musdl 3.0.0

[mus]escore [d]own[l]oader, ported from TypeScript

musdl

NOTE

The author of musdl does NOT condone piracy in any way, and is not responsible for anything that happens as a result of piracy arising from the use of musdl.

musdl (musscore downloader) is a downloader for Musescore, written in Python. musdl can now only download scores as .mscz (use the Musescore software to export to other formats).

The easiest way to download the score is through the CLI:

musdl (musescore url here)

But, it can also be imported and used as a module:

from musdl import Score, OnlineScore

my_score = OnlineScore("musescore_url")

# Read the score's metadata,
name = my_score["workTitle"]
# or save the whole score...
my_score.save_to("my_score.mscz")

# ...and then load it again.
my_score = Score.from_file("my_score.mscz")

How it works

I just copied over the IPNS-specific constants and used them to download the score file from Xmader's dataset.

Other approaches were considered:

• Create a native Python interpreter just to decode the musescore js (like Js2py). (Will problably break with every API change.)
• Use a headless JS browser with Selenium. (Resource-heavy, not really cross-platform.)
• Run a JS engine like PyMiniRacer/PyQt5. (Better speed, but needs C extensions.)

So I decided to roll with the IPNS dataset.

takedown request, et tu

Take a look at this issue in the same repo.

Requirements

• requests - Downloader.
• beautifulsoup4 - Powerful HTML parser.

Hacking

All my python projects now use flit to build and publish. So you need to python3 -m pip install flit first.

flit build

Install

(python3 -m) pip install musdl

License

MIT.

Changelog

3.0.0

Complete re-write of musdl to use IPNS instead of relying on MuseScore's API, which can break anytime.

2.3.0

Better PDF page detection (now the exact number of pages are known at runtime). The attributes .user, .id, and .pages have been added to the Score class. (user and pages are self-explanatory, id is the score id.) Added support for multi-threaded downloading of PDF pages (specify with -t). Downloading (may) be faster.

2.2.2

Fixed PDF scaling issue on Windows.

2.2.1

Made PDF downloading a extra feature (because svglib and reportlab dependencies are not pure-Python).

2.2.0

Fixed PDF support for Windows. Now musdl uses svglib as the backend for adding SVGs to PDFs.

2.1.1-2.1.2

Added PDF support.

2.0.0

Added -O option to use the score's title as the output filename (y'know, like curl -O). Refactored code to be object-oriented, in the form of the Score class. This should make adding new features easier. (Plus, you can use musdl more programmatically!)

1.1.2

Version bump, use Flit to build instead of setup.py. Updated README too.

1.1.0

Added MusicXML download option, thanks to JPFrancoia for the update!

1.0.0

Initial version.