Podcastindex-org / aggregator

Code, docs and discussion related to the Aggregator.
MIT License
16 stars 10 forks source link

Aggregator

Code, docs and discussion related to the Aggregator.

Aggrivate

The "puller" for the aggregators is called "Aggrivate" and is a node.js application. It uses Iconv, so you will need to install your distribution's "build-essential" package prior to running npm install.

Partytime

The "parser" for the aggregators is called "Partytime" and is a node.js application.

The puller and parser run independently of each other. Aggrivate is constantly polling for updated feed content. When it finds updated content, it downloads that content into a file named with the feed id and updates the feed record in the database with updated=<node id of this server>. The parser is then constantly polling the database for feeds marked with updated=<node id of this server>. When any are returned it looks for their files and parses them into the DB.

Batch DB inserts are used whenever possible.