Re: keep LilyPond process running in sort of "server mode"

[Wilbert Berendsen]

Op Do, 5 augustus, 2010 13:01, I wrote:
> Hi,
>
> some time ago there has been some discussion about a way to run LilyPond
> once and keep it running, waiting for a filename to compile. I think it
> was a special .ly file containing Scheme code that was started and ran all
> the init stuff, then waited for the name of a .ly file to process (via a
> socket, pipe or stdin, don't remember).
>
> After processing the file, the process would remain running and waiting
> for another filename to process. Just like specifying multiple files on
> the lilypond command line, all init files are read only once. This could
> speed up repeated compiles of LilyPond files inside (e.g.) Frescobaldi.

Well, I found out about ly:parse-file which seems to be able to do the
job. Now on to implement a simple protocol to talk to a running Scheme
procedure which reads a filename and calls ly:parse-file on the file and
then reports back a "failed" or "succeeded" status... I probably should
use a socket or a named pipe...

With best regards,
Wilbert Berendsen

Frescobaldi, LilyPond editor for KDE: http://www.frescobaldi.org/