lastBuildDate comes from the feed's updatedAt.
• guid is never a permalink: id is a stable URI, not necessarily a fetchable page. The clickable page is link (from url), omitted when the event has none.
• author is structurally an email address (RSS 2.0's own definition); without one, dc:creator (Dublin Core, name-only) is the fallback — same reasoning as @opentechevents/export-ics's ORGANIZER.
• Cancelled and moved-online events stay published, marked in the title and body — same rationale as the spec: removing them would leave dead entries downstream. Unlike the iCal exporter, RSS's status text is free-form here, so moved-online needs no special-casing beyond the title prefix — the word itself already survives in the body.