This is a prototype Python script which searches the ASCAP Repertory API to get a list of publishers.
I wrote this script in a few hours to see if it would be possible to have some sort of tool for content creators, which searches the publishers of the music they want to use in videos/on stream.
Before running this tool, it is important to have Python installed on your device. This can be done through any method of your choice. Personally I prefer to use Anaconda (or maybe Miniconda) when working on Windows, and the standard apt command-line package tool when working on Ubuntu.
The tool uses a few packages, most of them already included in the Python 3 Standard Library. The packages that need to be installed are pandas and requests, with the following commands.
pip install requests
pip install pandas
- Go to the TuneMyMusic website, and click on the Let's Start button.
- Choose the the source where your music is coming from, for this example I chose Spotify.
- On your Spotify playlist click the three dots and click Share > Copy link to playlist.
- Paste the link into the text box and click the Load playlist button.
- Once the playlist is loaded click the Next: Select Destination button.
- To select the destination, scroll down and click Export to file, this will give you an option to select format, it is important to choose CSV, and then click Select.
- Once this has been done, a summary page will be shown, with the chosen playlist and the text Spotify >> CSV below the summary heading. On this page, click the Start Moving My Music >> button.
- A download will start and the file My Spotify Playlist.csv should be saved on your device.
This tool is run via the command-line and includes an option to choose how to search for the songs and publishers.
usage: publisher_check.py [-h] -in INPUT_FILE -out OUTPUT_FILE
[-so {song-performer+song-only,song-performer}]
Run the ASCAP search script.
optional arguments:
-h, --help show this help message and exit
-in INPUT_FILE, --input_file INPUT_FILE
The path of the input CSV playlist file.
-out OUTPUT_FILE, --output_file OUTPUT_FILE
The path of the output CSV publishers file.
-so {song-performer+song-only,song-performer}, --search_options {song-performer+song-only,song-performer}
The type of search that is carried out (song-
performer-song-only - if the song-performer search
doesn't yield results, the song-only search is carried
out (can yield false positives); song-performer - only
searches song-performer combinations (more accurate
but can miss some results)) (default = ['song-
performer+song-only']).
For this option, the script first runs a query with song title and performer. This is the best way to make sure that the right song is found, but it can have some issues with certain characters in the song name (so far mainly the / character was found to be problematic).
If this search does not find any results, another search is carried out with only the song title. This tends to give a lot of results, so at the moment if more than one result is found, the script ignores the output. The song name only search can sometimes lead to false positives, so at the moment it might be best to not use it.
python publisher_check.py -in Playlist.csv -out PlaylistPublishers.csv -so song-performer+song-only
This option is very similar to the previous, but this only queries the song title and the performer. As mentioned, this tends to give fewer results but is somewhat more reliable.
python publisher_check.py -in Playlist.csv -out PlaylistPublishers.csv -so song-performer
After the script has run, make sure to open the output CSV file and go through the publishers that have been found. At this moment, it is important to manually go through the publishers and double check that these match with the ASCAP results (or any other repertory), I know this is a little counterintuitive, but as stated, this is only a prototype.
The script has not been thoroughly tested and might run into errors if the song name, performer or results are in a format which has not been encountered yet.
It has been stated clearly that the script is a basic prototype and is not always accurate, thus, you should take its results with a level of discretion.
If you would fully like to know the rights of a certain song and have not obtained results from this script, I advise you to manually check through the ASCAP, BMI and SESAC repertories, and see if any results come up on there. Keeping this in mind, these sources might still have issues with finding the song you are looking, so there is always a level of risk involved.
This project is not affiliated with any of the aforementioned repertories, and only makes use of available API links to gather information without having to type it manually.
At this moment, I cannot really afford to spend much time on this project, so I am leaving a list of ideas for anyone who would like to make a fork/pull-request and improve on this project.
- Adding a Graphical User Interface (GUI) to improve accessability.
- Transforming the script into a browser-based application (probably requires migration from Python to JavaScript).
- Improving the way the songs are searched, this might include:
- searching song name only then trying to match the performer;
- searching for the performer only then trying to match the song (should require less processing).
- Increasing the number of repertories that are searched. This might be difficult if the APIs are not easily available, but this should make the results more accurate.
- Adding some sort of multiprocessing to query multiple songs at once and speed-up the script.
- Most of these repertoires have their data publicly available for download It might be interesting to experiment with searching through a repertory file instead of sending requests over the internet. This method would require the repertory to be re-downloaded regularly, to keep up-to-date with any newly added songs.
- Adding a more fluid way to get songs from a playlist by only asking the user to input their Spotify (or any other platform) credentials, and choose the playlist from there. This might require quite a bit of time to work out how to information from such apps and is probably a very late stage addition.