[glyphdata] glyphdata.get_glyph sometimes returns multiple chars for `unicode` when `name` implies a ligature

Here’s an example:

>>> from glyphsLib import glyphdata
>>> glyphdata.get_glyph('T_h')
Glyph(name='T_h', production_name='T_h', unicode=u'Th', category=u'Letter', subCategory=u'Ligature')

The unicode value comes from the call to agl.toUnicode(prodname

>>> from fontTools import agl
>>> agl.toUnicode('T_h')
u'Th'

agl.toUnicode splits the string on _ and maps each component to the result string.

I did expect as result a unicode string with a length of 1 or None. So that I can do ord(result.unicode) if unicode is not None.

[anthrotype]

you mean, “unicode stuff is FUBAR in Python 2”.

⌛️ https://pythonclock.org/

[madig]

Should be fixed in master, we only return what the XML says now.