David has pointed out to me that what was tripping me up with this file was a CR character (0x0D) which is white space and is treated as such, but editors should not treat it as starting a line, which the GtkSourceView widget was doing. So looked at in (for example) gedit, the line count was wrong, but only because it treated an isolated CR as starting a new line.
Richard On Fri, 2014-10-24 at 21:02 +0200, Simon Albrecht wrote: > Am 24.10.2014 um 13:59 schrieb Richard Shann: > > On Fri, 2014-10-24 at 11:15 +0200, Helge Kruse wrote: > > Thank you for your response. > >> I think you should not added control characters like newline to the > >> markup text. > > yes, perhaps that is the case - I assumed any white space was > > acceptable, but it seems in some circumstances to trigger this undesired > > behavior. > >> When you want to have more than one line, > > that was not the intention here - the intention was just to make the > > markup source code easier to read. The original example where I hit this > > "feature" involved markup with fret diagrams, where it is natural to > > want to lay out the markup on several lines, and indeed this sometimes > > works without side effects. > > > > If this curious side effect from using different types of white space is > > difficult to avoid in the LilyPond compiler then I can easily live with > > it. > Well, I think you are right in expecting that any number of superfluous > whitespace and newline characters should not influence output (except > inside strings, of course) or in any way disturb Lily. So this should > definitely be discussed. > > Best, Simon _______________________________________________ lilypond-user mailing list lilypond-user@gnu.org https://lists.gnu.org/mailman/listinfo/lilypond-user
