A JSON list of podcast hosts, and the patterns they use in audio URLs.

If you have the URL of a podcast's audio file, this JSON file will help you work out who the host is; and abilities that host may have that may impact listener privacy.

This data is synched to podcast-privacy.com, which offers an API.

Each podcast host has an image, in optimised .png format. The filename is programmatically calculated from the podcast host's name, in lower case, with non-alphanumeric characters encoded as a "-". Here's our ugly code:

$imageFilename=str_replace(' ','-',str_replace('.','-',strtolower($host['hostname'])));

For now, the simplest way is to add to the file at src/hosts.json. Each podcast host may have multiple entries, but the URL patterns should be unique.

Each entry must contain the following properties:

• pattern: a unique string to spot within the audio URL (more accurately, within the domain portion of the URL)

• hostname: a humanly-readable name of the podcast that this is hosted with

• retailer: a boolean showing whether podcasters can buy hosting on this podcast host (BBC isn't; Libsyn is).

• iab: a boolean showing whether a podcast host is able to offer IAB Certified Compliant statistics. This does not mean that all statistics from this host are certified compliant.

• abilities_stats: a boolean showing whether this podcast host uses its logs to provide stats

• abilities_tracking: a boolean showing whether this podcast host uses its logs to provide tracking and attribution

• abilities_dynamicaudio: a boolean showing whether this podcast host is capable of dynamic content insertion, used for advertising or content.

• hosturl: a website link (escaped) that links to the homepage of the podcast host.

Each entry can contain the following properties:

• rss-pattern: an ADDITIONAL pattern for the RSS URL, to help with cases like LibsynPro which uses the same infrastructure for hosting retail and non-retail shows

• privacyhosturl: a link to the privacy policy of the podcast host

• notes: a freeform text link with details of evidence of 'abilities' claims. Ideally, all podcast hosts that have these indicated will have evidence in this field.

Podnews uses the below to extract a host's name, and privacy details, in podcast pages (example).

$stmt = $db->prepare("SELECT * FROM `podcasts-hosts` WHERE (INSTR(:url,pattern) AND INSTR(:rssurl,`rss-pattern`)) OR INSTR(:url,pattern) ORDER BY `rss-pattern` DESC LIMIT 1"); $stmt->execute(array(':url'=>$podcast['audiourl'],':rssurl'=>$podcast['feedUrl'])); $host = $stmt->fetch(PDO::FETCH_ASSOC);

Improvements are welcome. This list won't adequately spot original hosts via Feedburner or similar, but otherwise will catch most podcasts.