Urdu ہ & ھ are not visually distinguished in pairs like کہا & کھا

[khaledhosny]

INTRODUCTION

In Urdu, ـہـ and ـھـ are not stylistic variants as they are in some other Arabic-script languages like Arabic and Farsi. ہ (HEH GOAL) is semantically different from ھ (HEH DOACHASHMEE); the former is the consonant ‘h’ and the latter causes the preceding consonant to be aspirated (i.e. ک क k becomes کھ ख kh).

PROBLEM

As shown in the attached screenshot, the sequences کہا (Urdu: “[he/she] said”) and کھا (Urdu: “eat!”) look identical in Amiri currently. This is a major usability issue for Amiri in this language. However, it is worth noting that the bug does not reproduce in all cases. For example, the words بہار (Urdu: “spring”) and بھار (Urdu: “weight”) are correctly distinguished.

NOTES

Standard Urdu fonts will maintain the distinction between ہ and ھ in all forms (initial, medial, final, isolate). However, since initial and isolate are semantically meaningless for ھ (which needs a preceding consonant), many authors choose to write the ھ glyph in the initial position when it will be unambiguous and equivalent to ہ. I have noticed that Amiri does this also in that words like ہے are written to look like ھے*. While this is a valid approach and is used in certain styles, I just wanted to point out that it’s not the standard approach.

• The sequence ھے is only given for demonstration purposes. Using ھ in this manner would be semantically invalid.

Original comment by: karanm

Original Ticket: amiri/bugs/63

[khaledhosny]

Here’s a screenshot of one of the examples in which Amiri is rendering the distinction correctly.

Original comment by: karanm

[khaledhosny]

And here’s a screenshot showing the stylistic variation for ہے. Amiri at the top is using a ھ-looking initial, while Jameel Noori Nastaleeq and Geeza Pro stick to ہـ.

Original comment by: karanm

[khaledhosny]

It took me a while to work on this, but here is my plan, please confirm if this is sound approach:

• No changes to the current handling of HEH DOACHASHMEE.
• For HEH GOAL:
  • [ ] Initial form will look a bit like dotless Yeh with comma below.
  • [x] The medial form will always be the open one (is the comma necessary here?).
  • [x] The final form will be an open one as well (like the one used in Nastaliq).

[khaledhosny]

Sounds like a good plan!

Original comment by: karanm

[khaledhosny]

So I have done the 2nd and 3rd points above, the first one however is harder to get in a Naskh font like Amiri and the result is ugly in my eyes, since the current approach is still considered valid, I’ll keep it for now.