This is a support script for my Music Every Day project. I have very specific needs, and I haven't made any effort to make the code more general than that, so I don't think anyone else will want to use it. But the code might be of interest, especially to anyone who wants to use MacRuby to script iTunes (and didn't get enough out of Matt Aimonetti's post).
There are three commands (as of this writing... but check med --help
for current info):
med add [-g GENRE] [-a ARTIST] PATH
Adds a file to your iTunes library. Automatically sets the album name and album artist to "Music Every Day" and me. Sets "year" to the current year. If the filename starts with an integer, extracts that and uses it as the track number.
med convert TRACK_NUMBER
Finds the track in my "Music Every Day" album with the given number, and creates an MP3 version (or whatever iTunes is configured to create in its preferences).
med upload TRACK_NUMBER
Finds the track (a la med convert) and uploads it to my web site,
with more info in the filename because sometimes MP3s get separated
from their context and it's nice to be able to identify them at a
glance.
The most interesting parts of the code are the parts that control iTunes:
- add a track to the library
- modify track metadata
- search for an existing track
iTunes might not be the right tool for this job. Certain things can't
be controlled by this script - for example, the output format of the
convert command, and the question of whether the library gets its
own copy of the original audio file or uses the existing copy in
place. And I might not even want the tracks in my library when the
process is over.
But it was a good chance to finally get some experience with MacRuby! Which is pretty neat.
The main thing I want to add is a command that drafts a Tumblr post with some tags and the track URL and a link to the project, so I only have to enter my notes about the new track.
It might also be fun to add a GUI, but I'm pretty comfortable with the command line.
Obviously MacRuby is key. Other dependencies, for which I'm grateful:
- Mike Williams's Clamp, a nice library for parsing command-line options and subcommands.
- require_all by Tony Arciero and Jarmo Pertman, which lets me load all my code without listing all of it.
Copyright © 2011 Erik Ostrom. All rights reserved.
The code is available under a simplified BSD license.