Integrating a music library: an example with Beets

Liquidsoap’s native sources can read from files and folders, but if your radio uses an important music library (more than a thousand tracks) sorting this library by folders may not be enough. In that case you would better maintain a music library queried by Liquidsoap. In this section we’ll do this with Beets. Beets holds your music catalog, cleans tracks’ tags before importing, and most importantly has a command-line interface we can leverage from Liquidsoap.

The following examples may also inspire you to integrate another library program or your own scripts. If you’re going with Beets, you’ll need an installation having the random plug-in enabled. We’ll examine how you can pick Beets tracks (local files) from Liquidsoap, using 3 techniques fitting different Liquidsoap sources :

The most intuitive is an infite track source, that queries Beets for each track via request.dynamic.list.
We can also add Beets as a protocol so you can query beets from requests
Finally we use Beets to generate a file suitable for playlist.

Before that, let’s see Beets queries that are the most interesting for a radio.

Beet queries

Queries are parameters that you usually append to the beet ls command : Beets will use them to find matching tracks. The random plug-in works the same, except that it returns only one track matching the query (see the plug-in’s documentation). Once your library is imported, you can try the following queries on the command line by typing beet ls [query] or beet random [query]. To test quickly, add the -t 60 option to beet random so it will select an hour worth of tracks matching your query.

Without selectors, queries search in a track’s title, artist, album name, album artist, genre and comments. Typing an artist name or a complete title usually match the exact track, by you could do a lovely playlist just by querying love.

But in a radio you’ll usually query on other fields. You can select tracks by genre with the genre: selector. Be careful that genre:Rock also matches Indie Rock, Punk Rock, etc. To select songs having english lyrics, use language:eng. Or pick 80s songs with year:1980..1990.

Beets also holds internal meta-data, like added : the date and time when you imported each song. You can use it to query tracks inserted over the past month with added:-1m... Or you can query track imported more than a year ago with added:..-1y. Beets also lets you set your own tags.

You can use the info plug-in to see everything Beets knows about title(s) matching a query by typing beet info -l [query]. See also the Beets’ documentation for more details on queries operators. All these options should allow you to create both general and specialiazed sources.

To use the following examples, replace the beet path by the complete path on your own installation (on UNIX systems, find it with which beet). We also always add the -f '$path' option, so beets only returns the matching track’s path.

A source querying each next track from Beets

As of Liquidsoap 1.4.2 we can use the request.dynamic.list to create a source calling Beets every time it needs to prepare its the next track:

def beets() =
  list.map(fun(item) -> request.create(item),
    process.read.lines(
      "/home/me/path/to/beet random -f '$path' my_own_query"
    )
  )
end

from_beets = request.dynamic.list(beets)

process.read.lines returns the command’s output as a list of strings. Each item in this list (so, each line returned by beet random) is processed (list.map) by an anonymous function that turns it into a request. Thus our function returns a list of requests, as needed by request.dynamic.list.

A more re-usable implementation would store beet’s path in a constant and make a function returning a function, so you can re-use it for all queries to Beets:

BEET = "/home/me/path/to/beet"

def beets(arg="") =
  fun() -> list.map(fun(item) -> request.create(item),
    process.read.lines(
      "#{BEET} random -f '$path' #{arg}"
    )
  )
end

music = request.dynamic.list(beets())
recent_music = request.dynamic.list(beets("added:-1m.."))
rock_music = request.dynamic.list(beets("genre:Rock"))

Beets as a requests protocol

If you’re queueing tracks with request.equeue, you may prefer to integrate Beets as a protocol. In that case, the list of paths returned by beet random -f '$path' fits directly what’s needed by protocol resolution:

def beets_protocol(~rlog,~maxtime,arg) =
  process.read.lines(
    "/home/me/path/to/beet random -f '$path' #{arg}"
  )
end

Then declare the beets protocol with

add_protocol("beets", beets_protocol,
  syntax = "Beets queries, see https://beets.readthedocs.io/en/stable/reference/query.html"
)

Once this is done, you can push a beets query from the telnet server : if you created request.equeue(id="userrequested"), the server command userrequested.push beets:All along the watchtower will push the Jimi Hendrix’s song.

Generate playlists with Beets

For Liquidsoap versions prior to 1.4.2, or if you don’t want to call Beets before every track, you can ask beet random to return a track list and save this to a temporary file suitable for playlist sources. Note that this example uses a redirection that is only supported by UNIX systems. Also it’s hacky and error-prone, we just leave it here as an example.

To do this we will again integrate Beets as a protocol ; the difference with the above protocol resolution is that our function will only return the path to the playlist’s temporary file. To use both protocols, be careful to give different protocol names (the first argument in add_protocol).

def beets_protocol(~rlog,~maxtime,arg) =
  tmp_playlist = file.temp("beetsplaylist", ".m3u8")
  ignore(process.read.lines(
    "/home/me/path/to/beet random -f '$path' #{arg} > #{tmp_playlist}"
  ))
  [tmp_playlist]
end

add_protocol("beets", beets_protocol)

Before cleaning files created by file.temp: beet random returns only one track matching the query, remember ? We can add -t 60 to the query, so it will return at most one hour of music matching the query. Sources would look like:

music = playlist("beets:-t 60", mode="normal", reload=3600)
recent_music = playlist("beets:-t 60 added:-1m..", mode="normal", reload=3600)
rock_music = playlist("beets:-t 60 genre:Rock", mode="normal", reload=3600)

We use mode="normal" because we don’t need Liquidsoap to re-randomize. reload=3600 will cause Liquidsoap to ask Beets for a new playlist every hour.

If you’re playing from only one playlist, be careful that beet random -t 60 returns a list of 60 minutes at most but it’s usually less : it’s likely that the first song will play again at the end of the hour. Also Beets tends to fill the last minutes with short songs, therefore short songs are picked more frequently than others. You can mitigate these two problems by reloading before the playlist end, by increasing 60 decreasing 3600.

Another downside of this technique is that it creates a different playlist file each time we reload the playlist. After a while, your /tmp will be filled of files like beetsplaylistXXXX123.m3u8. On UNIX we can add a cleanup function that regularly calls find -delete:

exec_at(freq=3600., pred={ true },
  fun () -> list.iter(fun(msg) -> log(msg, label="playlists_cleaner"),
    process.read.lines("find /tmp -iname beetsplaylist*m3u8 -mtime +0 -delete"),
  )
)