fixing #1189

[eroux]

This should fix #1189, currently only the TeX part is here

[henryso]

I added the C part. Running the tests, I think the results look as they should, given the code, but the fix-1192 test shows a weird edge case, where the clef change to a larger clef occurs on the last (in this case, the only) line, adding space to the first clef which really doesn't align it with anything below. I presume something similar would occur if the clef changes from C to F to C on the same line.

[eroux]

Thanks a lot! Indeed the current mechanism doesn't cover the kind of edge cases you're describing, but they should be very rare and easy to fix in gabc. I'll run the tests

[henryso]

Perhaps how to fix this edge case in gabc should be documented. I'll admit I have no idea how you'd do that.

[eroux]

Strange: trying to run the tests, I have a lot of message like

_: ligne 18 : printf: 0.999984: nombre non valable

and all tests fail, although the diff-*.png show no visible difference... any idea where this may come from?

[henryso]

Must be the test scripts. I'll look into it... Maybe a locale issue.

[henryso]

It's definitely a locale problem. compare is returning 0.999984 and your locale most likely is expecting 0,999984. I'll have to think about the best way to fix this one.

[henryso]

I set LC_ALL=C in gregorio-test.sh. Let me know if that helps.

[eroux]

that solved the problem, thanks!

[eroux]

Ok, I think this is ready to review

[henryso]

Since this is the user documentation, I would explicitly mention that this should be used in a <v>texverb</v> in gabc.

[henryso]

I can't review the C code since I wrote it, but I'm done reviewing the TeX code and documentation.

[eroux]

done, thanks

[henryso]

Looks fine to me. If you're OK with the C code, then I think this can be merged.

[eroux]

Thanks!