Re: Need someone testing XML file in Finale and Sibelius (and Dorico)

[Jacques Menu Muzhic]

Would it be a reasonable approach to have LilyPond first produce data in 
memory, i.e. Scheme data representing the whole score, and then either to 
produce the score or a MIDI file or supply this data to other applications, 
such as a MusicXML exporter?

Of course this would mean a lot of work.

> Le 7 mai 2018 à 18:18, Urs Liska <address@hidden> a écrit :
> 
> 
> 
> Am 07.05.2018 um 18:13 schrieb Jacques Menu Muzhic:
>> Hello Urs,
>> 
>> With the LP code:
>> 
>> {
>>   c'8 [ d' e' f' ]
>> }
>> 
>> what is the difficulty for producing <beam/> markups in the MusicXML output?
> 
> None. This works now (although the Pull Request hasn't been merged.
> 
> What does *not* work is producing beams in MusicXML from *that*:
> 
> {
>  c'8 d' e' f
> }
> 
> Since the implicit beams are not explicit in the input file python-ly doesn't 
> know about them.
> 
> Similarly:
> 
> {
>  c'8 [ d'16 e'16 ]
> }
> 
> Can't *correctly* be exported because MusicXML expects detailed information 
> about the beaming pattern (how many beams at each note), which again isn't 
> explicitly available in the LilyPond input.
> 
> To overcome this limitation python-ly would either have to reimplement all 
> the stuff the LilyPond parser does (think about what happens when we come to 
> Scheme functions :-O ) - or we find a way to have LilyPond parse the score 
> and pass the parsed expressions to some exporter module (which is the 
> approach Jan-Peter is exploring).
> 
> Urs
> 
>> 
>> JM
>> 
>>> Le 6 mai 2018 à 12:20, Urs Liska <address@hidden> a écrit :
>>> 
>>> 
>>> 
>>> Am 6. Mai 2018 12:08:39 MESZ schrieb Jacques Menu Muzhic <address@hidden>:
>>>> I’m not familiar with python.ly <http://python.ly/>’s internals: does
>>>> it use LP’s internal description, shown below by \displayMusic?
>>> No, it parses the input file(s) - which points to the single most 
>>> significant limitation of python-ly's/Frescobaldi's approach.
>>> 
>>>> Starting lilypond 2.19.80 [nobeam.ly]...
>>>> Processing
>>>> `/var/folders/jc/xrpy67_x6_vcjfzpzds_9_6m0000gn/T/frescobaldi-nx6e_mso/tmp2kwa263p/nobeam.ly'
>>>> Parsing...
>>>> Interpreting music...
>>>> Preprocessing graphical objects...
>>>> Finding the ideal number of pages...
>>>> Fitting music on 1 page...
>>>> Drawing systems...
>>>> Layout output to
>>>> `/var/folders/jc/xrpy67_x6_vcjfzpzds_9_6m0000gn/T//lilypond-pwPv1B'...
>>>> Converting to `nobeam.pdf'...
>>>> Deleting
>>>> `/var/folders/jc/xrpy67_x6_vcjfzpzds_9_6m0000gn/T//lilypond-pwPv1B'...
>>>> 
>>>> (make-music
>>>> 'SequentialMusic
>>>> 'elements
>>>> (list (make-music
>>>> 'NoteEvent
>>>> 'articulations
>>>> (list (make-music
>>>> 'BeamEvent
>>>> 'span-direction
>>>> -1))
>>>> 'duration
>>>> (ly:make-duration 3)
>>>> 'pitch
>>>> (ly:make-pitch 0 0))
>>>> (make-music
>>>> 'NoteEvent
>>>> 'pitch
>>>> (ly:make-pitch 0 1)
>>>> 'duration
>>>> (ly:make-duration 3))
>>>> (make-music
>>>> 'NoteEvent
>>>> 'pitch
>>>> (ly:make-pitch 0 2)
>>>> 'duration
>>>> (ly:make-duration 3))
>>>> (make-music
>>>> 'NoteEvent
>>>> 'articulations
>>>> (list (make-music
>>>> 'BeamEvent
>>>> 'span-direction
>>>> 1))
>>>> 'duration
>>>> (ly:make-duration 3)
>>>> 'pitch
>>>> (ly:make-pitch 0 3))))
>>>> 
>>>> Success: compilation successfully completed
>>>> Completed successfully in 0.8".
>>>> 
>>>>> Le 6 mai 2018 à 11:58, Urs Liska <address@hidden> a écrit :
>>>>> 
>>>>> 
>>>>> 
>>>>> Am 06.05.2018 um 11:43 schrieb Jacques Menu Muzhic:
>>>>>> Hello Urs,
>>>>>> 
>>>>>> musicsml2ly and Finale 2014 behave as expected, but MuseScore 2.2.1
>>>> produces beams in both cases, i.e. even when there are no beams in the
>>>> MusicXML data.
>>>>> OK, that seems that between 2.1.0 and 2.2.1 MuseScore has added that
>>>> feature.
>>>>> Still it's of course inacceptable not to export beams at all.
>>>>> 
>>>>> Unfortunately we can so far only deal with explicit beams, there's no
>>>> way (without substantial work to redo LilyPond's autobeaming (and
>>>> beaming pattern) in Python).
>>>>> Thanks for testing
>>>>> Urs
>>>>> 
>>>>>> Strange…
>>>>>> 
>>>>>> JM
>>>>>> 
>>>>>>> Le 6 mai 2018 à 11:09, Urs Liska <address@hidden> a écrit :
>>>>>>> 
>>>>>>> Hello Jacques,
>>>>>>> 
>>>>>>> 
>>>>>>> Am 06.05.2018 um 10:11 schrieb Jacques Menu Muzhic:
>>>>>>>> Hello Urs,
>>>>>>>> 
>>>>>>>> I’m doing tests with MusicXML, could you send me an example where
>>>> MuseScore doesn’t show any beams after exporting from LilyPond?
>>>>>>> Attached you'll find a simple nobeam.ly and the file nobeam.xml,
>>>> which is the result from exporting with the current master of
>>>> python-ly. This will be opened without beams in MuseScore 2.
>>>>>>> nobeam-withbeams.xml is the result of exporting from the branch
>>>> xml-handle-beams (my current pull request). The XML file I originally
>>>> sent was also created from that branch.
>>>>>>> Thanks for helping to look into it.
>>>>>>> Urs
>>>>>>> 
>>>>>>>> Thanks!
>>>>>>>> 
>>>>>>>> JM
>>>>>>>> 
>>>>>>>>> Le 5 mai 2018 à 11:50, address@hidden a écrit :
>>>>>>>>> 
>>>>>>>>> Hi Torsten,
>>>>>>>>> 
>>>>>>>>> 5. Mai 2018 11:44, "Torsten Hämmerle" <address@hidden>
>>>> schrieb:
>>>>>>>>>> Hi Urs,
>>>>>>>>>> 
>>>>>>>>>> Sibelius works fine:
>>>>>>>>>> 
>>>>>>>>>> 
>>>> <http://lilypond.1069038.n5.nabble.com/file/t3887/05-beam-sibelius.png>
>>>>>>>>> Thanks.
>>>>>>>>> 
>>>>>>>>>> As my Dorico evaluation period has expired, I can't even start
>>>> it.
>>>>>>>>>> All the others allow testing w/o save/export functionality, but
>>>> Steinberg is
>>>>>>>>>> much more restrictive in that respect.
>>>>>>>>>> 
>>>>>>>>> On the Facebook group I got a screenshot of Dorico, so (not
>>>> unexpectedly) this works too.
>>>>>>>>> So my conclusion is: we can export faulty XML, and Finale,
>>>> Sibelius, Dorico, MuseScore and Verovio will still properly handle it.
>>>>>>>>> I find that somewhat embarrassing, but I think it's much better
>>>> than nothing (I was pretty surprised to see that MusesScore didn't show
>>>> *any* beams after exporting a small example from LilyPond ...)
>>>>>>>>> Thanks to all who responded.
>>>>>>>>> Urs
>>>>>>>>> 
>>>>>>>>>> All the best,
>>>>>>>>>> Torsten
>>>>>>>>>> 
>>>>>>>>>> --
>>>>>>>>>> Sent from: http://lilypond.1069038.n5.nabble.com/User-f3.html
>>>>>>>>>> 
>>>>>>>>>> _______________________________________________
>>>>>>>>>> lilypond-user mailing list
>>>>>>>>>> address@hidden
>>>>>>>>>> https://lists.gnu.org/mailman/listinfo/lilypond-user
>>>>>>>>> _______________________________________________
>>>>>>>>> lilypond-user mailing list
>>>>>>>>> address@hidden
>>>>>>>>> https://lists.gnu.org/mailman/listinfo/lilypond-user
>>>>>>> <nobeam.ly><nobeam.xml><nobeam-withbeams.xml>
>