SuffixAutomator for one-to-one substitutions

Michael Rafailyk

Hello! Type design applications are smart enough and automatically generate most features based on character suffixes. But sometimes it is necessary to write something custom, so I wrote a small automation for this.

Features

• Automate a substitution for an OpenType features.
• Make a receipts for generating composites.
• Wrap substituted characters into classes.
• Reverse substitution direction.
• Capitalize / decapitalization source / substituted characters.
• Sort the characters list.

Links

SuffixAutomator in action.
Project page on GitHub with feature explanation.

Preview

Michael Rafailyk

Updated links:
In Action: https://michaelrafailyk.github.io/SuffixAutomator/
On GitHub: https://github.com/michaelrafailyk/SuffixAutomator

Grzegorz Luk (gluk)

nice tool!

btw. if i can give you some advice - feature "decapitalize" is quite limited in use, because such problems occur:

  sub ae by Ae.alt;
  sub germandbls by Germandbls.alt;

instead
  sub ae by AE.alt;
  sub germandbls by uni1E9E.alt;

a better solution would be a "lowercase" feature using a table of names. For my own use I made one (limited to glyphs I use in my production):

tU2L={
'Aring':'aring',
'uni0216':'uni0217',
'Ograve':'ograve',
'Uring':'uring',
'Edieresis':'edieresis',
'Obreve':'obreve',
'Ccedilla':'ccedilla',
'uni020A':'uni020B',
'Kcommaaccent':'kcommaaccent',
'Zdotaccent':'zdotaccent',
'Hcircumflex':'hcircumflex',
'Otilde':'otilde',
'Ccaron':'ccaron',
'Gcircumflex':'gcircumflex',
'Oslash':'oslash',
'Ecircumflex':'ecircumflex',
'Uacute':'uacute',
'Cacute':'cacute',
'Gcaron':'gcaron',
'Ccircumflex':'ccircumflex',
'Dcroat':'dcroat',
'Ecaron':'ecaron',
'Omacron':'omacron',
'D':'d',
'uni0208':'uni0209',
'Lcaron':'lcaron',
'uni0206':'uni0207',
'H':'h',
'uni0204':'uni0205',
'F':'f',
'Ydieresis':'ydieresis',
'L':'l',
'uni0200':'uni0201',
'Amacron':'amacron',
'P':'p',
'T':'t',
'X':'x',
'Aogonek':'aogonek',
'Udieresis':'udieresis',
'uni01F8':'uni01F9',
'Scircumflex':'scircumflex',
'uni0202':'uni0203',
'Rcaron':'rcaron',
'Agrave':'agrave',
'Jcircumflex':'jcircumflex',
'Umacron':'umacron',
'Edotaccent':'edotaccent',
'uni022E':'uni022F',
'Egrave':'egrave',
'Eogonek':'eogonek',
'uni0214':'uni0215',
'Utilde':'utilde',
'uni0210':'uni0211',
'uni0212':'uni0213',
'Ntilde':'ntilde',
'Ucircumflex':'ucircumflex',
'Atilde':'atilde',
'Sacute':'sacute',
'Gcommaaccent':'gcommaaccent',
'Ncommaaccent':'ncommaaccent',
'Gbreve':'gbreve',
'Zcaron':'zcaron',
'Odieresis':'odieresis',
'Scommaaccent':'scommaaccent',
'Ebreve':'ebreve',
'Acircumflex':'acircumflex',
'uni021E':'uni021F',
'uni021A':'uni021B',
'uni0228':'uni0229',
'Oslashacute':'oslashacute',
'C':'c',
'G':'g',
'Imacron':'imacron',
'Rcommaaccent':'rcommaaccent',
'K':'k',
'O':'o',
'S':'s',
'W':'w',
'Eth':'eth',
'Hbar':'hbar',
'Uhungarumlaut':'uhungarumlaut',
'Ocircumflex':'ocircumflex',
'Igrave':'igrave',
'Zacute':'zacute',
'uni020E':'uni020F',
'uni0232':'uni0233',
'AE':'ae',
'Dcaron':'dcaron',
'Uogonek':'uogonek',
'Lslash':'lslash',
'Scaron':'scaron',
'Ycircumflex':'ycircumflex',
'uni01CD':'uni01CE',
'Iacute':'iacute',
'AEacute':'aeacute',
'uni020C':'uni020D',
'uni1E9E':'germandbls',
'B':'b',
'Iogonek':'iogonek',
'J':'j',
'N':'n',
'R':'r',
'V':'v',
'Z':'z',
'Idieresis':'idieresis',
'OE':'oe',
'Tcaron':'tcaron',
'Nacute':'nacute',
'Lcommaaccent':'lcommaaccent',
'Ugrave':'ugrave',
'uni0226':'uni0227',
'Abreve':'abreve',
'Racute':'racute',
'Itilde':'itilde',
'uni01E8':'uni01E9',
'Yacute':'yacute',
'Lacute':'lacute',
'Ibreve':'ibreve',
'uni01E2':'uni01E3',
'Ubreve':'ubreve',
'Gdotaccent':'gdotaccent',
'Adieresis':'adieresis',
'Ohungarumlaut':'ohungarumlaut',
'uni01EA':'uni01EB',
'Cdotaccent':'cdotaccent',
'Scedilla':'scedilla',
'Aacute':'aacute',
'uni01EC':'uni01ED',
'Icircumflex':'icircumflex',
'uni01CF':'uni01D0',
'Emacron':'emacron',
'Eacute':'eacute',
'A':'a',
'Wcircumflex':'wcircumflex',
'E':'e',
'I':'i',
'M':'m',
'Tcedilla':'tcedilla',
'Tbar':'tbar',
'Q':'q',
'U':'u',
'Y':'y',
'Ncaron':'ncaron',
'uni01D1':'uni01D2',
'uni01D3':'uni01D4',
'Oacute':'oacute',
'Thorn':'thorn',
'Wgrave':'wgrave',
'Wacute':'wacute',
'Wdieresis':'wdieresis',
'uni1E86':'uni1E87',
'uni1E8E':'uni1E8F',
'uni1EBC':'uni1EBD',
'Ygrave':'ygrave',
'uni1EF8':'uni1EF9',
'uni01F4':'uni01F5',
'Ldot':'ldot',
'IJ':'ij',
'Tcommaaccent':'tcommaaccent',
'uni1E88':'uni1E89'
}

Michael Rafailyk

@Grzegorz Luk (gluk) it's a good point, thank you!

Just upload an update. I cover it using exceptions.
If the first two letters of the character name is AE OE IJ NJ (in both lower and upper cases) the first two letters both will be converted.
If the first three letters of the character name is uni or UNI, the conversion doesn't happens.

This is how capitalize and decapitalize now works with exceptions (I marked them with red):

Michael Rafailyk

@Grzegorz Luk (gluk) I also noticed that your table cover case conversion on the Unicode level like from uni021A to uni021B.

Interesting. It's really a bit more difficult to implement, however it's also possible with a few numeral and alphabetic rules. Not sure if I will implement it, because developers usually use friendly glyph names.

Also, most type design application do it all automatically, covering correct unicode name conversion. So this tool is for some non-standard decisions when type design application can't handle with your non-standard suffix.

Grzegorz Luk (gluk)

Michael Rafailyk said:

... I also noticed that your table cover case conversion on the Unicode level like from uni021A to uni021B....developers usually use friendly glyph names...

as I mentioned above, my solution is very custom to my production, for some glyphs I use uniXXXX names, like uni1E9E - LATIN CAPITAL LETTER SHARP S (lowercase is germandbls)

Michael Rafailyk

Any reasons or requirements of using the production names in your process?
Is it some external text feature file that should not contain the friendly/nice names?
I mean, usually applications automatically convert the glyph names to production names at the export.

Grzegorz Luk (gluk)

I think I gave you an example: uni1E9E. I use this name for CAPITAL LETTER SHARP S (and probably many other typedrawers and my application don't convert it automatically in any frendly name.

I could probably change it, but I prefer to use my own scripts for this conversion

Michael Rafailyk

I understand. Thanks for sharing your experience!

Michael Rafailyk

Here's two updates for SuffixAutomator.

1. Capitalize / decapitalize now available for both source (from) and substituted (to) characters. Previously that was possible only for substituted (to) characters.



2. Switch from feature substitution code to composite receipts. For example, it could be useful for generating a list of a small caps like a.sc from a list of capitals like A – for this you need to enter the list of capital characters in the left text area, press the composite button and press the decapitalize to button.