Add special chars for Mahan Kosh
dsomel21 opened this issue · comments
Summary
(More of a note for myself).
Description
Reference in Mahan Kosh.
Is that thing under the Kakka corect? This is what it looks like in Sant Lipi:
Approach
It's a style thing. Can you find hand written sarups for reference?
I found it. I don't have a camera on me, but it basically looks like a Kakka with a line through the bottom hole.
Kinda like this:
Okay, I got Jaspreet to send me a pic from her phone:
What you're referencing is a special character used in Mahan Kosh. It will be added later. I'm going to update this issue for that.
Notes (going down column 1)
Even though Devanagari and Gurmukhi both stem from Brahma/Gupta scripts, text rendering engines will not allow gurmukhi vowels on devanagari code points. So can't just style gurmukhi letters on the devanagari unicode points unless they are being used as is (without vowel modification). Others which can fit into Gurmukhi without issue can just be added as is (e.g. half-m and r w/ nukta).
- Half s (typed as स् + char)
- ṣ, no clue why the dot is added in the gurmukhi version
- 1st is half ṣ (typed as ष् + char), 2nd one is half ś (typed as श् + char)
- The conjunct for kṣa (क् + ष = क्ष)
- Half kṣa (typed as क्ष् + char)
- The conjunct for jña (ज् + ञ = ज्ञ)
- Half r (typed as र् + char). This is not pairin-rara in punjabi, which comes after the consonant, instead this is a half-r preceding the conjunct, even though it is written at the far top right of the conjunct, it is pronounced initially in the cluster.
- Addak Bindi. This exists in Gurmukhi already. However, in Sant Lipi it is specifically being used to produce a bindi before bihari. Either have to change that strategy, or need to use Devanagari Addak Bindi whenever it is needed in Gurmukhi (aside from Mahan Kosh, I have zero examples except obscure and niche old dictionaries, whose legitimacy I haven't researched/looked into, but regardless could also be satisifed with Devanagari Addak Bindi U+0901, Chandrabindu).
- Half m (typed ੍ + ਮ = ੍ਮ). This already exists in Unicode fonts today, it is already in Sant Lipi as well. I don't know if this is being used at the beginning of a word, in which case we may need to evaluate the impact of showing the dotted circle at the start (as this is meant to be used non-initially).
- Independent form of vocalic r (r̥ = U+090B), basically extinct today
Notes (going down column 2)
This seems even harder to integrate into the font. Will have to investigate further.
- ṣâd (U+0635)
- s̱e (U+062B), shows up 220 times
- ẕâl (U+0630)
- že (U+0698), shows up around a dozen times in the entire Mahan Kosh, and only thrice without it's corresponding arabic/farsi equivalent.
- zâd (U+0636)
- ẓâ (U+0638)
- ʿayn (U+0639)
- ġayn (U+063A)
- fe (U+0641)
- qâf (U+0642)
- xe (U+062E)
- ze (U+0632)
The most likely approach will be to add ligatures with control chars for Gurmukhi. The same way we use addak bindi for bindi-before-bihari. Or the way we use north-indic fraction chars to vary yayyas. Will have a play around and see what works.
Questions / things to confirm or look into:
- Any instance of adak-bindi with bihari? If no, then can use adak bindi as is
- Does using north indic fractions keep the word together when reaching the end of a line (does the entire word go to the next line or will half remain above and other half come to the next line, acting as a separator of sorts)?
BTW, how are you looking up the number of instances in Mahan Kosh?
If you search the character in the PDF, it shows the number of results. You can find a PDF of Mahan Kosh with a quick google search.
Print of Mahan Kosh that I have at home:
