Added seven neumes to gregall

[xmarteo]

Added toS!cl-lsi8 from Hartker, and derived toS!cl- and toS!cl from it.
Added ql-ppt1sut2su1 from Hartker.
Derived to!cl> from to!cl->.
Please review the following facts:

• The list of ms sources in GregorioNabcRef.pdf looks broken, but just as broken as before - maybe this is my setup of MikTeX.
• The SFD diff is abnormally large, as if I had "touched" dozens of background images.

[rpspringuel]

I know our nabc coverage is severely lacking in the gregorio-test, but could you make some tests for these new glyphs so that we at least get started on testing the nabc glyphs?

[rpspringuel]

If there are no objections this week, I'm going to accept and merge this PR despite the lack of tests.

[xmarteo]

Please don't. I am not happy with the results and I beg @jakubjelinek to have a look at it first.

[jakubjelinek]

Sorry for not getting into this earlier.
My primary issue with this patch is that it clearly is created by different version of the fontforge than I used so far (see the first line change in the sfd:
-SplineFontDB: 3.0
+SplineFontDB: 3.2
and contains a lot of changes to existing characters.
+InvalidEm: 0
-Layer: 0 0 "Back" 1
-Layer: 1 1 "Fore" 0
+Layer: 0 1 "Back" 1
+Layer: 1 1 "Fore" 1
in the file header I assume (etc.), then in clB, and then hundreds of other characters.
So essentially rewrites the whole file, it is unclear what actually changed and what has not.
Compare that to the earlier updates to the font file, which looked like:
--- a/fonts/gregall.sfd
+++ b/fonts/gregall.sfd
@@ -20,7 +20,7 @@ OS2Version: 4
OS2_WeightWidthSlopeOnly: 0
OS2_UseTypoMetrics: 1
CreationTime: 1388509729
-ModificationTime: 1677262955
+ModificationTime: 1677343725
PfmFamily: 17
TTFWeight: 400
TTFWidth: 5
@@ -745,11 +745,11 @@ DisplaySize: -96
AntiAlias: 1
FitToEm: 1
ExtremaBound: 10
-WinInfo: 58862 19 8
+WinInfo: 58900 19 8
BeginPrivate: 0
EndPrivate
TeXData: 1 0 0 346030 173015 115343 717226 1048576 115343 783286 444596 497025 792723 393216 433062 380633 303038 157286 324010 404750 52429 2506097 1059062 262144
-BeginChars: 65538 565
+BeginChars: 65538 595

StartChar: space
Encoding: 32 32 0
and then added a bunch of new stuff at the end of file, nothing else. That helps with tracking that nothing else has really changed.
So, e..g. last committed change was
git diff 5745141^ 5745141 gregall.sfd | diffstat
gregall.sfd |27486 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++-
1 file changed, 27483 insertions(+), 3 deletions(-)
which added 30 new characters and didn't change anything else, while this one
gregall.sfd |386880 +++++++++++++++++++++++++++++++++++++++++++++++++++---------------------------------------------------------------------------------------------------------------
1 file changed, 123707 insertions(+), 263173 deletions(-)
even when it is supposed to add just 7 neumes.
I'll try to manually tweak it so that it just adds the new neumes and whether my fontforge can grok it at all now.

[jakubjelinek]

And the fontforge I have crashes on the added characters. The main reason I'm still on the old fontforge is because they've changed the UI for creating splines and it creates different style looking characters from the old one.
I'll try to redo the 7 neumes myself from just the mentioned locations.

[xmarteo]

The mess with the rest of the file is clearly caused by the version discrepancy, of which I was unaware. I could try and install the old fontforge along the new (what version is yours?) and copy-paste the characters from one to the other. Do you think it would be viable?

[jakubjelinek]

From the *.tex files, seems it is just 4 new glyphs and then 3 glyphs done by removing some part, I'll just try to do those myself, shouldn't take that long.
Anyway, for further glyphs, my fontforge says
Copyright (c) 2000-2012 by George Williams.
Executable based on sources from 14:57 GMT 31-Jul-2012.
Library based on sources from 14:57 GMT 31-Jul-2012.
I don't remember when exactly was the UI rewrite, but I think soon after this. And unfortunately at least the newer fontforge versions I've tried didn't have any way to use the old way of doing the splines.

[jakubjelinek]

Regarding the neume naming, the neume in https://www.e-codices.unifr.ch/en/csg/0391/204 certainly doesn't look like toS (the crooked torculus extra slow form), but either just to or maybe toS1.
Now, the font already has to!cl and to!cl- neumes and so far I just didn't make distinction between those in the combined forms. E.g. see to!cl in the font which really looks like having the to form of the torculus first, while to!cl-!cl- or to!po- which look more like toS1!cl-!cl- or toS1!po.
So, do you think we need to introduce toS1!cl- and toS1!cl neumes? If so, I'd prefer to use as the non-episemed to use the one in the same responsorium next word.
For pqS- I agree and will add now. Ditto ql!cl>1.
Regarding to!cl>, my preference would be not remove episema clearly connected from the rest of glyph, but rather find where in the manuscript you have a glyph without the episema. Where have you used this?

[jakubjelinek]

So, I have
jakubjelinek@c1afdb6
so far (added pqS-, ql!cl>1, ql-ppt1sut2su1 (and what wasn't in the original also ql-sut2su1).
For toS!cl- etc. vs. toS1!cl- vs. to!cl- I'm waiting for feedback above, for to!cl> I'd really like to hear where you saw it in the mss, so that we can use a glyph directly.

[xmarteo]

to!cl> : https://www.e-codices.unifr.ch/en/csg/0391/213 R/ Peto domine, "hUius" (first line)

with respect to to!cl vs toS!cl, well, I am not a professional semiologist, but I tend to distinguish between cases where the articulation between the first two notes of a torculus is very round (the first note and the articulation forming, together, half of a circle, as in the usual 'to' and 'toS1'), and cases where the first note is a straight dash with an inflexion point, followed by a more angled articulation (like in gregall's pe>1). I think the glyphs in https://www.e-codices.unifr.ch/en/csg/0391/204 , R/ Emitte domine, which you saw, belong to the second kind, where the first two notes really look like the beginning of pe>1 (or toS for that matter).