This logic is almost completely removed from Podverse (except for the Add by RSS feeds feature on mobile), and we plan to remove it completely soon and replace it with Ryan Hirsch's partytime.
A highly customizable package for fetching and parsing podcast feeds into simple and manageable JavaScript objects. For use with node and in the browser.
See the contrib.md file for information on how to get up-and-running and make contributions to this repository.
By default, podcast-feed-parser
will parse a podcast's xml feed and return an object with the following properties. meta
contains all of the information pertinent to the podcast show itself, and episodes
is list of episode objects which contain the information pertinent to each individual episode of the podcast.
{
meta: {
title: 'My podcast',
description: 'A podcast about whatever',
// ...
},
episodes: [
{
title: 'My Episode 1',
description: 'Episode 1',
pubDate: '2018-11-29T10:30:00.000Z',
// ...
}, {
title: 'My Episode 2',
description: 'Episode 2',
pubDate: '2018-11-28T10:30:00.000Z',
// ...
}
]
}
}
podcast-feed-parser
has two main functions: getPodcastFromFeed
and getPodcastFromURL
.
For fetching remote feeds from urls, use getPodcastFromURL
:
const podcastFeedParser = require("podcast-feed-parser")
// for fetching remote feeds, use getPodcastFromURL.
// Note that function must be async
async function printPodcastTitle (url) {
const podcast = await podcastFeedParser.getPodcastFromURL(url)
console.log(podcast.meta.title)
}
printPodcastTitle('http://feeds.gimletmedia.com/hearreplyall')
// "Reply All"
If you already have the podcast feed xml, use getPodcastFromFeed
:
const podcastFeedParser = require("podcast-feed-parser")
const fs = require('fs')
// if you already have the feed xml, you can parse
// synchronously with getPodcastFromFeed
const podcastFeed = fs.readFileSync('path/to/podcast-feed.xml', 'utf8')
const podcast = podcastFeedParser.getPodcastFromFeed(podcastFeed)
console.log(podcast.meta.title)
// "My Podcast"
podcast.episodes.forEach( (episode) => {
console.log(episode.title)
})
// "My Episode 1"
// "My Episode 2"
By default, podcast-feed-parser
will parse a feed for the following default fields, based on this standard. If a field is not found in a feed, it will quietly return undefined
.
{
meta: {
title: '',
description: '',
subtitle: '',
imageURL: '',
lastUpdated: '',
link: '',
language: '',
editor: '',
author: '',
summary: '',
categories: [],
owner: {
name: '',
email: ''
},
explicit: true,
complete: true,
blocked: true
},
episodes: [
{
title: '',
description: '',
imageURL: '',
pubDate: '',
link: '',
language: '',
enclosure: {
length: '0',
type: '',
url: ''
},
duration: 0,
summary: '',
blocked: true,
explicit: true,
order: 1
}
]
}
You can customize podcast-feed-parser
by passing an optional options
object to either of parsing functions, getPodcastFromFeed
and getPodcastFromURL
. The options object consists of three components: fields
, required
, and uncleaned
.
const options = {
// specifies the fields to be parsed from the podcast feed
fields: {
meta: [],
episodes: []
},
// specifies the fields which must be present for the function to return without
// an error
required: {
meta: [],
episodes: []
},
// specifies which fields should not have any of the cleaning functions applied
uncleaned: {
meta: [],
episodes: []
}
}
If no options object is passed to the parsing function, or if no fields are specified, then the fields listed in the Default section are applied.
If you specify particular fields for either meta
or episodes
, the final podcast object will only consist of those fields.
const options = {
fields : {
'meta': ['title', 'description', 'webMaster'],
'episodes': ['title', 'pubDate', 'timeline']
}
}
const podcast = podcastFeedParser.getPodcastFromFeed(sampleFeed, options)
console.log(podcast)
// { meta:
// { title: 'All Things Chemical',
// description: 'All Things Chemical is a podcast...',
// webMaster: 'Jackson Bierfeldt (jbierfeldt@gmail.com)'
// },
// episodes:
// [ { title: 'Confidential Business Information under TSCA',
// pubDate: '2018-11-29T10:30:00.000Z',
// timeline: 'http://timelinenotation.com/pages/documentation/notation.php' }
// ] }
// }
If you wish to use the default fields listed in the Default section, but to also parse an additional field, you can include 'default'
in the list of desired fields, along with the names of the additional fields you wish to parse.
const options = {
fields : {
'meta': ['default', 'webMaster'],
'episodes': ['default', 'timeline']
}
}
const podcast = podcastFeedParser.getPodcastFromFeed(sampleFeed, options)
console.log(podcast)
// { meta:
// { title: 'All Things Chemical',
// description: 'All Things Chemical is a podcast...',
// subtitle: 'A Podcast...',
// ...
// [all default meta fields]
// ...
// webMaster: 'Jackson Bierfeldt (jbierfeldt@gmail.com)'
// },
// episodes:
// [ { title: 'Confidential Business Information under TSCA',
// ...
// [all default episode fields]
// ...
// timeline: 'http://timelinenotation.com/pages/documentation/notation.php' }
// ] }
// }
By default, podcast-feed-parser
will quietly return an undefined
value if it tries to parse a field in a podcast feed that does not exist. If you wish for the function to halt and throw requiredError when a particular field is missing, you can specify those fields in the required
options object.
const options = {
fields : {
'meta': ['title', 'description'],
'episodes': ['title', 'pubDate']
},
required: {
'meta': ['title']
}
}
const podcast = podcastFeedParser.getPodcastFromFeed(sampleFeed, options)
// If podcast feed does not have a title attribute, parser will throw a requiredError
// If podcast feed does not have a description attribute, parsing will continue
// and the resulting podcast object will have an undefined attribute for meta.description
By default, podcast-feed-parser
will clean and standardize the data for several fields. For example, the podcast object returned by podcast-feed-parser
will always return durations as integer numbers of seconds, not as formatted strings. This is for convenience when working with many different unstandardized podcast feeds from different sources.
A full list of the fields which are cleaned and the functions used to clean them can be found in the CLEAN FUNCTIONS
section of index.js
.
If you would like the data in the podcast object to resemble exactly that of the podcast feed,
you can list fields which should remain uncleaned in the uncleaned
options object. These fields will have no cleaning applied to them after parsing.
// sampleFeed
<xml>
<itunes:duration>39:58</itunes:duration>
</xml>
// -------------
// default behavior with no options supplied
const podcast = podcastFeedParser.getPodcastFromFeed(sampleFeed)
console.log(podcast.episodes[0].duration)
// 2398
// -------------
const options = {
uncleaned: {
'episodes': ['duration']
}
}
const podcast = podcastFeedParser.getPodcastFromFeed(sampleFeed, options)
console.log(podcast.episodes[0].duration)
// ['39:58']
podcast-feed-parser
can also fetch and parse remote feeds in both the browser and server environment thanks to isomorphic-fetch
. Simply call getPodcastFromURL(url, options)
. Functions which fetch remote feeds must be asynchronous and utilize async/await.
const podcastFeedParser = require("podcast-feed-parser")
async function getNumberOfEpisodes (url) {
const podcast = await podcastFeedParser.getPodcastFromURL(url)
console.log(podcast.meta.title, podcast.episodes.length)
}
getNumberOfEpisodes('http://feeds.gimletmedia.com/hearreplyall')
// "Reply All"
// 148
podcast-feed-parser
has a variety of custom errors. These are exposed under exports.ERRORS
and are as follows:
exports.ERRORS = {
'parsingError' : new Error("Parsing error."),
'requiredError' : new Error("One or more required values are missing from feed."),
'fetchingError' : new Error("Fetching error."),
'optionsError' : new Error("Invalid options.")
}