Automation Scripts

This repository contains a collection of scripts for the purpose of automating certain jobs in my podcast workflow.

launch-episode

launch-episode is the master script for launching an episode of our podcast, Lovey Dummies. It runs in a directory that contains the master WAV file, an exported Audacity label TXT file for chapters, and a JPEG image for social media. These 3 files should all have the same "episode-#" file name. The script takes 3 parameters: 1. episode number, 2. episode title, and 3. episode subtitle. It compresses the master file to OPUS and AAC formats, writes metadata, adds the chapters to the AAC file, renames the JPEG image to match the episode title, then uploads them to our AWS S3 bucket.

Lastly, it calculates the episode's duration (for the iTunes feed), puts together the front matter for the episode's blog entry, then opens it in vim for editing.

Dependencies

launch-episode requires the following packages, all of which can be found in Ubuntu 16.04's repos:

opus-tools fdkaac mp4v2-utils awscli vim

It also requires the chappy script in this repo, so make sure it's available in your $PATH! An easy and sudo-free way to do that is the following:

$ mkdir -p ~/.local/bin && cp chappy ~/.local/bin/chappy

chappy

chappy is a cute little 30-line bash script that parses exported label tracks from Audacity. It is briefly used in the launch-episode script to create properly formatted chapters for the website (WebVTT), as well as a chapter file integrated into our M4A file so that listeners can browse chapters in apps like Pocket Casts and Overcast.

Try it!

Clone the repository and run ./chappy tests/labels.txt from the root directory. You should now have two files in your directory, labels.vtt and labels.chapters.txt.

lovey-stats

lovey-stats is a script that pulls logfiles from an AWS S3 bucket, turns them into NCSA CLF logs, and feeds them to Webalizer to generate approximate download statistics. It's designed to run daily as a cron job to get yesterday's stats, although you can feed it a date parameter in the terminal to get stats for a particular date, month, or year. For example, lovey-stats "2016-10-20" will generate statistics for October 20, 2016; lovey-stats "2016-10" will generate them for the entire month of October 2016; and lovey-stats "2016" will generate them for all of 2016.

Dependencies

lovey-stats requires the following packages, all of which can be found in Ubuntu 16.04's repos:

awscli webalizer

Note about using cron

If you plan to use cron for generating daily statistics, keep in mind that it runs in a very limited env. I decided to use pip to install the latest version of awscli, but pip installed it in /usr/local/bin, which was not in the pared-down $PATH of cron. Just be sure to include your desired $PATH in your crontab!