Can you add Musicbed.com support?

[becauseim]

SoudScrape works really great but it will be very good to add support of this website. Are you interested in it?

[Miserlou]

Never heard of that website. Would accept a well written, well tested PR though.

[becauseim]

Are you interested to write the code for downloading from Musicbed.com yourself?

[Miserlou]

Nah.

[brachna]

I tried to download a track just for the hell of it, but it's not that simple.
Tracks lie in the open (inside json), but downloading them - no luck.
Seems some kinda encryption.

[becauseim]

I download tracks from this site with help of following JS script, but it does't write title of the tracks:

App.components.SongRows.loadedSongs.forEach(function(song) {var temp = document.createElement('a'); temp.href = song.playback_url;temp.download=song.name; temp.click();});

[brachna]

Where do you execute such script? Python library? In a browser?

[becauseim]

In a browser.

[brachna]

What browser? I tried it in Firefox and using PyQt5.QWebEngineView. Didn't work in either.
Also tried just one track:
function test() {
var song = App.components.SongRows.loadedSongs[0];
var temp = document.createElement('a');
temp.href = song.playback_url;
temp.download=song.name;
temp.click();
}
test();

[becauseim]

Chrome.
Copy and paste. One line.

[brachna]

Confirm that Chrome works. Though at least in case of https://www.musicbed.com/albums/be-still/2828
it will only download first 30 seconds of complete song.
So I don't know, Chrome works, Firefox doesn't, but more importantly Python HTTP requests don't work.
I might give it a try with phantomjs.

[becauseim]

Reg and Log in.

[brachna]

Ok, but log in is not good for Soundscrape.
PhantomJS is out the window, because it's not supported on newer Python versions (at least I couldn't download it through pip). So javascript idea wouldn't work I'm afraid.
I tried to get a track using Python again, but stuck on "SignatureMismatch" error:

import json, requests


def main():
    exampleUrl = "https://www.musicbed.com/albums/be-still/2828"
    session = requests.Session()

##    # example structure
##    url = "https://s3.amazonaws.com/themusicbed/files/production/songs/73410/30-128-mp3_ffd0c5faf256a697a982d5e1417ff54c23bdf161.mp3"
##
##    details = { "X-Amz-Content-Sha256": "UNSIGNED-PAYLOAD",
##                "X-Amz-Algorithm": "AWS4-HMAC-SHA256",
##                "X-Amz-Credential": "AKIAJ4L3VOKMHHM4QLAA%2F20170105%2Fus-east-1%2Fs3%2Faws4_request",
##                "X-Amz-Date": "20170105T071222Z",
##                "X-Amz-SignedHeaders": "host",
##                "X-Amz-Expires": "604800",
##                "X-Amz-Signature": "eb4856af1da8bf92d38f4828b06eb7843370eeac285f33883eb9d932448532d1" }

    result = session.get( exampleUrl )
    
    jsonText = result.text.split( "App.components.SongRows = " )[1].split("</script>")[0]
    jsonData = json.loads( jsonText )
    url, data = jsonData["loadedSongs"][0]["playback_url"].split("?")

    details = dict()
    for each in data.split( "&" ):
        name, value = each.split( "=" )
        details.update( { name: value } )

    print( "url: " + url )
    print( str( details ) )

    # musicbed warns about it if it's not fixed
    details["X-Amz-Credential"] = details["X-Amz-Credential"].replace( "%2F", "/" )

    result = session.post( url, data=details )

    file = open( "test.txt", "w" )
    file.write( result.text )
    file.close()

if ( __name__ == "__main__" ):
    main()

Maybe someone more experienced will tackle this issue.

[brachna]

omg I just needed to change POST request to GET request.

[becauseim]

How can i test your code?

[brachna]

What code? You mean the one I posted above? That's just a draft, it doesn't even download a track.
Anyway, I figured out how to login and download using Python, no additional libraries needed.
Gotta tinker some more.