musicxml2ly enhancements

classic Classic list List threaded Threaded
18 messages Options
Reply | Threaded
Open this post in threaded view
|

musicxml2ly enhancements

Jacques Menu TvTMail
Hello folks,

I've been experimenting with Sibelius 7.1.3 as a front-end to Lily, i.e. as a note input engine.

This lead me to fix some issues in musicxml2ly:
        - the comment bar number at the end of regular lines is one too much;
        - it's missing when a \ barNumberCheck is present;
        - musicxml2ly crashes if there's no title specified in the XML file.

The first two points are important to me since they help locating the guilty lines (it's cumbersome to have to do the math each time you switch from an error message to the .ly file).

The experience is positive by the way, since Lily will pin-point errors not detected by Sibelius.

Compare the 2.16.2 original with the modified version used on the attached file:

<     \clef "treble" \key c \major \numericTimeSignature\time 4/4
<     \pageBreak a4 ( a4 ) ( b4 ) d4 | % 2
<     \key c \major a4 ( a4 ) ( b4 ) d4 | % 3
<     a4 ( a4 ) ( b4 ) d4 | % 4
<     g1 | % 5
<     a,4 ( a4 ) ( b4 ) d4 | % 6
<     g1 \break | % 7
<     a,4 ( a4 ) ( b4 ) d4 | % 8
<     g1 | % 9
<     a,4 ( a4 ) ( b4 ) d4 | \barNumberCheck #10
<     g1 | % 11
<     a,4 ( a4 ) ( b4 ) d4 | % 12
---

>     \clef "treble" \key c \major \numericTimeSignature\time 4/4 a4 ( a4
>     ) ( b4 ) d4 | % 1
>     \key c \major a4 ( a4 ) ( b4 ) d4 | % 2
>     a4 ( a4 ) ( b4 ) d4 | % 3
>     g1 | % 4
>     a,4 ( a4 ) ( b4 ) d4 | % 5
>     g1 | % 6
>     a,4 ( a4 ) ( b4 ) d4 | % 7
>     g1 | % 8
>     a,4 ( a4 ) ( b4 ) d4 | % 9
>     g1 | \barNumberCheck #11 % 10
>     a,4 ( a4 ) ( b4 ) d4 | % 11
I could spend time on the TODOs mentioned in the code, hence my question: is there work going on on musicxml2ly, and if so, can I contribute to it?

Hope this helps!




--

Jacques Menu
Ch. de la Pierre 12
1023 Crissier

mailto:[hidden email]





_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user

TieAndSlur.xml (26K) Download Attachment
dak
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

dak
Jacques Menu <[hidden email]> writes:

> Hello folks,
>
> I've been experimenting with Sibelius 7.1.3 as a front-end to Lily,
> i.e. as a note input engine.
>
> This lead me to fix some issues in musicxml2ly:
> - the comment bar number at the end of regular lines is one
> too much;

I don't see that.  It is the number that would be printed above the bar.

> - it's missing when a \ barNumberCheck is present;

Because then the bar number is given by the \barNumberCheck command.

> - musicxml2ly crashes if there's no title specified in the XML
> file.

That's certainly an error.

> The first two points are important to me since they help locating the
> guilty lines (it's cumbersome to have to do the math each time you
> switch from an error message to the .ly file).

Math?  You just look for the right bar number and then read on from
there.

> Compare the 2.16.2 original with the modified version used on the
> attached file:
>
> <     \clef "treble" \key c \major \numericTimeSignature\time 4/4
> <     \pageBreak a4 ( a4 ) ( b4 ) d4 | % 2
> <     \key c \major a4 ( a4 ) ( b4 ) d4 | % 3
> <     a4 ( a4 ) ( b4 ) d4 | % 4
> <     g1 | % 5
> <     a,4 ( a4 ) ( b4 ) d4 | % 6
> <     g1 \break | % 7
> <     a,4 ( a4 ) ( b4 ) d4 | % 8
> <     g1 | % 9
> <     a,4 ( a4 ) ( b4 ) d4 | \barNumberCheck #10
> <     g1 | % 11
> <     a,4 ( a4 ) ( b4 ) d4 | % 12

Looks consistent to me.

>>     a,4 ( a4 ) ( b4 ) d4 | % 9
>>     g1 | \barNumberCheck #11 % 10
>>     a,4 ( a4 ) ( b4 ) d4 | % 11

Not exactly an improvement here, is it?

> I could spend time on the TODOs mentioned in the code, hence my
> question: is there work going on on musicxml2ly, and if so, can I
> contribute to it?

Wouldn't it be even more important to contribute to it if there was _no_
work going on on musicxml2ly?

Currently there is not much happening with it.  The Philomelos guys have
their own fork of it, but so far not much has been reintegrated into
LilyPond IIRC.

--
David Kastrup


_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

Urs Liska-4
Am 09.04.2013 08:49, schrieb David Kastrup:

> Jacques Menu <[hidden email]> writes:
>
>> Hello folks,
>>
>> I've been experimenting with Sibelius 7.1.3 as a front-end to Lily,
>> i.e. as a note input engine.
>>
>> This lead me to fix some issues in musicxml2ly:
>> - the comment bar number at the end of regular lines is one
>> too much;
> I don't see that.  It is the number that would be printed above the bar.
>
>> - it's missing when a \ barNumberCheck is present;
> Because then the bar number is given by the \barNumberCheck command.
>
>> - musicxml2ly crashes if there's no title specified in the XML
>> file.
> That's certainly an error.
>
>> The first two points are important to me since they help locating the
>> guilty lines (it's cumbersome to have to do the math each time you
>> switch from an error message to the .ly file).
> Math?  You just look for the right bar number and then read on from
> there.
>
>> Compare the 2.16.2 original with the modified version used on the
>> attached file:
>>
>> <     \clef "treble" \key c \major \numericTimeSignature\time 4/4
>> <     \pageBreak a4 ( a4 ) ( b4 ) d4 | % 2
>> <     \key c \major a4 ( a4 ) ( b4 ) d4 | % 3
>> <     a4 ( a4 ) ( b4 ) d4 | % 4
>> <     g1 | % 5
>> <     a,4 ( a4 ) ( b4 ) d4 | % 6
>> <     g1 \break | % 7
>> <     a,4 ( a4 ) ( b4 ) d4 | % 8
>> <     g1 | % 9
>> <     a,4 ( a4 ) ( b4 ) d4 | \barNumberCheck #10
>> <     g1 | % 11
>> <     a,4 ( a4 ) ( b4 ) d4 | % 12
> Looks consistent to me.
>
>>>      a,4 ( a4 ) ( b4 ) d4 | % 9
>>>      g1 | \barNumberCheck #11 % 10
wouldn't it be better to write

g1 | % 10
\barNumberCheck #11


in such cases?

Urs


>>>      a,4 ( a4 ) ( b4 ) d4 | % 11
> Not exactly an improvement here, is it?
>
>> I could spend time on the TODOs mentioned in the code, hence my
>> question: is there work going on on musicxml2ly, and if so, can I
>> contribute to it?
> Wouldn't it be even more important to contribute to it if there was _no_
> work going on on musicxml2ly?
>
> Currently there is not much happening with it.  The Philomelos guys have
> their own fork of it, but so far not much has been reintegrated into
> LilyPond IIRC.
>


_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

pls-2
On 09.04.2013, at 09:01, Urs Liska <[hidden email]> wrote:

> Am 09.04.2013 08:49, schrieb David Kastrup:
>> Jacques Menu <[hidden email]> writes:
>>
>>> Hello folks,
>>>
>>> I've been experimenting with Sibelius 7.1.3 as a front-end to Lily,
>>> i.e. as a note input engine.
>>>
>>> This lead me to fix some issues in musicxml2ly:
>>> - the comment bar number at the end of regular lines is one
>>> too much;
>> I don't see that.  It is the number that would be printed above the bar.
>>
>>> - it's missing when a \ barNumberCheck is present;
>> Because then the bar number is given by the \barNumberCheck command.
>>
>>> - musicxml2ly crashes if there's no title specified in the XML
>>> file.
>> That's certainly an error.
Which has been fixed on https://github.com/Philomelos/lilypond-musicxml2ly-dev

>>
>>> The first two points are important to me since they help locating the
>>> guilty lines (it's cumbersome to have to do the math each time you
>>> switch from an error message to the .ly file).
>> Math?  You just look for the right bar number and then read on from
>> there.
>>
>>> Compare the 2.16.2 original with the modified version used on the
>>> attached file:
>>>
>>> <     \clef "treble" \key c \major \numericTimeSignature\time 4/4
>>> <     \pageBreak a4 ( a4 ) ( b4 ) d4 | % 2
>>> <     \key c \major a4 ( a4 ) ( b4 ) d4 | % 3
>>> <     a4 ( a4 ) ( b4 ) d4 | % 4
>>> <     g1 | % 5
>>> <     a,4 ( a4 ) ( b4 ) d4 | % 6
>>> <     g1 \break | % 7
>>> <     a,4 ( a4 ) ( b4 ) d4 | % 8
>>> <     g1 | % 9
>>> <     a,4 ( a4 ) ( b4 ) d4 | \barNumberCheck #10
>>> <     g1 | % 11
>>> <     a,4 ( a4 ) ( b4 ) d4 | % 12
>> Looks consistent to me.
>>
>>>>     a,4 ( a4 ) ( b4 ) d4 | % 9
>>>>     g1 | \barNumberCheck #11 % 10
> wouldn't it be better to write
>
> g1 | % 10
> \barNumberCheck #11
>
>
> in such cases?
>
> Urs
>
>
>>>>     a,4 ( a4 ) ( b4 ) d4 | % 11
>> Not exactly an improvement here, is it?
>>
>>> I could spend time on the TODOs mentioned in the code, hence my
>>> question: is there work going on on musicxml2ly, and if so, can I
>>> contribute to it?
>> Wouldn't it be even more important to contribute to it if there was _no_
>> work going on on musicxml2ly?
>>
>> Currently there is not much happening with it.  
Wrong.  I spent quite a considerable amount of time tracking down and analyzing bugs that lead to compilation failures.  These bugs have a high priority.  There are a lot more issues of the type "ugly".  Apart from this there are still many white gaps in musicxml2ly.  But we only have limited time to fix all of these problems...
>> The Philomelos guys have
>> their own fork of it, but so far not much has been reintegrated into
>> LilyPond IIRC.
We have fixed quite a few issues but there is still a lot to do.  Help would be very much appreciated!  I can offer to help with the reintegration of the code and to share our findings / bug reports.

hth
patrick
_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

Jacques Menu TvTMail
The math I mentioned is because I interpreted the bar number in:

a,4 ( a4 ) ( b4 ) d4 | % 6

as a help to locate the offending bar number

David, sorry I offended you about work going on. I've got only one experience in open software contributions so far, which lead to a dead-end: these guys didn't even answer to my proposal...

I'll be glad to help on musicxml2ly, with guidance from Patrick or others as to how to proceed.

Jacques

Le 9 avr. 2013 à 10:08:00, pls <[hidden email]> a écrit :

On 09.04.2013, at 09:01, Urs Liska <[hidden email]> wrote:

Am 09.04.2013 08:49, schrieb David Kastrup:
Jacques Menu <[hidden email]> writes:

Hello folks,

I've been experimenting with Sibelius 7.1.3 as a front-end to Lily,
i.e. as a note input engine.

This lead me to fix some issues in musicxml2ly:
- the comment bar number at the end of regular lines is one
too much;
I don't see that.  It is the number that would be printed above the bar.

- it's missing when a \ barNumberCheck is present;
Because then the bar number is given by the \barNumberCheck command.

- musicxml2ly crashes if there's no title specified in the XML
file.
That's certainly an error.
Which has been fixed on https://github.com/Philomelos/lilypond-musicxml2ly-dev

The first two points are important to me since they help locating the
guilty lines (it's cumbersome to have to do the math each time you
switch from an error message to the .ly file).
Math?  You just look for the right bar number and then read on from
there.

Compare the 2.16.2 original with the modified version used on the
attached file:

<     \clef "treble" \key c \major \numericTimeSignature\time 4/4
<     \pageBreak a4 ( a4 ) ( b4 ) d4 | % 2
<     \key c \major a4 ( a4 ) ( b4 ) d4 | % 3
<     a4 ( a4 ) ( b4 ) d4 | % 4
<     g1 | % 5
<     a,4 ( a4 ) ( b4 ) d4 | % 6
<     g1 \break | % 7
<     a,4 ( a4 ) ( b4 ) d4 | % 8
<     g1 | % 9
<     a,4 ( a4 ) ( b4 ) d4 | \barNumberCheck #10
<     g1 | % 11
<     a,4 ( a4 ) ( b4 ) d4 | % 12
Looks consistent to me.

   a,4 ( a4 ) ( b4 ) d4 | % 9
   g1 | \barNumberCheck #11 % 10
wouldn't it be better to write

g1 | % 10
\barNumberCheck #11


in such cases?

Urs


   a,4 ( a4 ) ( b4 ) d4 | % 11
Not exactly an improvement here, is it?

I could spend time on the TODOs mentioned in the code, hence my
question: is there work going on on musicxml2ly, and if so, can I
contribute to it?
Wouldn't it be even more important to contribute to it if there was _no_
work going on on musicxml2ly?

Currently there is not much happening with it.  
Wrong.  I spent quite a considerable amount of time tracking down and analyzing bugs that lead to compilation failures.  These bugs have a high priority.  There are a lot more issues of the type "ugly".  Apart from this there are still many white gaps in musicxml2ly.  But we only have limited time to fix all of these problems...
The Philomelos guys have
their own fork of it, but so far not much has been reintegrated into
LilyPond IIRC.
We have fixed quite a few issues but there is still a lot to do.  Help would be very much appreciated!  I can offer to help with the reintegration of the code and to share our findings / bug reports.

hth
patrick
_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user


Bien à toi,
Bien à vous,

--

Jacques Menu
Ch. de la Pierre 12
1023 Crissier





_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
dak
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

dak
Jacques Menu TvTMail <[hidden email]> writes:

> The math I mentioned is because I interpreted the bar number in:
>
> a,4 ( a4 ) ( b4 ) d4 | % 6
>
> as a help to locate the offending bar number…

Well, it is a help but probably not immediately what you'd like best.  I
think it is reasonable to have bar checks at the end of a line where
they don't distract from the content, and putting the comment containing
the bar number with it makes sense.  It is no accident that
\barNumberCheck fits well with that numbering.  Putting the bar number
at the start of the line would require

  %{ 5 %} a,4 ( a4 ) ( b4 ) d4 |

instead.

> David, sorry I offended you about work going on.

No offense was taken.  What might have looked like offense was just my
personal rendition of "playful banter".  I might be somewhat
overspecialized regarding the roles I play.

> I've got only one experience in open software contributions so far,
> which lead to a dead-end: these guys didn't even answer to my
> proposal...
>
> I'll be glad to help on musicxml2ly, with guidance from Patrick or
> others as to how to proceed.

We have a "contributor's guide" for LilyPond in the documentation, it
outlines the general procedures pretty well.

I doubt you'll get much tutoring regarding musicxml2ly itself.  Like
with many other areas of LilyPond, work being done on it is spotty
enough that one can't really talk about coordinated efforts.  That does
not mean that organizing a concerted effort might not be a good idea,
but it might require a better overview over potential players and the
involved code passages and how to best approach them (or replace them)
than just fixing a few of your personal pet peeves might.  So it is
probably easiest to just try starting to work on those, and look for the
greater picture once you get more of a feeling for how things work.

There is also a certain likelihood of getting the "these guys didn't
even answer to my proposal..." impression: a large ratio of currently
active developers is mainly focused on their own projects.  We have a
"bug squad" that tries to make sure that no contributions fall through
the cracks, but even if things are tracked in the issue tracker, at some
point of time somebody needs to pick them up.



--
David Kastrup


_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

Werner LEMBERG
> Putting the bar number at the start of the line would require
>
>   %{ 5 %} a,4 ( a4 ) ( b4 ) d4 |
>
> instead.

I like that.  It's tedious to type manually, but a computer program
doesn't know that word :-)


    Werner

_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

Urs Liska-4
Am 09.04.2013 11:38, schrieb Werner LEMBERG:

>> Putting the bar number at the start of the line would require
>>
>>    %{ 5 %} a,4 ( a4 ) ( b4 ) d4 |
>>
>> instead.
> I like that.  It's tedious to type manually, but a computer program
> doesn't know that word :-)
>
>
>      Werner
Would something like

%@5 a,4 ( a4 ) ( b4 ) d4 |

be an option?
The combination'%@' would indicate that the next token is meant as a
measure number (or any other pointer to the structure of the content).
An editor could use such entries (i.e. specific syntax, starting at the
beginning of a line) to create a document outline, e.g. point to @5
entries in all parts.

Of course I know that syntax changes shouldn't be taken lightly, though ...

Urs
>
> _______________________________________________
> lilypond-user mailing list
> [hidden email]
> https://lists.gnu.org/mailman/listinfo/lilypond-user


_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
dak
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

dak
Urs Liska <[hidden email]> writes:

> Am 09.04.2013 11:38, schrieb Werner LEMBERG:
>>> Putting the bar number at the start of the line would require
>>>
>>>    %{ 5 %} a,4 ( a4 ) ( b4 ) d4 |
>>>
>>> instead.
>> I like that.  It's tedious to type manually, but a computer program
>> doesn't know that word :-)
>>
>>
>>      Werner
> Would something like
>
> %@5 a,4 ( a4 ) ( b4 ) d4 |
>
> be an option?

In the same sense that drilling a hole in your knee and pouring in milk
is an option, definitely.

> The combination'%@' would indicate that the next token is meant as a
> measure number (or any other pointer to the structure of the content).

Except that inside of comments, there is no such thing as a "token".
Which is part of a comment's appeal.  And a comment that is a "pointer
to the structure of the content"?  I don't see how a comment be any such
thing, but then I have a problem parsing this parenthetical addition
anyway.

> An editor could use such entries (i.e. specific syntax, starting at
> the beginning of a line) to create a document outline, e.g. point to
> @5 entries in all parts.

You lost me _way_ back.

> Of course I know that syntax changes shouldn't be taken lightly,
> though ...

There is nobody keeping you to put whatever editor-hinting comments into
your LilyPond source file using the existing comment syntax.  I don't
see what another comment syntax would buy you in that regard, though.

--
David Kastrup


_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

Thomas Morley
2013/4/9 David Kastrup <[hidden email]>:

> In the same sense that drilling a hole in your knee and pouring in milk
> is an option, definitely.

ROFLMAO

_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

Urs Liska-4
In reply to this post by dak
OK. Please consider my suggestion as obsolete, but I'd like to clarify  
what I intended to say (but obviously failed to do so)

Quoting David Kastrup <[hidden email]>:

> Urs Liska <[hidden email]> writes:
>
>> Would something like
>>
>> %@5 a,4 ( a4 ) ( b4 ) d4 |
>>
>> be an option?
>>
>> The combination'%@' would indicate that the next token is meant as a
>> measure number (or any other pointer to the structure of the content).
>
> Except that inside of comments, there is no such thing as a "token".
> Which is part of a comment's appeal.  And a comment that is a "pointer
> to the structure of the content"?  I don't see how a comment be any such
> thing, but then I have a problem parsing this parenthetical addition
> anyway.
What I thought of was that
   %@
would assume an implicit closing tag after the next white space, so  
you can write
   %@5_ (the underscore meaning a space character)
instead of
   %{ 5 %}
so that the rest of the line is considered as non-comment.

That probably was a silly idea because its only real advantage is some  
slightly less typing.

Of course it was misleading to use 'keywords' like pointer etc. I  
meant special comments that indicate a measure number or any other  
indicator of a musical structure (e.g. partI, verse, coda etc.)

>
>> An editor could use such entries (i.e. specific syntax, starting at
>> the beginning of a line) to create a document outline, e.g. point to
>> @5 entries in all parts.
>
> You lost me _way_ back.
>
>> Of course I know that syntax changes shouldn't be taken lightly,
>> though ...
>
> There is nobody keeping you to put whatever editor-hinting comments into
> your LilyPond source file using the existing comment syntax.  I don't
> see what another comment syntax would buy you in that regard, though.
You're probably right.
Everything I could want could also be achieved with the existing  
command syntax.

Urs

>
> --
> David Kastrup
>
>
> _______________________________________________
> lilypond-user mailing list
> [hidden email]
> https://lists.gnu.org/mailman/listinfo/lilypond-user




_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

janek.lilypond
In reply to this post by Werner LEMBERG
2013/4/9 Werner LEMBERG <[hidden email]>:
>> Putting the bar number at the start of the line would require
>>
>>   %{ 5 %} a,4 ( a4 ) ( b4 ) d4 |
>>
>> instead.
>
> I like that.  It's tedious to type manually, but a computer program
> doesn't know that word :-)

Indeed, but i find it unreadable, especially if it was placed at every line.
I think that the best way to mark bar numbers in source is to make a
separate-line comment every 5 or so measures:

>     \clef "treble" \key c \major \numericTimeSignature \time 4/4
>     a4 ( a4) ( b4 ) d4 |
>     \key c \major a4 ( a4 ) ( b4 ) d4 |
>     a4 ( a4 ) ( b4 ) d4 |
>     g1 |
>     % m.5
>     a,4 ( a4 ) ( b4 ) d4 |
>     g1 |
>     a,4 ( a4 ) ( b4 ) d4 |
>     g1 |
>     a,4 ( a4 ) ( b4 ) d4 |
>     % m.10
>     g1 |
>     a,4 ( a4 ) ( b4 ) d4 |

etc.
Janek

_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

Urs Liska-4
Am 09.04.2013 13:19, schrieb Janek Warchoł:

> 2013/4/9 Werner LEMBERG <[hidden email]>:
>>> Putting the bar number at the start of the line would require
>>>
>>>    %{ 5 %} a,4 ( a4 ) ( b4 ) d4 |
>>>
>>> instead.
>> I like that.  It's tedious to type manually, but a computer program
>> doesn't know that word :-)
> Indeed, but i find it unreadable, especially if it was placed at every line.
> I think that the best way to mark bar numbers in source is to make a
> separate-line comment every 5 or so measures:
>
>>      \clef "treble" \key c \major \numericTimeSignature \time 4/4
>>      a4 ( a4) ( b4 ) d4 |
>>      \key c \major a4 ( a4 ) ( b4 ) d4 |
>>      a4 ( a4 ) ( b4 ) d4 |
>>      g1 |
>>      % m.5
>>      a,4 ( a4 ) ( b4 ) d4 |
>>      g1 |
>>      a,4 ( a4 ) ( b4 ) d4 |
>>      g1 |
>>      a,4 ( a4 ) ( b4 ) d4 |
>>      % m.10
>>      g1 |
>>      a,4 ( a4 ) ( b4 ) d4 |
> etc.
> Janek
I have come to the conclusion that it is a _very_ good idea to make a
measure number for _every_ measure.
That way you have a measure indicator in each Git diff.
Which proves useful to map the change to the location in the file if the
line numbers have changed considerably in the meantime.

Urs
>
> _______________________________________________
> lilypond-user mailing list
> [hidden email]
> https://lists.gnu.org/mailman/listinfo/lilypond-user


_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

RE: musicxml2ly enhancements

Curt McDowell
How about allowing bar numbers as a letter+number, parsed similar to musical note+duration:
        B78 a,4 ( a4 ) ( b4 ) d4 |
        B79 e,4 ( f4 ) ( g4 ) a4 |
Perhaps they could also double as bar checks:
        B78 a,4 ( a4 ) ( b4 ) d4
        B79 e,4 ( f4 ) ( g4 ) a4
Or alternately, augment the existing bar check to allow an optional value:
        a,4 ( a4 ) ( b4 ) d4 |78
        e,4 ( f4 ) ( g4 ) a4 |79
Or (to taste):
        |78  a,4 ( a4 ) ( b4 ) d4
        |79  e,4 ( f4 ) ( g4 ) a4
The actual bar numbers might be ignored initially, but eventually
- they might be the numbers printed
- bar checks might also verify that bar numbers match across staves, voices, \alternatives.

Having already entered some large pieces, I only recently discovered Frescobaldi. What a wonderful program. The point-and-click cursor matching between source and PDF nearly obviates the need for measure numbers. I wish I'd found it sooner!

Cheers,
-Curt

-----Original Message-----
From: lilypond-user-bounces+lilypond=[hidden email] [mailto:lilypond-user-bounces+lilypond=[hidden email]] On Behalf Of Urs Liska
Sent: Tuesday, April 09, 2013 4:26 AM
To: [hidden email]
Subject: Re: musicxml2ly enhancements

Am 09.04.2013 13:19, schrieb Janek Warchoł:

> 2013/4/9 Werner LEMBERG <[hidden email]>:
>>> Putting the bar number at the start of the line would require
>>>
>>>    %{ 5 %} a,4 ( a4 ) ( b4 ) d4 |
>>>
>>> instead.
>> I like that.  It's tedious to type manually, but a computer program
>> doesn't know that word :-)
> Indeed, but i find it unreadable, especially if it was placed at every line.
> I think that the best way to mark bar numbers in source is to make a
> separate-line comment every 5 or so measures:
>
>>      \clef "treble" \key c \major \numericTimeSignature \time 4/4
>>      a4 ( a4) ( b4 ) d4 |
>>      \key c \major a4 ( a4 ) ( b4 ) d4 |
>>      a4 ( a4 ) ( b4 ) d4 |
>>      g1 |
>>      % m.5
>>      a,4 ( a4 ) ( b4 ) d4 |
>>      g1 |
>>      a,4 ( a4 ) ( b4 ) d4 |
>>      g1 |
>>      a,4 ( a4 ) ( b4 ) d4 |
>>      % m.10
>>      g1 |
>>      a,4 ( a4 ) ( b4 ) d4 |
> etc.
> Janek
I have come to the conclusion that it is a _very_ good idea to make a measure number for _every_ measure.
That way you have a measure indicator in each Git diff.
Which proves useful to map the change to the location in the file if the line numbers have changed considerably in the meantime.

Urs
>
> _______________________________________________
> lilypond-user mailing list
> [hidden email]
> https://lists.gnu.org/mailman/listinfo/lilypond-user


_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user


_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

RE: musicxml2ly enhancements

m.tarenskeen


>>>>    %{ 5 %} a,4 ( a4 ) ( b4 ) d4 |
>>>>
>>>> instead.
>>> I like that.  It's tedious to type manually, but a computer program
>>> doesn't know that word :-)
>> Indeed, but i find it unreadable, especially if it was placed at every line.

In general I think tools like musicxml2ly and other programs that convert
to or export .ly files (xxx2ly, mscore, denemo, rosegarden) should do
their best to make the lilypond output look like it was handwritten by a
"real" human lilypond user and likewise should be easily readable and
editable by a real human. With proper formatting, indentation, and coding
style.

I don't like the way it going for example in today's world of webdesign:
There are advanced tools to create professional websites "without knowing
anything about HTML and CSS". But if you want to edit the html code these
tools produce by hand it can be a real nightmare.

Just like my lilypond scores I often still write (simple) html code using
just Vim. Maybe I'm just getting old and old fashioned :-)

--

MT


_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

Werner LEMBERG

>>>>>    %{ 5 %} a,4 ( a4 ) ( b4 ) d4 |
>>>>
>>>> I like that.  It's tedious to type manually, but a computer
>>>> program doesn't know that word :-)
>>>
>>> Indeed, but i find it unreadable, especially if it was placed at
>>> every line.

Really?  It's like having a BASIC program which always starts with a
line number:

  %{  1 %}  a,4 ( a4 ) ( b4 ) d4 |
  %{  2 %}  a,4 ( a4 ) ( b4 ) d4 |
  %{  3 %}  a,4 ( a4 ) ( b4 ) d4 |
  %{  4 %}  a,4 ( a4 ) ( b4 ) d4 |
  %{  5 %}  a,4 ( a4 ) ( b4 ) d4 |
  %{  6 %}  a,4 ( a4 ) ( b4 ) d4 |
  %{  7 %}  a,4 ( a4 ) ( b4 ) d4 |
  %{  8 %}  a,4 ( a4 ) ( b4 ) d4 |
  %{  9 %}  a,4 ( a4 ) ( b4 ) d4 |
  %{ 10 %}  a,4 ( a4 ) ( b4 ) d4 |

As mentioned earlier, I don't like to type this, but having such a
format returned by musicxml2ly is good IMHO.  After some time you get
used to it, and line numbers are only looked at if you really need
them.

> In general I think tools like musicxml2ly and other programs that
> convert to or export .ly files (xxx2ly, mscore, denemo, rosegarden)
> should do their best to make the lilypond output look like it was
> handwritten by a "real" human lilypond user and likewise should be
> easily readable and editable by a real human.  With proper
> formatting, indentation, and coding style.

I fully agree.


    Werner

_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

Wim van Dommelen-4
In reply to this post by m.tarenskeen


On 10 Apr 2013, at 11:38 , Martin Tarenskeen wrote:

Just like my lilypond scores I often still write (simple) html code using just Vim. Maybe I'm just getting old and old fashioned :-)

+1, we're not alone......


Regards,
Wim.


_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user
Reply | Threaded
Open this post in threaded view
|

Re: musicxml2ly enhancements

Urs Liska-4
In reply to this post by Werner LEMBERG
Am Mittwoch, den 10.04.2013, 12:28 +0200 schrieb Werner LEMBERG:

> >>>>>    %{ 5 %} a,4 ( a4 ) ( b4 ) d4 |
> >>>>
> >>>> I like that.  It's tedious to type manually, but a computer
> >>>> program doesn't know that word :-)
> >>>
> >>> Indeed, but i find it unreadable, especially if it was placed at
> >>> every line.
>
> Really?  It's like having a BASIC program which always starts with a
> line number:
>
>   %{  1 %}  a,4 ( a4 ) ( b4 ) d4 |
>   %{  2 %}  a,4 ( a4 ) ( b4 ) d4 |
>   %{  3 %}  a,4 ( a4 ) ( b4 ) d4 |
>   %{  4 %}  a,4 ( a4 ) ( b4 ) d4 |
>   %{  5 %}  a,4 ( a4 ) ( b4 ) d4 |
>   %{  6 %}  a,4 ( a4 ) ( b4 ) d4 |
>   %{  7 %}  a,4 ( a4 ) ( b4 ) d4 |
>   %{  8 %}  a,4 ( a4 ) ( b4 ) d4 |
>   %{  9 %}  a,4 ( a4 ) ( b4 ) d4 |
>   %{ 10 %}  a,4 ( a4 ) ( b4 ) d4 |
>
> As mentioned earlier, I don't like to type this, but having such a
> format returned by musicxml2ly is good IMHO.  After some time you get
> used to it, and line numbers are only looked at if you really need
> them.

You could define a snippet in Frescobaldi with
%{ $SELECTION %}|
as its content (note the bar check and the empty space at the end) and
assign a shortcut.

With some more effort one could create a Python snippet that
- takes a single integer on a line
- places this comment around it and
- right-aligns the number to three digits

If one then added for example an '@' after the opening comment one could
maybe talk Wilbert into making use of this for Frescobaldi's document
outline.

Urs


_______________________________________________
lilypond-user mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/lilypond-user