#13 noon ghunna (U+06BA) initial and medial letterforms have an incorrect dot above

[Philippe Verdy]

Arabic letter “noon ghunna” (U+06BA) or “nūn-e ġunna”, used in Urdu and archaic Arabic; indicates nasalized vowels in Urdu; dotless in all four contextual forms.

However the PakType font shows a dot abot the initial and medial form of this letter.

This is incorrect, as these dotted contextual letterforms only apply to Arabic letter African “noon” (U+08BD): initial and medial forms have one dot above.

This error occurs in 'PakType Naskh Basic', 'PakType Ajrak', 'PakType Tehreer', and 'PakType Naqsh'; but does not occur with other Arabic fonts, notably 'Scheherazade New', 'Noto Naskh Arabic', 'Noto Sans Arabic', 'Noto Nastaliq Urdu', 'Times New Roman' (Windows 11 fonts), ' Awami Nastaliq', 'Jameel Noori Nastaleeq Regular'.

In summary, the PakType fonts are incorrectly mapping these dotted glyphs in U+06BA instead of U+08BD.

[Philippe Verdy]

Note that the isolated and final forms of U+06BA are effectively dotless in PakType fonts.

It seems that you've forgotten to include the initial and medial forms with two rows missing in your "development" Excel speadshet for letterforms and their association with other Arabic diacritics. And then you've filled the "holes" using incorrect assumption. Noon ghunna (U+06BA) does NOT behave the same as the African noon (U+08BD).

This bug is confirmed in Unicode charts and namelist, with the explicit note "dotless in all four contextual forms" for U+06BA.

This bug has a consequence: standard Urdu is not rendered correctly with PakType fonts.
If ever this letter in Urdu used in Pakistan behaves differently than in India, you need to include a localized "ur-PK" feature mapping in OpenType (to differentiate from "ur" or "ur-IN"), so that it behaves like U+08BD (African noon). But I doubt this is the case (or this is a common error where the behavior is just considered in Pakistan like a minor variant, and perceived there as equivalent): Unicode does not indicates anywhere in the standard the existence of such local variation. Even the meaning of "noon ghunna" indicates that it is dotless.

"Noon ghunna" is in fact the common base for ALL "noon" letters encoded distinctly in Unicode, with a single "polyline closed path" needed for each of its 4 letter forms (it should not use any mapping to a composite glyph, unlike what the other "noon" letters need for one or more of their contextual letterforms).

[Lateef Sagar Shaikh]

I'll fix the initial and medial forms of noon ghunna.

Yahoo Mail: Search, Organize, Conquer

On Tue, Apr 29, 2025 at 7:17 AM, Philippe Verdyverdy_p@users.sourceforge.net wrote:
Note that the isolated and final forms of U+06BA are effectively dotless. It seems that you've forgotten to include the initial and medial forms in your development Excel speadshet. And then filled the "holes" using incorrect assumption. U+06BA does NOT bzhave the same as U+08BD.

This bug is confirmed in Unicode charts and namelist, with the explicit note "dotless in all four contextual forms" for U+06BA.

[bugs:#13] noon ghunna (U+06BA) initial and medial letterforms have an incorrect dot above

Status: open
Group: v7.2
Created: Tue Apr 29, 2025 10:51 AM UTC by Philippe Verdy
Last Updated: Tue Apr 29, 2025 10:51 AM UTC
Owner: nobody

Arabic letter “noon ghunna” (U+06BA) or “nūn-e ġunna”, used in Urdu and archaic Arabic; indicates nasalized vowels in Urdu; dotless in all four contextual forms.

However the PakType font shows a dot abot the initial and medial form of this letter.

This is incorrect, as these dotted contextual letterforms only apply to Arabic letter African “noon” (U+08BD): initial and medial forms have one dot above.

This error occurs in 'PakType Naskh Basic', 'PakType Ajrak', 'PakType Tehreer', and 'PakType Naqsh'; but does not occur with other Arabic fonts, notably 'Scheherazade New', 'Noto Naskh Arabic', 'Noto Sans Arabic', 'Noto Nastaliq Urdu', 'Times New Roman' (Windows 11 fonts), ' Awami Nastaliq', 'Jameel Noori Nastaleeq Regular'.

In summary, the PakType fonts are incorrectly mapping these dotted glyphs in U+06BA instead of U+08BD.

Sent from sourceforge.net because you indicated interest in https://sourceforge.net/p/paktype/bugs/13/

To unsubscribe from further messages, please visit https://sourceforge.net/auth/subscriptions/

[Lateef Sagar Shaikh]

I'll work on it

[Philippe Verdy]

Thanks for your work, have a good continuation on this project. May be you'll work at the same time on the second (more minor bug) about kerning pairs (#12) for the next version 7.3 when you'll plan it.
It would be useful to include some texts in Urdu, Pashto, Bosnian, and Jawi in your test samples (which currently target mostly Quranic texts in Arabic, or Sindhi, or possibly some other Pakistanese languages).

[Lateef Sagar Shaikh]

Noon Ghunna shapes are fixed in the latest release.