In Malayalam (iso639-2 language code : mal) there are 37 'vyanchanangal'
(consonants).
All these consonants are usually pronounced with a support of 'swaram'
(vowel) sound
A [U0D06]. The pure forms of consonats is writing with a 'chandrakkala'
(virama [U0D4D])
above the consonant. While pronouncing the pure forms of consonants there
should be
clear sound of vowel U [U0D09]. Some consonants another form, which is
called 'chillu'.
A 'chillu' is a consonant which do not require any vowel support to
prounce. It is writing
with a vowel sign U [U0D41] and 'chandrakkala' (virama [U0D4D]) above
that. Infact
Malayalam has seperate 'lipi' (script) for 7 chillu forms of consonants
which are widely
using in Malayalam. Since we have seperate scripts for most of the
chillus, in writing system
we almost stopped writing chillu forms of other consonants (which is
rarely occurs) as explained
above. Eventhough still you can see some texts written in this style.
Antoine said this
is half form of u that is the 'samvrutokaram' of U [U0D09] (infact
'samvrutokaram' has
a sound of A and U, so the 'virama', 'vowel sign U' and 'combination of
this two' is
used in diffnerent places and texts, some lingusits says that
'samvrutokaram' has
a vowel value.) Now many are writing consonants with virama for chillu
forms of other consonats
One example is that Antoine said : U0D15 + U0D41 + U0D4D (ka, u, virama).
So internaly a chillu can be represented with unicode character sequence
like this :
consonant + vowel sign U [U0D41] + virama [U0D4D].
Then you can render 7 chillu forms with correct script. I will explain how
to do this below.
For making inputting very easy you can use the inscript keyboard layout
standardised by
kerala govt. (See they just added chillus to original inscript keyboard
layout at appropriate
positions, they considered the frequency of occurense of this chillu
forms. I will explain
the drawback of this keyboard layout below.)
The proposal for inclusion of scripts of chillus forms of consonants as
basic characters should
not be accepted by Unicode consortium. (This is going to be submitted (or
already submitted?)
by Ministry of Information Technology (Govt. of India), a member of
Unicode consortium)
The prosal includeds some other things, in my opinion those changes should
be accepted.
Now I will explain howto represent chillu forms of consonants in unicode
sequence.
An important thing to be noticed is that two (or more) consonants may have
same script
for their chillu forms. And its pronouciation is also same. Though it
should be represented
in correct unicode sequence. Script for chillu forms of both RA [U0D30]
and RRA [U0D31]
are same. Similary script for chillu forms of both LLA [U0D33] and LLLA
[U0D34] are same.
Other consonants which has chillu forms with unique scripts are NNA
[U0D23], NA [U0D28]
and LA [U0D32].
Why 5 scripts of 7 chillus forms of consonants should not be included in
unicode ?
--
* The basic reason is that those 5 'lipi' (script) are not part of
Malayalam 'Aksharamala'
(character set). instead these are chillus only (See it is not a
'koottaksharam'
(consonanat conjunct) )
Sopporting reasons :-
+ As I explained above two (or more) consonants is using same script for
their chillu
forms. So if these 'simple shapes' are going to be part of unicode
hard encoding of
hard encoding of chillus wll be impossible. If someone input in
correct unicode seqence
the renderer should render those characters, this will make more
problems.
+ Sorting rule cannot impliment effectively.
Inscript keyboard layout problems :-
-
I think the drawback of new inscript keyboard layout standardised by
Kerala govt.
will be clear from the above discussion. Eventhough the layout can be
accepted with
practical consideration. Since we are only using those scripts, we can
compose
any character sequence to keys allocated to them. Here the choice is
coiming in between
RA [U0D30] and RRA [U0D31] chillu and LLA [U0D33] and LLLA [U0D34]. By
considering the
accent of pronounciation and freequency of occurense of these chillus, you
can choose
RRA [U0D31] and LLA [U0D33]. Infact this only can be decided by cosidering
the words.
For example :-
RA [U0D30] + vowel sign U [U0D41] + virama [U0D4D] is correct in words :
neer - neere (water), avar - avare (they), aar - aare (who) etc.
and RRA [U0D31] + vowel sign U [U0D41] + virama [U0D4D] is correct in words :
car - caRe (car), kiNar - kiNaRe (well), sir - saRe (sir) etc.
So if someone input the other correct sequences (without using those keys),
it should render properly.
P.S : please reply to [EMAIL PROTECTED]
Regards,
Baiju M
--
http://baijum81.tripod.com
--- In [EMAIL PROTECTED], Antoine LECA [EMAIL PROTECTED] wrote:
Hi folks,
A problem was signaled in the Microsoft VOLT mailing list (this list
should be dedicated
