Ben Harris
d85ec66568
Simon had the name "Bedstead" lying around doing nothing, so I've swiped
...
it for this one as an improvement on "TTXT".
2009-10-11 23:17:04 +01:00
Ben Harris
cf0d1e0429
I don't think there's any need for my "palt" lookup to restrict itself from
...
considering ligatures. For one thing, I don't have any. Why FontForge
chose that as its default is anybody's guess.
2009-10-11 23:14:55 +01:00
Ben Harris
8870351a3c
Add a 'palt' feature providing alternative metrics for proportional
...
spacing. They're generated automatically (of course), but they work
quite well. Also apply what I think is a correction to the script
lists for 'salt' and 'aalt'.
2009-10-11 10:29:11 +01:00
Ben Harris
7789d32d30
Add a 'salt' feature for the characters that vary between chips, though
...
I wonder if it really ought to be 'locl'.
2009-10-10 23:11:53 +01:00
Ben Harris
6a529f89e2
Pull out creation of lookups into its own function, since main() was
...
getting a bit big.
2009-10-10 18:05:11 +01:00
Ben Harris
5720bcd9c3
A few additional glyphs.
2009-10-10 13:23:47 +01:00
Ben Harris
a4b9779941
Add flags and ways to pass them around so that the character renderer can
...
take into account the joining behaviour of Arabic letters. This produces
a few new glyphs that aren't incorporated into the OpenType tables yet.
2009-10-10 00:37:44 +01:00
Ben Harris
394a3a28d6
Create a simple Web page for the font.
2009-10-08 21:33:21 +01:00
Ben Harris
e5f22fa557
Add a couple of odd glyphs: a .notdef to stop FontForge generating one that
...
violates our design rules, and a dotted circle because Uniscribe would
apparently like it (though it probably won't use it unless we gain proper
combining marks).
2009-10-07 23:18:30 +01:00
Ben Harris
a4f66225b2
Tweaks to Arabic shaping following tests.
2009-10-07 22:56:49 +01:00
Ben Harris
16ef6bb038
Add an appropriate GaspTable, even if I haven't seen anything pay
...
attention to it yet.
2009-10-06 23:40:26 +01:00
Ben Harris
4cfb429224
Adobe's current recommendation is that U+00B5 should be called "mu", so
...
follow that.
2009-10-06 23:25:15 +01:00
Ben Harris
dcf01d9324
We need tails for isolated forms as well.
2009-10-06 23:14:28 +01:00
Ben Harris
a6e5df1f25
Name and provide final forms for a bunch more Arabic glyphs.
2009-10-06 22:59:54 +01:00
Ben Harris
319337567a
Initial attempt at Arabic shaping. Not sure if it'll work.
2009-10-06 22:38:24 +01:00
Ben Harris
3de04d502a
Use uni* names for most Arabic glyphs.
2009-10-06 21:14:48 +01:00
Ben Harris
7e12083599
Arrange to connect many Arabix glyphs to the left to help with joined-up
...
writing.
2009-10-06 20:43:37 +01:00
Ben Harris
ac30cc18bf
It would help if our 'c2sc' table actually specified its feature as 'c2sc'.
2009-10-06 19:54:34 +01:00
Ben Harris
3aa319a0c2
Now that lots of glyphs have "uni*" names, arrange that an unspecified name
...
defaults to that. Much more space for flags!
2009-10-06 00:44:22 +01:00
Ben Harris
f58f6e68c6
More lurking "afii*" names
2009-10-06 00:34:27 +01:00
Ben Harris
04acde30bc
As of AGLFN 1.8 (November 2008) Adobe recommends that afii* names not
...
be used in new fonts. Follow their lead and switch to uni* (which are
simpler).
2009-10-06 00:32:54 +01:00
Ben Harris
acde7d410a
Another Arabic glyph identified (and one moved out of the way).
2009-10-06 00:21:50 +01:00
Ben Harris
a7f9cff84a
Add proper lookups so the small caps are accessible.
2009-10-05 23:40:09 +01:00
Ben Harris
f47076ea3e
Don't bother mentioning the bottom row of each character cell in the
...
initialiser for glyphs, since it's always empty. This should give space in
the source to add extra character metadata.
2009-10-05 23:02:05 +01:00
Ben Harris
3074eef3fb
More comprehensive comments on where I got the glyphs from.
2009-10-05 22:53:23 +01:00
Ben Harris
39641a68f8
Some interesting Arabic-related stuff found in an old EBU document. I've
...
got the glyphs, even if I don't really know what to do with them yet.
2009-10-05 00:57:03 +01:00
Ben Harris
5c5473b0c9
Produce a properly Unicodeish font, rather than one in some bizarre encoding.
2009-10-04 14:31:13 +01:00
Ben Harris
c8b9285383
Add comments explaining some of the less obvious glyph names. Names of
...
Cyrillic and Hebrew letters are from the Unicode Standard.
2009-10-04 11:16:39 +01:00
Ben Harris
220316384e
Great big comment explaining what this program is trying to achieve.
2009-10-04 10:51:28 +01:00
Ben Harris
1e20450eba
Gareth has successfully identified the mystery character in the Hebrew
...
set as the Old Sheqel sign. I have yet to find either a Unicode code
point or a definitive glyph name for it, but I can at least be descriptive.
2009-10-03 23:24:11 +01:00
Ben Harris
827ad4cc67
I've found a more comprehensive version of the SAA5050 datasheet,
...
including various versions intended for western continental Europe,
along with Cyrillic and Hebrew ones. This has given me a load more
glyphs, including canonical versions of ones I'd previously designed
myself. Add them all. Also sort my glyphs better.
2009-10-03 22:28:33 +01:00
Ben Harris
c7e64f025b
A second attempt at "MICRO SIGN", and I seem to have something that actually
...
looks right. Hurrah.
2009-10-03 18:38:59 +01:00
Ben Harris
2ec60179cc
It seems that the "yen" glyph is almost always drawn with two horizontal
...
strokes. After some experimentation, here's a new version that has that
and doesn't look too horrid.
2009-10-03 18:31:42 +01:00
Ben Harris
aaea654549
Easy hints: StdHW and StdVW are 100.
2009-10-03 18:02:45 +01:00
Ben Harris
dbd90bd72e
Cleanups to SFD emitter:
...
* Don't emit Fore, SplineSet, or EndSplineSet if a glyph has no splines.
* Emit a blank line between glyphs.
* Set the "W" flag on all glyphs to indicate that the width is explicitly set.
The last seems to be what's necessary to get Fontforge to believe in my
"space" glyph.
2009-10-03 15:32:25 +01:00
Ben Harris
d6b1d044e9
Reword g.sc to match the corrected full-cap version.
...
Also a POI: gid's font is called "Mallard", not "Mullard".
2009-10-03 14:33:34 +01:00
Ben Harris
0e985e885f
Fixes to "G" and "sterling" from the SAA5050 datasheet. All characters now
...
checked against that and, where it disagrees with the Beeb UG, against a
real SAA5050. In all cases, the datasheet was correct.
2009-10-03 14:22:02 +01:00
Ben Harris
abafbdb41b
Comparing my font with gid's "Mullard Smooth" led me to notice a missing
...
pixel in my 'M'. Now fixed.
2009-10-03 13:53:41 +01:00
Ben Harris
303330aca2
correct Unicode code point for quoteleft
2009-10-02 23:52:55 +01:00
Ben Harris
176e47aeed
More quotation marks, a better degree sign, and some fixes to metadata.
2009-10-02 22:23:32 +01:00
Ben Harris
02aab1064b
Sort additional punctuation into Unicode order.
2009-10-02 20:53:03 +01:00
Ben Harris
4cbd7230f8
Ooh, the SAA5050 datasheet includes character diagrams, including an ASCII
...
set, so steal that. Also add a few quotation mark characters that I think
I've done before but seem to have lost.
2009-10-02 20:50:47 +01:00
Ben Harris
f4191c69fa
Minor silliness: a few obvious extra glyphs and a whole set of small
...
caps, because they turned out far better that they had any right to.
2009-10-01 00:49:39 +01:00
Ben Harris
b5faa814ae
Tidy glyph list a little, sort it more authentically, and so on.
2009-09-30 21:02:28 +01:00
Ben Harris
40a5383c82
Arrange to emit an SFD file instead of PostScript. The result is broken
...
in various ways, but it can at least be loaded into Fontforge.
2009-09-30 20:28:06 +01:00
Ben Harris
90aa1f14a7
Fix a couple of minor warnings from GCC.
2009-09-30 17:44:04 +01:00
Ben Harris
a3c77690c7
Clear up old output code, debugging code, and an unused routine.
2009-09-30 17:42:03 +01:00
Ben Harris
a5b30ba095
Hurrah. Seems to be working, finally.
2009-09-30 17:39:23 +01:00
Ben Harris
adcd38b253
Now producing the right outlines, and ones which don't do anything rude
...
(unlike the old ones).
2009-09-30 14:25:53 +01:00
Ben Harris
50a6eb7b38
Beginnings of an attempt at building character outlines in memory so we
...
can simplify them before emission. No simplification yet and a blatant
bug in whitepixel(), but it's kind of interesting.
2009-09-30 13:41:07 +01:00
