When using the Nafees Nastaleeq font for Urdu, I noticed various differences or unexpected outcomes between versions 1.00 and 1.01 which I will list here.
| code points | v1.00 | v1.01 |
|---|---|---|
|
0635:
062F:
0627:
0020:
06D3:
0020:
0628:
0644:
0646:
062F:
|
صدا ۓ بلند | صدا ۓ بلند |
|
0635:
062F:
0627:
06D3:
0628:
0644:
0646:
062F:
|
صداۓبلند | صداۓبلند |
|
0635:
062F:
0627:
0020:
06D2:
0654:
0628:
0644:
0646:
062F:
|
صدا ۓبلند | صدا ۓبلند |
|
0635:
062F:
0627:
06D2:
0654:
0628:
0644:
0646:
062F:
|
صداۓبلند | صداۓبلند |
The shape of the yeh baree with precomposed vs decomposed characters is different. Since these are supposed to be canonically equivalent, I thought this must be an error.
| code points | v1.00 | v1.01 |
|---|---|---|
|
0648:
0644:
0626:
0020:
06A9:
0627:
0645:
|
ولئ کام | ولئ کام |
|
0648:
0644:
0626:
06A9:
0627:
0645:
|
ولئکام | ولئکام |
|
0648:
0644:
064A:
0654:
06A9:
0627:
0645:
|
ولئکام | ولئکام |
|
0648:
0644:
06CC:
0654:
06A9:
0627:
0645:
|
ولیٔکام | ولیٔکام |
Here I am trying to reproduce a yeh with hamza above for the izafat.
The 1.01 font seems to force you to never have a space or end of text after the precomposed ARABIC LETTER YEH WITH HAMZA ABOVE: otherwise a dotted circle remains (also in the PDF file listing the characters in the font that accompanies the font download). Note that this seems inconsistent with use of the izafat in the following, where I think a space is needed to achieve the correct form of beh: طالبِ علم
The other thing I am confused about is that ARABIC LETTER YEH WITH HAMZA ABOVE canonically decomposes into ARABIC LETTER YEH plus HAMZA ABOVE in NFD, whereas the font doesn't provide a glyph for ARABIC LETTER YEH. I would have expected that to achieve yeh with hamza the font would make you use ARABIC LETTER FARSI YEH and ARABIC HAMZA ABOVE (there is no precomposed form of this combination that I'm aware of). Using the farsi yeh makes sense in my mind because the hamza is not always shown in this izafat situation, as I understand it. That would mean that the user would use ARABIC LETTER FARSI YEH. That then would lead to an issue, surely, when comparing the same text with and without the hamza: it should be considered the same, but will use different characters when decomposition takes place prior to collation.
Version: $Id: about.html,v 1.31 2003/06/23 16:16:10 rishida Exp $