From gmhossain at gmail.com  Sun Feb  8 11:55:43 2009
From: gmhossain at gmail.com (Golam Mortuza Hossain)
Date: Sun, 8 Feb 2009 15:55:43 -0400
Subject: [Padma] Malformed Conjuncts
Message-ID: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>

Hi,

While trying to add a new font (Datasoft, www.aajkaal.net)
in Padma,  I am facing couple of issues that are breaking
conjunct formation.

The current handling of prefixList in Padma breaks conjunct
formation, if the conjunct follows a matra which is in prefixList.
For example: If the ordering in original font is say

Vowelsn_E  + (Letter_LA + Halant + Letter_PA).   { "?" + "???" }

As Vowelsn_E is in prefixList, Padma correctly pushes it down to

 Letter_LA + Vowelsn_E  + Halant + Letter_PA.   { "????" }

However, notice that it now breaks conjunct formation. The correct
ordering should be

 (Letter_LA + Halant + Letter_PA)  + Vowelsn_E  .   { "????" }

This along with few other issues can be easily handled if Padma
API is expanded to have a post processing subroutine. This
should be analogous to the current "preprocessMessage"  API.
For example, I am considering following code

     if (encoding.postprocessMessage != undefined)
        this.text = encoding.postprocessMessage(input);

However, I am not sure about the best place to put the
code in Padma. Can anyone please shed some light on this?

Thanks,
Golam

From vnagarjuna at gmail.com  Tue Feb 10 12:59:33 2009
From: vnagarjuna at gmail.com (Nagarjuna Venna)
Date: Tue, 10 Feb 2009 15:59:33 -0500
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
Message-ID: <89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>

Hi Golam,

I don't think the parser works as you described. It actually pushes
the vowel sign to the end (modulo other things that can go at the end)
or at least that's what I remember. I think if it did not do that,
most of the transformations would be broken.

I didn't have a chance to look at the code yet, but may be the problem
is something else?

Thanks,
Nagarjuna

On Sun, Feb 8, 2009 at 2:55 PM, Golam Mortuza Hossain
<gmhossain at gmail.com> wrote:
> Hi,
>
> While trying to add a new font (Datasoft, www.aajkaal.net)
> in Padma,  I am facing couple of issues that are breaking
> conjunct formation.
>
> The current handling of prefixList in Padma breaks conjunct
> formation, if the conjunct follows a matra which is in prefixList.
> For example: If the ordering in original font is say
>
> Vowelsn_E  + (Letter_LA + Halant + Letter_PA).   { "?" + "???" }
>
> As Vowelsn_E is in prefixList, Padma correctly pushes it down to
>
>  Letter_LA + Vowelsn_E  + Halant + Letter_PA.   { "????" }
>
> However, notice that it now breaks conjunct formation. The correct
> ordering should be
>
>  (Letter_LA + Halant + Letter_PA)  + Vowelsn_E  .   { "????" }
>
> This along with few other issues can be easily handled if Padma
> API is expanded to have a post processing subroutine. This
> should be analogous to the current "preprocessMessage"  API.
> For example, I am considering following code
>
>     if (encoding.postprocessMessage != undefined)
>        this.text = encoding.postprocessMessage(input);
>
> However, I am not sure about the best place to put the
> code in Padma. Can anyone please shed some light on this?
>
> Thanks,
> Golam
> _______________________________________________
> Padma mailing list
> Padma at mozdev.org
> https://www.mozdev.org/mailman/listinfo/padma
>

From gmhossain at gmail.com  Tue Feb 10 14:08:06 2009
From: gmhossain at gmail.com (Golam Mortuza Hossain)
Date: Tue, 10 Feb 2009 18:08:06 -0400
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
Message-ID: <141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>

Hi Nagarjuna,

Thanks for your response.

May be, I should clarify the situation here. These conjuncts are
not in the font  as separate glyphs rather they are being
formed by putting two half letters together.  These half letters
have separate glyphs in the font.




On Tue, Feb 10, 2009 at 4:59 PM, Nagarjuna Venna <vnagarjuna at gmail.com> wrote:
> Hi Golam,
>
> I don't think the parser works as you described. It actually pushes
> the vowel sign to the end (modulo other things that can go at the end)
> or at least that's what I remember. I think if it did not do that,
> most of the transformations would be broken.
>
> I didn't have a chance to look at the code yet, but may be the problem
> is something else?
>
> Thanks,
> Nagarjuna
>
> On Sun, Feb 8, 2009 at 2:55 PM, Golam Mortuza Hossain
> <gmhossain at gmail.com> wrote:
>> Hi,
>>
>> While trying to add a new font (Datasoft, www.aajkaal.net)
>> in Padma,  I am facing couple of issues that are breaking
>> conjunct formation.
>>
>> The current handling of prefixList in Padma breaks conjunct
>> formation, if the conjunct follows a matra which is in prefixList.
>> For example: If the ordering in original font is say
>>
>> Vowelsn_E  + (Letter_LA + Halant + Letter_PA).   { "?" + "???" }
>>
>> As Vowelsn_E is in prefixList, Padma correctly pushes it down to
>>
>>  Letter_LA + Vowelsn_E  + Halant + Letter_PA.   { "????" }
>>
>> However, notice that it now breaks conjunct formation. The correct
>> ordering should be
>>
>>  (Letter_LA + Halant + Letter_PA)  + Vowelsn_E  .   { "????" }
>>
>> This along with few other issues can be easily handled if Padma
>> API is expanded to have a post processing subroutine. This
>> should be analogous to the current "preprocessMessage"  API.
>> For example, I am considering following code
>>
>>     if (encoding.postprocessMessage != undefined)
>>        this.text = encoding.postprocessMessage(input);
>>
>> However, I am not sure about the best place to put the
>> code in Padma. Can anyone please shed some light on this?
>>
>> Thanks,
>> Golam
>> _______________________________________________
>> Padma mailing list
>> Padma at mozdev.org
>> https://www.mozdev.org/mailman/listinfo/padma
>>
> _______________________________________________
> Padma mailing list
> Padma at mozdev.org
> https://www.mozdev.org/mailman/listinfo/padma
>

From vnagarjuna at gmail.com  Wed Feb 11 07:35:43 2009
From: vnagarjuna at gmail.com (Nagarjuna Venna)
Date: Wed, 11 Feb 2009 10:35:43 -0500
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
Message-ID: <89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>

Golam - The parser should be able to deal with it - didn't you have
similar conjunt combinations with ABP?

Nagarjuna


On Tue, Feb 10, 2009 at 5:08 PM, Golam Mortuza Hossain
<gmhossain at gmail.com> wrote:
> Hi Nagarjuna,
>
> Thanks for your response.
>
> May be, I should clarify the situation here. These conjuncts are
> not in the font  as separate glyphs rather they are being
> formed by putting two half letters together.  These half letters
> have separate glyphs in the font.
>
>
>
>
> On Tue, Feb 10, 2009 at 4:59 PM, Nagarjuna Venna <vnagarjuna at gmail.com> wrote:
>> Hi Golam,
>>
>> I don't think the parser works as you described. It actually pushes
>> the vowel sign to the end (modulo other things that can go at the end)
>> or at least that's what I remember. I think if it did not do that,
>> most of the transformations would be broken.
>>
>> I didn't have a chance to look at the code yet, but may be the problem
>> is something else?
>>
>> Thanks,
>> Nagarjuna
>>
>> On Sun, Feb 8, 2009 at 2:55 PM, Golam Mortuza Hossain
>> <gmhossain at gmail.com> wrote:
>>> Hi,
>>>
>>> While trying to add a new font (Datasoft, www.aajkaal.net)
>>> in Padma,  I am facing couple of issues that are breaking
>>> conjunct formation.
>>>
>>> The current handling of prefixList in Padma breaks conjunct
>>> formation, if the conjunct follows a matra which is in prefixList.
>>> For example: If the ordering in original font is say
>>>
>>> Vowelsn_E  + (Letter_LA + Halant + Letter_PA).   { "?" + "???" }
>>>
>>> As Vowelsn_E is in prefixList, Padma correctly pushes it down to
>>>
>>>  Letter_LA + Vowelsn_E  + Halant + Letter_PA.   { "????" }
>>>
>>> However, notice that it now breaks conjunct formation. The correct
>>> ordering should be
>>>
>>>  (Letter_LA + Halant + Letter_PA)  + Vowelsn_E  .   { "????" }
>>>
>>> This along with few other issues can be easily handled if Padma
>>> API is expanded to have a post processing subroutine. This
>>> should be analogous to the current "preprocessMessage"  API.
>>> For example, I am considering following code
>>>
>>>     if (encoding.postprocessMessage != undefined)
>>>        this.text = encoding.postprocessMessage(input);
>>>
>>> However, I am not sure about the best place to put the
>>> code in Padma. Can anyone please shed some light on this?
>>>
>>> Thanks,
>>> Golam
>>> _______________________________________________
>>> Padma mailing list
>>> Padma at mozdev.org
>>> https://www.mozdev.org/mailman/listinfo/padma
>>>
>> _______________________________________________
>> Padma mailing list
>> Padma at mozdev.org
>> https://www.mozdev.org/mailman/listinfo/padma
>>
> _______________________________________________
> Padma mailing list
> Padma at mozdev.org
> https://www.mozdev.org/mailman/listinfo/padma
>

From gmhossain at gmail.com  Wed Feb 11 15:22:26 2009
From: gmhossain at gmail.com (Golam Mortuza Hossain)
Date: Wed, 11 Feb 2009 19:22:26 -0400
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
Message-ID: <141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>

Hi Nagarjuna,

You are right.  Parser indeed correctly deals with prefixList and
half letters, if I declare them as  "Padma.halffm_XX". Earlier I was
writing them as  "Padma.consnt_XX + Padma.halant".

However, the problem remains :-(  Conjuncts are still
not forming for the situation I mentioned. Currently,
I am using "Padma_halffm_XX"  for upper half letters
and "Padma.vattu_YY" for lower half-letters.

For example,

"Padma_halffm_LA + Padma.vattu_PA"  is not forming a
conjunct.

Am I missing something?


Regarding ABP, they are using three fonts simultaneously.
Consequently, they could afford to be over-generous to assign
separate glyphs for each Bengali conjuncts :-)  So I didn't
face this issue earlier.

Thanks,
Golam






On Wed, Feb 11, 2009 at 11:35 AM, Nagarjuna Venna <vnagarjuna at gmail.com> wrote:
> Golam - The parser should be able to deal with it - didn't you have
> similar conjunt combinations with ABP?
>
> Nagarjuna
>
>
> On Tue, Feb 10, 2009 at 5:08 PM, Golam Mortuza Hossain
> <gmhossain at gmail.com> wrote:
>> Hi Nagarjuna,
>>
>> Thanks for your response.
>>
>> May be, I should clarify the situation here. These conjuncts are
>> not in the font  as separate glyphs rather they are being
>> formed by putting two half letters together.  These half letters
>> have separate glyphs in the font.
>>
>>
>>
>>
>> On Tue, Feb 10, 2009 at 4:59 PM, Nagarjuna Venna <vnagarjuna at gmail.com> wrote:
>>> Hi Golam,
>>>
>>> I don't think the parser works as you described. It actually pushes
>>> the vowel sign to the end (modulo other things that can go at the end)
>>> or at least that's what I remember. I think if it did not do that,
>>> most of the transformations would be broken.
>>>
>>> I didn't have a chance to look at the code yet, but may be the problem
>>> is something else?
>>>
>>> Thanks,
>>> Nagarjuna
>>>
>>> On Sun, Feb 8, 2009 at 2:55 PM, Golam Mortuza Hossain
>>> <gmhossain at gmail.com> wrote:
>>>> Hi,
>>>>
>>>> While trying to add a new font (Datasoft, www.aajkaal.net)
>>>> in Padma,  I am facing couple of issues that are breaking
>>>> conjunct formation.
>>>>
>>>> The current handling of prefixList in Padma breaks conjunct
>>>> formation, if the conjunct follows a matra which is in prefixList.
>>>> For example: If the ordering in original font is say
>>>>
>>>> Vowelsn_E  + (Letter_LA + Halant + Letter_PA).   { "?" + "???" }
>>>>
>>>> As Vowelsn_E is in prefixList, Padma correctly pushes it down to
>>>>
>>>>  Letter_LA + Vowelsn_E  + Halant + Letter_PA.   { "????" }
>>>>
>>>> However, notice that it now breaks conjunct formation. The correct
>>>> ordering should be
>>>>
>>>>  (Letter_LA + Halant + Letter_PA)  + Vowelsn_E  .   { "????" }
>>>>
>>>> This along with few other issues can be easily handled if Padma
>>>> API is expanded to have a post processing subroutine. This
>>>> should be analogous to the current "preprocessMessage"  API.
>>>> For example, I am considering following code
>>>>
>>>>     if (encoding.postprocessMessage != undefined)
>>>>        this.text = encoding.postprocessMessage(input);
>>>>
>>>> However, I am not sure about the best place to put the
>>>> code in Padma. Can anyone please shed some light on this?
>>>>
>>>> Thanks,
>>>> Golam
>>>> _______________________________________________
>>>> Padma mailing list
>>>> Padma at mozdev.org
>>>> https://www.mozdev.org/mailman/listinfo/padma
>>>>
>>> _______________________________________________
>>> Padma mailing list
>>> Padma at mozdev.org
>>> https://www.mozdev.org/mailman/listinfo/padma
>>>
>> _______________________________________________
>> Padma mailing list
>> Padma at mozdev.org
>> https://www.mozdev.org/mailman/listinfo/padma
>>
> _______________________________________________
> Padma mailing list
> Padma at mozdev.org
> https://www.mozdev.org/mailman/listinfo/padma
>

From harshita at atc.tcs.com  Wed Feb 11 23:10:48 2009
From: harshita at atc.tcs.com (Harshita)
Date: Thu, 12 Feb 2009 12:40:48 +0530
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
Message-ID: <1234422648.8119.13.camel@localhost.localdomain>

Hi Golam,
  As far as I understand, writing mapping for conjuncts( ???   ) as
"Padma_consnt_LA + Padma.vattu_PA" should work.


Regards,
Harshita


On Wed, 2009-02-11 at 19:22 -0400, Golam Mortuza Hossain wrote:
> Hi Nagarjuna,
> 
> You are right.  Parser indeed correctly deals with prefixList and
> half letters, if I declare them as  "Padma.halffm_XX". Earlier I was
> writing them as  "Padma.consnt_XX + Padma.halant".
> 
> However, the problem remains :-(  Conjuncts are still
> not forming for the situation I mentioned. Currently,
> I am using "Padma_halffm_XX"  for upper half letters
> and "Padma.vattu_YY" for lower half-letters.
> 
> For example,
> 
> "Padma_halffm_LA + Padma.vattu_PA"  is not forming a
> conjunct.
> 
> Am I missing something?
> 
> 
> Regarding ABP, they are using three fonts simultaneously.
> Consequently, they could afford to be over-generous to assign
> separate glyphs for each Bengali conjuncts :-)  So I didn't
> face this issue earlier.
> 
> Thanks,
> Golam
> 
> 
> 
> 
> 
> 
> On Wed, Feb 11, 2009 at 11:35 AM, Nagarjuna Venna <vnagarjuna at gmail.com> wrote:
> > Golam - The parser should be able to deal with it - didn't you have
> > similar conjunt combinations with ABP?
> >
> > Nagarjuna
> >
> >
> > On Tue, Feb 10, 2009 at 5:08 PM, Golam Mortuza Hossain
> > <gmhossain at gmail.com> wrote:
> >> Hi Nagarjuna,
> >>
> >> Thanks for your response.
> >>
> >> May be, I should clarify the situation here. These conjuncts are
> >> not in the font  as separate glyphs rather they are being
> >> formed by putting two half letters together.  These half letters
> >> have separate glyphs in the font.
> >>
> >>
> >>
> >>
> >> On Tue, Feb 10, 2009 at 4:59 PM, Nagarjuna Venna <vnagarjuna at gmail.com> wrote:
> >>> Hi Golam,
> >>>
> >>> I don't think the parser works as you described. It actually pushes
> >>> the vowel sign to the end (modulo other things that can go at the end)
> >>> or at least that's what I remember. I think if it did not do that,
> >>> most of the transformations would be broken.
> >>>
> >>> I didn't have a chance to look at the code yet, but may be the problem
> >>> is something else?
> >>>
> >>> Thanks,
> >>> Nagarjuna
> >>>
> >>> On Sun, Feb 8, 2009 at 2:55 PM, Golam Mortuza Hossain
> >>> <gmhossain at gmail.com> wrote:
> >>>> Hi,
> >>>>
> >>>> While trying to add a new font (Datasoft, www.aajkaal.net)
> >>>> in Padma,  I am facing couple of issues that are breaking
> >>>> conjunct formation.
> >>>>
> >>>> The current handling of prefixList in Padma breaks conjunct
> >>>> formation, if the conjunct follows a matra which is in prefixList.
> >>>> For example: If the ordering in original font is say
> >>>>
> >>>> Vowelsn_E  + (Letter_LA + Halant + Letter_PA).   { "?" + "???" }
> >>>>
> >>>> As Vowelsn_E is in prefixList, Padma correctly pushes it down to
> >>>>
> >>>>  Letter_LA + Vowelsn_E  + Halant + Letter_PA.   { "????" }
> >>>>
> >>>> However, notice that it now breaks conjunct formation. The correct
> >>>> ordering should be
> >>>>
> >>>>  (Letter_LA + Halant + Letter_PA)  + Vowelsn_E  .   { "????" }
> >>>>
> >>>> This along with few other issues can be easily handled if Padma
> >>>> API is expanded to have a post processing subroutine. This
> >>>> should be analogous to the current "preprocessMessage"  API.
> >>>> For example, I am considering following code
> >>>>
> >>>>     if (encoding.postprocessMessage != undefined)
> >>>>        this.text = encoding.postprocessMessage(input);
> >>>>
> >>>> However, I am not sure about the best place to put the
> >>>> code in Padma. Can anyone please shed some light on this?
> >>>>
> >>>> Thanks,
> >>>> Golam
> >>>> _______________________________________________
> >>>> Padma mailing list
> >>>> Padma at mozdev.org
> >>>> https://www.mozdev.org/mailman/listinfo/padma
> >>>>
> >>> _______________________________________________
> >>> Padma mailing list
> >>> Padma at mozdev.org
> >>> https://www.mozdev.org/mailman/listinfo/padma
> >>>
> >> _______________________________________________
> >> Padma mailing list
> >> Padma at mozdev.org
> >> https://www.mozdev.org/mailman/listinfo/padma
> >>
> > _______________________________________________
> > Padma mailing list
> > Padma at mozdev.org
> > https://www.mozdev.org/mailman/listinfo/padma
> >
> _______________________________________________
> Padma mailing list
> Padma at mozdev.org
> https://www.mozdev.org/mailman/listinfo/padma


From gmhossain at gmail.com  Thu Feb 12 03:51:21 2009
From: gmhossain at gmail.com (Golam Mortuza Hossain)
Date: Thu, 12 Feb 2009 07:51:21 -0400
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <1234422648.8119.13.camel@localhost.localdomain>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
	<1234422648.8119.13.camel@localhost.localdomain>
Message-ID: <141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>

Hi Harshita,

On Thu, Feb 12, 2009 at 3:10 AM, Harshita <harshita at atc.tcs.com> wrote:
>  As far as I understand, writing mapping for conjuncts( ???   ) as
> "Padma_consnt_LA + Padma.vattu_PA" should work.

Yes, this combination works.  However, the problem here is that the
font doesn't have a separate glyph for these conjuncts. For example,
the font has separate glyphs for upper half letter SA, lower half-letter
BA and consnt TA. So I am mapping them as

Datasoft.toPadma[Datasoft.codepoint_A4] = Padma.vattu_BA ;
Datasoft.toPadma[Datasoft.codepoint_A6] = Padma.halffm_SA ;
Datasoft.toPadma[Datasoft.codepoint_EE] = Padma.consnt_TA ;

Now consider the following combination to see the issue:

(1) Padma.consnt_TA  + Padma.vattu_BA   => "???"  [OK]
(2) Padma.halffm_SA  + Padma.consnt_TA => "???"  [OK]
(3) Padma.halffm_SA  + Padma.vattu_BA   => "????" [MALFORMED]  should be "???"

Basically, two successive Halant is causing the problem. Two succesive
Halant should be merged into one Halant.

Thanks
Golam

From harshita at atc.tcs.com  Thu Feb 12 05:26:30 2009
From: harshita at atc.tcs.com (Harshita)
Date: Thu, 12 Feb 2009 18:56:30 +0530
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
	<1234422648.8119.13.camel@localhost.localdomain>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
Message-ID: <1234445190.8894.20.camel@localhost.localdomain>

Hi Golam,

Third combination(Padma.halffm_SA  + Padma.vattu_BA) will become
meaningful only if it is combined with some consonant. Other wise padma
will render it as it is (i.e "????"  ).
I think it is logically correct output [input(in dynamic font) and
output(in unicode font) should be equal].  

May be giving an occurrence of the combination(word -- written in
datasoft) will give some more idea.

Thanks & Regards,
Harshita


On Thu, 2009-02-12 at 07:51 -0400, Golam Mortuza Hossain wrote:
> Hi Harshita,
> 
> On Thu, Feb 12, 2009 at 3:10 AM, Harshita <harshita at atc.tcs.com> wrote:
> >  As far as I understand, writing mapping for conjuncts( ???   ) as
> > "Padma_consnt_LA + Padma.vattu_PA" should work.
> 
> Yes, this combination works.  However, the problem here is that the
> font doesn't have a separate glyph for these conjuncts. For example,
> the font has separate glyphs for upper half letter SA, lower half-letter
> BA and consnt TA. So I am mapping them as
> 
> Datasoft.toPadma[Datasoft.codepoint_A4] = Padma.vattu_BA ;
> Datasoft.toPadma[Datasoft.codepoint_A6] = Padma.halffm_SA ;
> Datasoft.toPadma[Datasoft.codepoint_EE] = Padma.consnt_TA ;
> 
> Now consider the following combination to see the issue:
> 
> (1) Padma.consnt_TA  + Padma.vattu_BA   => "???"  [OK]
> (2) Padma.halffm_SA  + Padma.consnt_TA => "???"  [OK]
> (3) Padma.halffm_SA  + Padma.vattu_BA   => "????" [MALFORMED]  should be "???"
> 
> Basically, two successive Halant is causing the problem. Two succesive
> Halant should be merged into one Halant.
> 
> Thanks
> Golam
> _______________________________________________
> Padma mailing list
> Padma at mozdev.org
> https://www.mozdev.org/mailman/listinfo/padma


From gmhossain at gmail.com  Thu Feb 12 07:06:40 2009
From: gmhossain at gmail.com (Golam Mortuza Hossain)
Date: Thu, 12 Feb 2009 11:06:40 -0400
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <1234445190.8894.20.camel@localhost.localdomain>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
	<1234422648.8119.13.camel@localhost.localdomain>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
	<1234445190.8894.20.camel@localhost.localdomain>
Message-ID: <141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>

Hi,

On Thu, Feb 12, 2009 at 9:26 AM, Harshita <harshita at atc.tcs.com> wrote:

> May be giving an occurrence of the combination(word -- written in
> datasoft) will give some more idea.


Here goes a screenshot.  [ Marked in red are incorrect where as
marked in green is perfect (for Padma.halffm_SA but
in different combination) ]

http://www.math.unb.ca/~ghossain/malformed-conjuncts.jpg

Hopefully, it will make the issue more clear. BTW,  "SA + halant + BA"
conjunct is very common in Bengali. Like  "Swadhin" (???????, independent)
"swabhabhik" (?????????, normal).

Thanks
Golam

From harshita at atc.tcs.com  Thu Feb 12 07:22:52 2009
From: harshita at atc.tcs.com (Harshita)
Date: Thu, 12 Feb 2009 20:52:52 +0530
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
	<1234422648.8119.13.camel@localhost.localdomain>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
	<1234445190.8894.20.camel@localhost.localdomain>
	<141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
Message-ID: <1234452172.9737.4.camel@localhost.localdomain>

Hi,

  Can I get a url(or at least the 2 incorrect words) to the original
content written in datasoft font. 

Regards,
Harshita

On Thu, 2009-02-12 at 11:06 -0400, Golam Mortuza Hossain wrote:
> Hi,
> 
> On Thu, Feb 12, 2009 at 9:26 AM, Harshita <harshita at atc.tcs.com> wrote:
> 
> > May be giving an occurrence of the combination(word -- written in
> > datasoft) will give some more idea.
> 
> 
> Here goes a screenshot.  [ Marked in red are incorrect where as
> marked in green is perfect (for Padma.halffm_SA but
> in different combination) ]
> 
> http://www.math.unb.ca/~ghossain/malformed-conjuncts.jpg
> 
> Hopefully, it will make the issue more clear. BTW,  "SA + halant + BA"
> conjunct is very common in Bengali. Like  "Swadhin" (???????, independent)
> "swabhabhik" (?????????, normal).
> 
> Thanks
> Golam
> _______________________________________________
> Padma mailing list
> Padma at mozdev.org
> https://www.mozdev.org/mailman/listinfo/padma


From gmhossain at gmail.com  Thu Feb 12 07:29:05 2009
From: gmhossain at gmail.com (Golam Mortuza Hossain)
Date: Thu, 12 Feb 2009 11:29:05 -0400
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <1234452172.9737.4.camel@localhost.localdomain>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
	<1234422648.8119.13.camel@localhost.localdomain>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
	<1234445190.8894.20.camel@localhost.localdomain>
	<141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
	<1234452172.9737.4.camel@localhost.localdomain>
Message-ID: <141c154a0902120729k4f90a264pd88afc50b17ba367@mail.gmail.com>

Hi,


>  Can I get a url(or at least the 2 incorrect words) to the original
> content written in datasoft font.

Sure, here it goes (the page from which I took the screenshot)

http://www.aajkaal.net/report.php?hidd_report_id=108867

Cheers
Golam

From vnagarjuna at gmail.com  Thu Feb 12 14:47:10 2009
From: vnagarjuna at gmail.com (Nagarjuna Venna)
Date: Thu, 12 Feb 2009 17:47:10 -0500
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
	<1234422648.8119.13.camel@localhost.localdomain>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
	<1234445190.8894.20.camel@localhost.localdomain>
	<141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
Message-ID: <89531ffe0902121447gc798c3ifed3643a9aada102@mail.gmail.com>

Golam,

Maybe I am missing something obvious but why isn't 'swa' in swabhabhik
not halfform_sa + consnt_ba?

Thanks,
Nagarjuna

On Thu, Feb 12, 2009 at 10:06 AM, Golam Mortuza Hossain
<gmhossain at gmail.com> wrote:
> Hi,
>
> On Thu, Feb 12, 2009 at 9:26 AM, Harshita <harshita at atc.tcs.com> wrote:
>
>> May be giving an occurrence of the combination(word -- written in
>> datasoft) will give some more idea.
>
>
> Here goes a screenshot.  [ Marked in red are incorrect where as
> marked in green is perfect (for Padma.halffm_SA but
> in different combination) ]
>
> http://www.math.unb.ca/~ghossain/malformed-conjuncts.jpg
>
> Hopefully, it will make the issue more clear. BTW,  "SA + halant + BA"
> conjunct is very common in Bengali. Like  "Swadhin" (???????, independent)
> "swabhabhik" (?????????, normal).
>
> Thanks
> Golam
> _______________________________________________
> Padma mailing list
> Padma at mozdev.org
> https://www.mozdev.org/mailman/listinfo/padma
>

From harshita at atc.tcs.com  Thu Feb 12 23:41:41 2009
From: harshita at atc.tcs.com (Harshita)
Date: Fri, 13 Feb 2009 13:11:41 +0530
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <89531ffe0902121447gc798c3ifed3643a9aada102@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
	<1234422648.8119.13.camel@localhost.localdomain>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
	<1234445190.8894.20.camel@localhost.localdomain>
	<141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
	<89531ffe0902121447gc798c3ifed3643a9aada102@mail.gmail.com>
Message-ID: <1234510901.10655.27.camel@localhost.localdomain>

Hi Golam,
 
Mappings sholud be as follows

 1)  Datasoft.conjct_SW = "\u00A6\u00A4"; // Declaring the sequence as
conjunct SW
  
 2) Datasoft.toPadma[Datasoft.conjct_SW] = Padma.consnt_SA +
Padma.vattu_VA ;  // Mapping it to appropriate combination
 
  and
  
  
3)  Datasoft.overloadList["\u00A6"]; // Making the padma to read the
sequence.

 This should solve the problem as far as I believe .

In some fonts individual glyphs (here 00A6 & 00A4) might not be
meaningful, but combination becomes meaningful(i.e SWA in this case).
This is how the designing of converter becomes little tough. I faced the
same situation.

Hope it works.


Regards,
Harshita



On Thu, 2009-02-12 at 17:47 -0500, Nagarjuna Venna wrote:
> Golam,
> 
> Maybe I am missing something obvious but why isn't 'swa' in swabhabhik
> not halfform_sa + consnt_ba?
> 
> Thanks,
> Nagarjuna
> 
> On Thu, Feb 12, 2009 at 10:06 AM, Golam Mortuza Hossain
> <gmhossain at gmail.com> wrote:
> > Hi,
> >
> > On Thu, Feb 12, 2009 at 9:26 AM, Harshita <harshita at atc.tcs.com> wrote:
> >
> >> May be giving an occurrence of the combination(word -- written in
> >> datasoft) will give some more idea.
> >
> >
> > Here goes a screenshot.  [ Marked in red are incorrect where as
> > marked in green is perfect (for Padma.halffm_SA but
> > in different combination) ]
> >
> > http://www.math.unb.ca/~ghossain/malformed-conjuncts.jpg
> >
> > Hopefully, it will make the issue more clear. BTW,  "SA + halant + BA"
> > conjunct is very common in Bengali. Like  "Swadhin" (???????, independent)
> > "swabhabhik" (?????????, normal).
> >
> > Thanks
> > Golam
> > _______________________________________________
> > Padma mailing list
> > Padma at mozdev.org
> > https://www.mozdev.org/mailman/listinfo/padma
> >
> _______________________________________________
> Padma mailing list
> Padma at mozdev.org
> https://www.mozdev.org/mailman/listinfo/padma


From gmhossain at gmail.com  Sat Feb 14 06:32:31 2009
From: gmhossain at gmail.com (Golam Mortuza Hossain)
Date: Sat, 14 Feb 2009 10:32:31 -0400
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <1234510901.10655.27.camel@localhost.localdomain>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
	<1234422648.8119.13.camel@localhost.localdomain>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
	<1234445190.8894.20.camel@localhost.localdomain>
	<141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
	<89531ffe0902121447gc798c3ifed3643a9aada102@mail.gmail.com>
	<1234510901.10655.27.camel@localhost.localdomain>
Message-ID: <141c154a0902140632j2e6c9066q897e7c9512aaa057@mail.gmail.com>

Thanks Harshita,

It worked!

Though, its a bit costly solution :-( It has already cost
me 80 lines of code to take care of all such combination
in this font alone!

I also tried a one-liner solution (simply replacing
"halant + halant" => "halant" after Padma convert).
It worked too. However, Padma seems to push
prefixList entries two steps down in such situation and thus
creating another issue!!

So at present, your solution seems to be the better one.

The fontmap is now almost ready for submission to Padma.
I will watch it for next couple of days to see any omission.

Once this is included, Padma will support three out Big Four
Bengali dailies published from West Bengal.

Thanks,
Golam







On Fri, Feb 13, 2009 at 3:41 AM, Harshita <harshita at atc.tcs.com> wrote:
> Hi Golam,
>
> Mappings sholud be as follows
>
>  1)  Datasoft.conjct_SW = "\u00A6\u00A4"; // Declaring the sequence as
> conjunct SW
>
>  2) Datasoft.toPadma[Datasoft.conjct_SW] = Padma.consnt_SA +
> Padma.vattu_VA ;  // Mapping it to appropriate combination
>
>  and
>
>
> 3)  Datasoft.overloadList["\u00A6"]; // Making the padma to read the
> sequence.
>
>  This should solve the problem as far as I believe .
>
> In some fonts individual glyphs (here 00A6 & 00A4) might not be
> meaningful, but combination becomes meaningful(i.e SWA in this case).
> This is how the designing of converter becomes little tough. I faced the
> same situation.
>
> Hope it works.
>
>
> Regards,
> Harshita
>
>
>
> On Thu, 2009-02-12 at 17:47 -0500, Nagarjuna Venna wrote:
>> Golam,
>>
>> Maybe I am missing something obvious but why isn't 'swa' in swabhabhik
>> not halfform_sa + consnt_ba?
>>
>> Thanks,
>> Nagarjuna
>>
>> On Thu, Feb 12, 2009 at 10:06 AM, Golam Mortuza Hossain
>> <gmhossain at gmail.com> wrote:
>> > Hi,
>> >
>> > On Thu, Feb 12, 2009 at 9:26 AM, Harshita <harshita at atc.tcs.com> wrote:
>> >
>> >> May be giving an occurrence of the combination(word -- written in
>> >> datasoft) will give some more idea.
>> >
>> >
>> > Here goes a screenshot.  [ Marked in red are incorrect where as
>> > marked in green is perfect (for Padma.halffm_SA but
>> > in different combination) ]
>> >
>> > http://www.math.unb.ca/~ghossain/malformed-conjuncts.jpg
>> >
>> > Hopefully, it will make the issue more clear. BTW,  "SA + halant + BA"
>> > conjunct is very common in Bengali. Like  "Swadhin" (???????, independent)
>> > "swabhabhik" (?????????, normal).
>> >
>> > Thanks
>> > Golam
>> > _______________________________________________
>> > Padma mailing list
>> > Padma at mozdev.org
>> > https://www.mozdev.org/mailman/listinfo/padma
>> >
>> _______________________________________________
>> Padma mailing list
>> Padma at mozdev.org
>> https://www.mozdev.org/mailman/listinfo/padma
>
> _______________________________________________
> Padma mailing list
> Padma at mozdev.org
> https://www.mozdev.org/mailman/listinfo/padma
>

From haridallas at gmail.com  Sat Feb 14 09:30:47 2009
From: haridallas at gmail.com (hariharan ramamurthy)
Date: Sat, 14 Feb 2009 09:30:47 -0800 (PST)
Subject: [Padma] hariharan wants to add you as a friend :)
Message-ID: <20090214173047.05E09110043@web0.grouply.com>

I want to add you as a friend in Grouply so you can see my profile with my
pictures, my groups, and my favorite group messages. 

  Here is the link: 
http://www.grouply.com/register.php?r=2813209&amp;vt=52416737  

  hariharan 

  

  ========================== 
   Block Grouply Invites
[http://www.grouply.com/unsubscribe.php?ea=b6x7ujLehpgARS28FJkgBA%3D%3D]  | 
Mark as SPAM
[http://www.grouply.com/unsubscribe.php?ea=b6x7ujLehpgARS28FJkgBA%3D%3D]  

  Click  here
[http://www.grouply.com/unsubscribe.php?ea=b6x7ujLehpgARS28FJkgBA%3D%3D]  to
block all emails from Grouply, 495 Seaport Court, Suite 103, Redwood City, CA
94063. 

  


From gmhossain at gmail.com  Sun Feb 15 05:05:07 2009
From: gmhossain at gmail.com (Golam Mortuza Hossain)
Date: Sun, 15 Feb 2009 09:05:07 -0400
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <89531ffe0902121447gc798c3ifed3643a9aada102@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<89531ffe0902101259q279d2a7au885841845c3f8091@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
	<1234422648.8119.13.camel@localhost.localdomain>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
	<1234445190.8894.20.camel@localhost.localdomain>
	<141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
	<89531ffe0902121447gc798c3ifed3643a9aada102@mail.gmail.com>
Message-ID: <141c154a0902150505k76a2dce7t34cfc4b0e5ed8688@mail.gmail.com>

Hi Nagarjuna,

On Thu, Feb 12, 2009 at 6:47 PM, Nagarjuna Venna <vnagarjuna at gmail.com>
> Maybe I am missing something obvious but why isn't 'swa' in swabhabhik
> not halfform_sa + consnt_ba?

The reason for this is that the font doesn't have a separate glyph for
"swa". Rather, they are using two separate glyphs for "halffm_SA"
and "vattu_BA" to write "Swa".

As fas as visual appearance of these Bengali conjuncts are concerned, it
make sense (at least to me) to express them as "halffm_XX + vattu_YY"
where "halffm_XX" denotes upper half-letter and "vattu_YY" lower
half-letter.

Harshita has pointed out a solution to me, by overloading the first
glyph and reading the conjunct as a whole. It works!

However, this becomes costly when you have 100s of such
combination. Let me give an example, the next font that I
am working on ("TonnyBanglaMJ") has  10 "halffm"s and
25 "vattu"s (including duplicate maps).

So this font could require as much as  500 lines (10X25 + 10X 25 + 10)
of additional codes to handle such conjuncts. I have written
a script to generate Padma fontmaps for such combination
automatically. So its not a big deal for me anymore. However,
I believe, there may be a better way of handling this issue.

Nagarjuna, I am now interested to hear your opinion on this.  Would
it be difficult to expand the Padma parser code to deal with
such situation? Otherwise,  I will follow the above solution.

Thanks for your reply.

Best
Golam

From harshita at atc.tcs.com  Mon Feb 16 01:36:52 2009
From: harshita at atc.tcs.com (Harshita)
Date: Mon, 16 Feb 2009 15:06:52 +0530
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <141c154a0902140632j2e6c9066q897e7c9512aaa057@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<141c154a0902101408p73cbbd2evf965b808474c57cd@mail.gmail.com>
	<89531ffe0902110735m4ce263e9w9b8f7b472bd54297@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
	<1234422648.8119.13.camel@localhost.localdomain>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
	<1234445190.8894.20.camel@localhost.localdomain>
	<141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
	<89531ffe0902121447gc798c3ifed3643a9aada102@mail.gmail.com>
	<1234510901.10655.27.camel@localhost.localdomain>
	<141c154a0902140632j2e6c9066q897e7c9512aaa057@mail.gmail.com>
Message-ID: <1234777012.6274.34.camel@localhost.localdomain>

Hi,

> Though, its a bit costly solution :-( It has already cost
> me 80 lines of code to take care of all such combination
> in this font alone!

Why is it so?? 

One more question Is the glyph at location "00A6" is halfform SA in
bengali?  Does it have a proper identity in the bengali script??

Regards,
Harshita




From gmhossain at gmail.com  Mon Feb 16 07:24:22 2009
From: gmhossain at gmail.com (Golam Mortuza Hossain)
Date: Mon, 16 Feb 2009 11:24:22 -0400
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <1234777012.6274.34.camel@localhost.localdomain>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<141c154a0902111522h41d34739x6374c70865fcab30@mail.gmail.com>
	<1234422648.8119.13.camel@localhost.localdomain>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
	<1234445190.8894.20.camel@localhost.localdomain>
	<141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
	<89531ffe0902121447gc798c3ifed3643a9aada102@mail.gmail.com>
	<1234510901.10655.27.camel@localhost.localdomain>
	<141c154a0902140632j2e6c9066q897e7c9512aaa057@mail.gmail.com>
	<1234777012.6274.34.camel@localhost.localdomain>
Message-ID: <141c154a0902160724h5a5eb262pfa387d7aee43a359@mail.gmail.com>

Hi Harshita,


On Mon, Feb 16, 2009 at 5:36 AM, Harshita <harshita at atc.tcs.com> wrote:
>> Though, its a bit costly solution :-( It has already cost
>> me 80 lines of code to take care of all such combination
>> in this font alone!
>
> Why is it so??

Because, in this font there are "halffm_SA", "halffm_MA", "halffm_SHA",
"halffm_SSA", "halffm_NA", "halffm_GA"... to make it worst some of these are
defined more than once!

Similarly, there are vattus like  "vattu_BA", "vattu_RA", "vattu_MA",
"vattu_LA",
"vattu_NA",  "vattu_NNA", "vattu_PA",  "vattu_KA" ...and also some of
these are defined several times.

Now, the no of combination (halffm_XX + vattu_YY) =  6X8  = 48.
So to write them in Padma, you need 48 + 48 + 6 = 98 lines.


> One more question Is the glyph at location "00A6" is halfform SA in
> bengali?  Does it have a proper identity in the bengali script??

Yes!

Also, this font has a TTF file.

http://www.aajkaal.net/downloads/DATAN___.TTF

So please check the glyph corresponding to A6. I believe the
glyph is standard in Bengali and widely used.

This issue is also present in all six fonts that I am currently
working on.

Thanks,
Golam

From gmhossain at gmail.com  Thu Feb 19 13:18:32 2009
From: gmhossain at gmail.com (Golam Mortuza Hossain)
Date: Thu, 19 Feb 2009 17:18:32 -0400
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <1234855789.10173.5.camel@localhost.localdomain>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
	<1234445190.8894.20.camel@localhost.localdomain>
	<141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
	<89531ffe0902121447gc798c3ifed3643a9aada102@mail.gmail.com>
	<1234510901.10655.27.camel@localhost.localdomain>
	<141c154a0902140632j2e6c9066q897e7c9512aaa057@mail.gmail.com>
	<1234777012.6274.34.camel@localhost.localdomain>
	<141c154a0902160724h5a5eb262pfa387d7aee43a359@mail.gmail.com>
	<1234855789.10173.5.camel@localhost.localdomain>
Message-ID: <141c154a0902191318u7aca10b7md5a6af10f797172f@mail.gmail.com>

Hi Harshita,

On Tue, Feb 17, 2009 at 3:29 AM, Harshita <harshita at atc.tcs.com> wrote:
>  Currently I am working on milleniumvarunweb font, I am facing a similar
> situation.
>
>  One more doubt, we are mapping 00A4 to halffm_SA. If the site is using
> 00A4 only in conjunction with half forms such as halffm_WA, making 00A4
> to consonant Va 'll reduce this overhead.
>
> If they are using 00A4 to write halffm_SA also(some times they might not
> use) we can't avoid writing all the mappings.

I tried it out as you suggested. Indeed, it reduces some of the overhead.
But as you have anticipated, they are also using the same halffm
in combination of "consnt". So it also creates some new conjunct
issues :-(

How are you planning to handle such situation for milleniumvarunweb font?

Also how difficult do you think, is to enhance the current Padma parser
such that it can re-map "halffm_XX + vattu_YY" combination to
"consnt_XX + vattu_YY" on-the-fly?

Thanks,
Golam

From harshita at atc.tcs.com  Thu Feb 26 05:07:15 2009
From: harshita at atc.tcs.com (Harshita)
Date: Thu, 26 Feb 2009 18:37:15 +0530
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <141c154a0902191318u7aca10b7md5a6af10f797172f@mail.gmail.com>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<141c154a0902120351u2ea4b609xe0b4410d6931499b@mail.gmail.com>
	<1234445190.8894.20.camel@localhost.localdomain>
	<141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
	<89531ffe0902121447gc798c3ifed3643a9aada102@mail.gmail.com>
	<1234510901.10655.27.camel@localhost.localdomain>
	<141c154a0902140632j2e6c9066q897e7c9512aaa057@mail.gmail.com>
	<1234777012.6274.34.camel@localhost.localdomain>
	<141c154a0902160724h5a5eb262pfa387d7aee43a359@mail.gmail.com>
	<1234855789.10173.5.camel@localhost.localdomain>
	<141c154a0902191318u7aca10b7md5a6af10f797172f@mail.gmail.com>
Message-ID: <1235653635.6249.10.camel@localhost.localdomain>

Hi Golam,

> How are you planning to handle such situation for milleniumvarunweb font?

I had to write mappings for every such occurrence. 

> Also how difficult do you think, is to enhance the current Padma parser
> such that it can re-map "halffm_XX + vattu_YY" combination to
> "consnt_XX + vattu_YY" on-the-fly?

I am not sure about the difficulty of enhancing, but I think mapping
"halffm_XX + vattu_YY" to "consnt_XX + vattu_YY" 'll be logically
incorrect.

Regards,
Harshita




From gmhossain at gmail.com  Thu Feb 26 17:51:14 2009
From: gmhossain at gmail.com (Golam Mortuza Hossain)
Date: Thu, 26 Feb 2009 21:51:14 -0400
Subject: [Padma] Malformed Conjuncts
In-Reply-To: <1235653635.6249.10.camel@localhost.localdomain>
References: <141c154a0902081155u123321c8t12bfa29484b448c0@mail.gmail.com>
	<141c154a0902120706n3886ba29w35a35e39cf4bdf17@mail.gmail.com>
	<89531ffe0902121447gc798c3ifed3643a9aada102@mail.gmail.com>
	<1234510901.10655.27.camel@localhost.localdomain>
	<141c154a0902140632j2e6c9066q897e7c9512aaa057@mail.gmail.com>
	<1234777012.6274.34.camel@localhost.localdomain>
	<141c154a0902160724h5a5eb262pfa387d7aee43a359@mail.gmail.com>
	<1234855789.10173.5.camel@localhost.localdomain>
	<141c154a0902191318u7aca10b7md5a6af10f797172f@mail.gmail.com>
	<1235653635.6249.10.camel@localhost.localdomain>
Message-ID: <141c154a0902261751h7e7b1fb9p5ed9f4c44c2a5dd3@mail.gmail.com>

Hi Harshita,

On Thu, Feb 26, 2009 at 9:07 AM, Harshita <harshita at atc.tcs.com> wrote:
>> Also how difficult do you think, is to enhance the current Padma parser
>> such that it can re-map "halffm_XX + vattu_YY" combination to
>> "consnt_XX + vattu_YY" on-the-fly?
>
> I am not sure about the difficulty of enhancing, but I think mapping
> "halffm_XX + vattu_YY" to "consnt_XX + vattu_YY" 'll be logically
> incorrect.

Could you please explain bit more? I am trying to understand the
problems with such remapping so that we can have a better
approach in dealing with such combinations.

Thanks,
Golam