Re: XML to .ly and Lilypond, again

[Phil Holmes]

----- Original Message ----- 
From: "Leszek Wronski" <address@hidden>
To: <address@hidden>
Sent: Friday, May 12, 2017 12:54 PM
Subject: Re: XML to .ly and Lilypond, again


> Dear Phil,
>
> I read through the 2 pages of comments at the link you provided. Is
> the ultimate suggestion the one in your last comment, that is, "I
> believe this patch requires Python 2.7, and we ship 2.4. As a result,
> windows users can no longer use musicxml2ly." ?
>
> If so, I should just mention that I'm getting the errors on an Ubuntu
> 16.04 system with Python version 2.7.12.
>
> Cheers,
>
> Leszek.

OK - I've done some more looking at this.  Using a test file produced by 
Sibelius, I get no problems using musicxml2ly on Ubuntu 14.04, Python 2.7.6. 
I think it would be helpful if people with other Ubuntu/Pytho versions could 
check an XML file (I can email my test to anyone who can check) to see 
whether the incorrect encoding of text is happening.

If I try the same file on Windows, with LilyPond 2.19.43, all is well.  With 
2.19.44 I get the same errors as Leszek:

Traceback (most recent call last):
 File "C:\Program Files (x86)\LilyPondV2.19.44\usr\bin\musicxml2ly.py", 
line 3269, in ?
   main()
 File "C:\Program Files (x86)\LilyPondV2.19.44\usr\bin\musicxml2ly.py", 
line 3263, in main
   voices = convert(filename, options)
 File "C:\Program Files (x86)\LilyPondV2.19.44\usr\bin\musicxml2ly.py", 
line 3175, in convert
   score_information.print_ly(printer)
 File "out/musicexp.py", line 855, in print_ly
 File "out/musicexp.py", line 848, in format_header_strings
 File "out/musicexp.py", line 123, in newline
 File "C:\Program Files 
(x86)\LilyPondV2.19.43\usr\lib\python2.4\codecs.py", line 501, in write
   return self.writer.write(data)
 File "C:\Program Files 
(x86)\LilyPondV2.19.43\usr\lib\python2.4\codecs.py", line 178, in write
   data, consumed = self.encode(object, self.errors)
UnicodeDecodeError: 'utf8' codec can't decode byte 0xa9 in position 39: 
unexpected code byte

However, if I try a file with a simpler header (e.g. 
02d-Rests-Multimeasure-TimeSignatures.xml in the regtest suite), then the 
script runs successfully, but I get an output file with spurious nulls in 
text fields:

\header {
texidoc =
"&#0;M&#0;u&#0;l&#0;t&#0;i&#0;-&#0;M&#0;e&#0;a&#0;s&#0;u&#0;r&#0;e&#0; 
&#0;r&#0;e&#0;s&#0;t&#0;s&#0; &#0;s&#0;h&#0;o&#0;u&#0;l&#0;d&#0; 
&#0;a&#0;l&#0;w&#0;a&#0;y&#0;s&#0; &#0;
&#0; &#0; &#0; &#0; &#0; &#0; &#0; &#0; &#0; &#0; &#0;b&#0;e&#0; 
&#0;c&#0;o&#0;n&#0;v&#0;e&#0;r&#0;t&#0;e&#0;d&#0; &#0;i&#0;n&#0;t&#0;o&#0; 
&#0;d&#0;u&#0;r&#0;a&#0;t&#0;i&#0;o&#0;n&#0;s&#0; &#0;t&#0;h&#0;a&#0;t&#0; 
&#0;a&#0;r&#0;e&#0; &#0;a&#0; &#0;m&#0;u&#0;l&#0;t&#0;i&#0;p&#0;l&#0;e&#0; 
&#0;o&#0;f&#0; &#0;t&#0;h&#0;e&#0; &#0;t&#0;i&#0;m&#0;e&#0; &#0;
&#0; &#0; &#0; &#0; &#0; &#0; &#0; &#0; &#0; &#0; 
&#0;s&#0;i&#0;g&#0;n&#0;a&#0;t&#0;u&#0;r&#0;e&#0;.&#0;"
&#0;
}

So it looks clear that something probably introduced in 
https://sourceforge.net/p/testlilyissues/issues/4751/ is causing encoding 
problems on Windows.  As I say, I can't reproduce the problem on Linux.

--
Phil Holmes