Microsoft OT specs consider just two akhand ligatures
This site shows the two above plus 5 akhand ligatures
Why the discrepancy ?
'Akhand' is a fairly loose concept. From an OpenType perspective, it simply refers to conjunct ligatures that are formed before other shaping occurs. The two that are identified in the MS Devanagari spec are conjuncts that are traditionally treated as separate letters in some situations, e.g. being mapped to individual keys on keyboards.
[The word 'akhand' had an entirely different meaning in some earlier typesetting technologies.]
"The two that are identified in the MS Devanagari spec are conjuncts that are traditionally treated as separate letters in some situations, e.g. being mapped to individual keys on keyboards.
Indeed, using the Devanagari keyboard I was able to get the glyphs क्ष and ञ, as I have just done here. For the glyph क्ष, I used the key corresponding to the '&' on the US keyboard, and it seems to be equal to the KaSsa ligature mentioned above. The second one though, which I obtained by typing '}' in the US kb, doesn't seem to equal the JaNya ligature. Just by looking at the Devanagari keyboard, I couldn't find any other candidate for the JaNya ligature glyph. Could you clarify ?
I'm still striving to correctly recognize the devanagari characters. After a few hours of research on this issue, I think I finally found out the answer for the doubts, I expressed above.
In the first case, the KaSsa ligature, that was no problem, it was correctly recognized and reproduced in MS Word, by typing :
क + ् + ष = क्ष
using the Devanagari keyboard.
The problem with the JaNya ligature was due to the fact that I didn't recognized correctly the character ज्ञ in the alluded keyboard, mistaking it with the character ञ. Making this correction, I obtained in MS Word the expected result :
ज + ् + ञ = ज्ञ
And more, these two characters (or glyphs), क्ष and ज्ञ, although not Unicode points, they are present in the Devanagari keyboard, on the keys corresponding, respectively, to the characters '&' and '%', in the American keyboard, as John mentioned above.
One other important point : when I said "obtained in MS Word", I meant that MS Word automatically replaced the typed characters with the glyphs at the right of the equal signs, when the ligatures were identified by the shaping engine, used by Word.
There seems to be a bug on this page. Just look at the way the spaces between the Indic characters show above. The strangest thing is that while I previewed the text yesterday, everything seemed to be all right. Even after I've posted it, I could see the text was in perfect conditions yesterday. But no more today. That's weird. Maybe I've done something wrong while mixing Indic and Latin characters. Any suggestion to avoid or work around this ?
Thanks in advance.
Edit : As soon as I posted this last thread, the problems related above suddenly disappeared again. I have no idea what's happening here !