Topic Multi key sequence having different combination

How can the following be achieved:

c > U+004C
c + h > U+004C
c + h + h > U+0052

q > U+004C
q + h > U+004D

k > U+004C
k + h > U+004D

If I use context (U+004C) the first combination rule gets fired and later two combinations does not work.

Is there a way to work with previous keystrokes rather than context. For example a dynamic variable array that stores few last keystrokes etc and conditional processing of the same.

Please help. I am a newbie and learning Keyman Developer while trying to achieve something. The keyboard I am designing is phonetic and creates different letters and conjuncts using different key combinations. I am also worried in inserting the akaar/eekar/matras before or after letters/conjuncts.

Please Help.

I'm not entirely clear on how your input is working. Do you mean:

(a) if you type 'c', you get 4C; if you type 'cc' you get 4C 4C; and if you type 'cch' you get 4C 4C 52, or

(b) if you type 'c', you get 4C; if you type 'cc' you also get 4C; but if you type 'cch' you get 52?

Both of these can be solved with the context model but I'd like to give the right answer :)

Good Morning Marc,

My bad Typo. c+h > U+0051
so the following would be correct: Left hand side are key combinations, right hand side desired output
The code tag is used for dexterity

:

c          > U+004C  c L (ক) c c + h     > U+0051  c Q(চ) c+h c + h + h > U+0052  c R (ছ) c+h+h q         > U+004C  c L (ক) c q + h     > U+004D  c M(খ) k+h    k         > U+004C  c L (ক) c  k + h     > U+004D  c M(খ) k+h

This is what I was doing:

:

+ 'c'         > U+004C c L (ক) c U+004C + 'h'   > U+0051 c Q (চ) c+h U+0051 + 'h'   > U+0052 c R (ছ) c+h+h  + 'k'         > U+004C c L (ক) k U+004C + 'h'   > U+004D c M (খ) k+h

The comment section has the remapped character, followed by bengali character it renders, followed by the keystrokes that it produced.
Naturally the last line is never getting fired. I need to understand "context" better. I have learnt from your help the context is whatever at the cursor, how it is related to previous keystroke.

Best Regards
Sougata

What you have is nearly there. In order to distinguish between the 'c' and 'k' keys, you'll need to use a deadkey (http://help.keyman.com/developer/language/reference/deadkey) placeholder (dk is an alias for deadkey):

:

+ 'c'                     > deadkey(c) U+004C c L (ক) c deadkey(c) U+004C + 'h'   > U+0051            c Q (চ) c+h U+0051 + 'h'              > U+0052            c R (ছ) c+h+h + 'k'                     > deadkey(k) U+004C c L (ক) k deadkey(k) U+004C + 'h'   > U+004D            c M (খ) k+h

Putting the deadkey first allows you to ignore if it you aren't interested in it.

Didn't know deadkey could be used like that. Awesome. Thanks.