Re: Accidental style in presence of mid-bar line breaks.

[Richard Shann]

On Mon, 2018-06-25 at 22:48 +0200, Simon Albrecht wrote:
> On 25.06.2018 13:25, Richard Shann wrote:
> > In 18th c. mss and prints it is quite common to create line breaks
> > during bars - I've found this much improves the legibility of some
> > florid slow movements.
> > 
> > However, it seems that the available accidental styles only take
> > account of barlines and ignore line breaks,
> 
> It’s not what you ask for, but of course in the 18th century
> accidentals 
> generally pertained only to one note and its immediate repetitions – 
> with exceptions according to common sense. Except for the exceptions 
> this can be modelled very well using a LilyPond accidental style –
> see 
> attachment.

Thank you very much for this code - it affords a good guide to the sort
of thing you can do to add accidental styles.  

As you observe it's not what I asked for, going rather in the direction
of mimicking old prints rather than easing the task for the modern
reader of the old music. (OT : I've had fun playing from facsimiles and
the biggest problems are not accidental style but the lack of alignment
of parts between barlines and - killer - the joining up of ledger lines
which makes the top line of the staff appear higher than it is.
Evidence that until the 19th c. musicians played by interval rather
than position - see https://www.jstor.org/stable/841141?refreqid=robots
txt-
sitemaps:b3d7f2442648678c96dec5651ef07226&seq=1#page_scan_tab_contents
for this old story).

Anyway, back to the topic: even if it's not possible to coax LilyPond
to flag up when it takes advantage of \bar "" to break a line, it would
help if it could emit a list of line-breaks - in what form I'm not sure
- that could be post-processed to determine if an a bar was split. (For
example if it could print the bar number before and after each line
break). 

In this connection, I noticed that LilyPond skips printing the bar
number at the beginning of each line if the bar is being split over a
line. This is better than the behavior you can see in another program
(Finale? Sib?) at 
https://imslp.org/wiki/Special:ImagefromIndex/323673
where the split bar gets assigned *two* bar numbers!

Thanks again for your engagement in this problem!

Richard