rss-parser

rss-parser is a small library that fetches or accepts RSS/Atom XML and converts it into a normalized JavaScript object with a title, description, link, and an items array - handling the inconsistencies between feed formats (RSS 2.0, Atom, dc: namespaced fields) so callers get a consistent shape regardless of the source feed’s dialect. Both parseURL (fetch and parse in one call) and parseString (parse an already-downloaded XML string) are supported, with either callback or Promise/async-await usage.

Built on top of xml2js for the underlying XML-to-object conversion, it adds feed-specific normalization on top: stripping HTML from a contentSnippet field, converting dates to ISO 8601, and unifying naming differences like Atom’s updated vs RSS’s pubDate. A customFields option lets consumers pull in feed-specific fields (like podcast <itunes:*> tags) that aren’t mapped by default, and a browser build is published in dist/ for use outside Node.

What You Get

• parser.parseURL(url) to fetch and parse a feed in one call, and parser.parseString(xml) for already-downloaded XML
• Normalized output fields across RSS and Atom - unified items, ISO 8601 isoDate, creator, and HTML-stripped contentSnippet
• A customFields option to map additional/non-standard XML tags (e.g. podcast itunes: fields) into the parsed feed or item objects
• Configurable HTTP behavior - request timeout, custom headers, and max redirect count for parseURL
• A pre-built browser distribution in dist/ alongside the Node.js module, plus TypeScript type definitions with generic custom-field typing

Common Use Cases

• Building a feed reader or aggregator that ingests many different RSS/Atom feeds and needs a single consistent item shape
• Pulling blog or news updates into a Node.js backend for further processing, storage, or re-publishing
• Parsing podcast RSS feeds by mapping itunes: namespaced fields via the customFields option
• Fetching and displaying a feed directly in the browser (with a CORS proxy) without a server-side component

Under The Hood

Architecture - The library is a thin, focused wrapper: lib/parser.js implements the Parser class with parseURL/parseString, delegating XML tokenization to xml2js and then running the result through lib/fields.js (defines the default RSS/Atom field mappings) and lib/utils.js (helpers for HTML stripping, date normalization, and applying customFields overrides); index.js is a 4-line entry point that just re-exports the Parser class. Tech Stack - Plain JavaScript (no TypeScript source, though index.d.ts ships hand-written types), depends on xml2js for XML parsing and entities for HTML entity decoding, built for the browser with Webpack 5, tested with Mocha/Chai plus Puppeteer for browser-environment verification. Code Quality - test/input and test/output hold golden-file fixtures for numerous real-world feeds (Reddit, podcasts, various blogs), and the test suite fails if any change alters the parsed output for those fixtures - a snapshot-style regression strategy that catches unintended parsing changes; however, the repo has had low recent commit activity, so newer edge cases may lag. API Design - The API mirrors common Node HTTP-client idioms (new Parser(), await parser.parseURL(url)) with both callback and Promise support, and the TypeScript definitions support generics (Parser<CustomFeed, CustomItem>) so custom fields get compile-time type checking.