-1

Background

The Arabic letter noon ghunna (ں) is displayed incorrectly on my Windows 10 PC (in Chrome, Edge, Notepad and Word). The sequence ALEF, NOON GHUNNA, ALEF is displayed as:

ana

The same sequence is displayed correctly on my Android phone without the dot:

ana

For completeness, the actual unicode string (for copy/paste purposes) is:

اںا

There has been some controversy regarding this letter (L2-12/381) which has settled by now as seen from the Unicode Standard which states (since version 7 and up to the current 11):

Rendering systems should display U+06BA as a dual-joining letter, with all four contextual forms shown dotless, regardless of the language of the text.

But the dot appears in word-initial (ںا) and mid-word (اںا) positions. Final (اں) and isolated (ں) forms are fine.

Question

Now my question is, how can this be fixed, other than by waiting for Microsoft to fix it? I want to understand where the problem lies. Is it in the Uniscribe library, or is it down to the font being used? Can it be fixed by using a specifically crafted TrueType/OpenType font?

Mike 'Pomax' Kamermans
  • 49,297
  • 16
  • 112
  • 153
Sergey Slepov
  • 1,861
  • 13
  • 33
  • 2
    FWIW it displays without the dot on my Windows 10 system. Tried Chrome and Edge. Did you try changing the font? – JJF Jun 11 '18 at 13:52
  • Interesting! Yes, I browsed lots of fonts in Word and also on fonts.com and they all either display a dot or show the isolated form. – Sergey Slepov Jun 11 '18 at 13:57
  • اںا <== here is the actual Unicode so you can see what it looks like on your system. I see a dot. – Sergey Slepov Jun 11 '18 at 13:59
  • That displays with the dot on my system. But... it is not the same sequence of code points that I tested. Copying the character out of your comment gives me utf8 encoded characters (big endian) 0xd8a7 0xdaba 0xd8a7. My test was done with with an html file with the contents ں which apparently is some sort of html entity. I'm very confused about this... – JJF Jun 11 '18 at 15:55
  • I type "html entity ں" into Google's search box and Google interpreted it as the un-dotted character and this was the first search result http://graphemica.com/%DA%BA – JJF Jun 11 '18 at 16:03
  • Oh, I forgot to mention that the dot only appears in word-initial (ںا) and mid-word (اںا) positions, but not in final (اں) or isolated (ں) forms. – Sergey Slepov Jun 11 '18 at 18:14
  • This is a matter of font, not a matter of OS: the font contains the instructions on which letters to show and how, not the font engine. The engine only runs through the font data and follows the instructions it finds. As such: which Arabic fonts are you seeing this happen for? – Mike 'Pomax' Kamermans Jun 13 '18 at 18:57

1 Answers1

0

This turned out to be a font problem. Quite a few fonts on fonts.google.com show this letter correctly:

Sergey Slepov
  • 1,861
  • 13
  • 33