compodio

Work in progress, currently online at https://www.compodio.com.

Some community radio stations post audio recordings of their shows online, but do not provide a podcast feed. This project provides rrs feeds for radio shows along side a UI with search and categoization features.

Overview

UI folder

Contains a next.js project using react-query for server state, CSS modules for styling, and fuse.js for search.

scraper folder

Contains a python Scrapy project that scrapes show data from radio station websites.

In addition:

OpenAI Whisper is used to transcribe audio files, identify language, and Open source summarization models are used to create episode summaries in ai_ep_summary.py

Sci-kit learn is used to:

• Generate tags for shows based on their descriptions in show_keywords.py
• Perform k-means clustering on show image colours in image_colour.py in order to generate the show specific CSS gradients used in the UI.

scrape_images.py downloads show images and uses Pillow to generate a set of responsive images for the UI.

data folder

Contains a php api, and the podcast feed generator.

The api fetches data from a mysql database and returns it as json for the UI to consume.

data/public/index.php is the entry point for poddcast feeds. It includes data/private/feed_retreiver.php to fetch data from the database and generates rss feeds for each show.

db folder

Contains database dumps

Development

Prerequisites

• python 3.11
• next.js 13.1.1
• mysql 8.0.23
• php 8.2.3

Setup

1. Clone the repo to local server folder (required for the php portion of the project)
2. Install python dependencies: pip install -r scraper/requirements.txt
3. Install next.js dependencies in the UI folder: npm install
4. import database from db/compodio.sql
5. add db credentials for php to data/private/db_credentials.php using data/private/db_credentials_default.php as a template
6. add db credentials for python to scraper/radio_scrape/radio_scrape/DBConfig.py using DBConfig-default.py as a template
7. define local path for show images in scraper/radio_scrape/radio_scrape/scrape_images.py
8. In /UI add .env.local file with the following contents:

NEXT_PUBLIC_API_URI = 'http://localhost/path-to-project-folder/data/'
NEXT_PUBLIC_feed_URI = 'http://localhost/path-to-project-folder/feed/'
NEXT_PUBLIC_image_server_URI = 'http://localhost/compodio_images/'