Re: [Ankur-core] Ankur Dictionary

[Golam M. H. <gmh...@gm...>]

Hi Salahuddin,

On Wed, May 13, 2009 at 2:26 AM, salahuddin66 <sal...@gm...> wrote:
> Yes, you are right. Personally I also think getxml.pl would be a
> standard name. Now BDictXML import BDictSQL and use the existing db
> function.
>
> File attached. Please feel free to let me know if you need any changes. :)


First of all, it is coming out very nicely.  I have one more suggestion.


>                <dict_entry_unedited id="7">


I would prefer not to have above element for the following reason.

The client side parsing would need to have separate code for reading
unedited entries. My suggestion would be to have

<status>UNEDITED</status>

for each entry. Currently, we have four separate status

(1) EDITED
(2) UNEDITED*
(3) OBSOLETE
(4) DELETED   (It helps to undo delete easily by simply changing the status).

Also, there is a unique "id" field for each dictionary entry. In case, you want
to have editing capability in your client app then you need to send
back this id.  So something like this would be good to have

<bdict_id>23232</bdict_id>

for each entry.

Given you are almost finishing up the spec, I would urge you to
write down the final specifications that we can put up in Abhidhan
page.  This will help those who would want to write their own
XML based app using Ankur dictionary.

Here is the current dict_table elements

  `id` int(11) NOT NULL auto_increment,
  `en_word` text,
  `pos_tag` text,
  `en_lemma` text,
  `bn_pronunciation` text,
  `bn_word` text,
  `explanation` text,
  `example` text,
  `status` text,


Finally, please make sure  to put your name in the license
headers and in the final XML specs.

Best,
Golam