My Comments at the bottom :-)

Jorge Marques Pelizzoni wrote:

The grammar checker already exists and implements "auto" checking in two text
processors to no significant performance loss. If this can not be reproduced in
OOo, that will be solely OOo's fault.
Cheers,

Jorge.

Quoting Christian Junker <[EMAIL PROTECTED]>:



To some degree I am pretty sure the UNO API has the power to allow you
a lot of things for such a grammar add-on.
I don't want to give you advice here, but I know of one attempt to
produce such a grammar checker in Python, what came out made
OpenOffice run very slow. So if you and others around you in Brazil
are concerned about OOo's speed in terms of simple usage, then
consider that it would probably be a lot of work writing such a
grammar checker and it would probably slow down the office program
considerably.
That's just my opinion and input to you, whether you take that into
consideration is up to you.

In case you're going to create such a grammar checker, I will happily
be of help if I can be in somewhat way.

Best Regards
Christian Junker

On Wed, 12 Jan 2005 12:37:11 -0200, Jorge Marques Pelizzoni
<[EMAIL PROTECTED]> wrote:


I represent a company that produces a grammar-checker for Brazilian


Portuguese


and have been comissioned a way to integrate it with OpenOffice.org. I've


done


my basic homework and seen lots of macro tricks, but am up to something a


bit


more sophisticated, like "auto" grammar checking (i.e. checking as you


type).


Basically, a have one single question now (and a further one, depending on


the


answer), as follows:

With OOo's current build (ok, take also version 2.0 into consideration if


need


be) and just that, can such a grammmar checker (i.e. involving concurrent
scanning and processing, "curly" underlining in different colours of


arbitrary


text ranges, associating such ranges with specific pop-up menus, etc.) be
implemented as a UNO something? I mean does the _current_ API and


architecture


support all that, even if it took lots of effort and expertise?

If so, that would make prototyping a grammar checker API much easier till


we


finally got to something worth integrating into the core, if ever.

IF NOT, what is the _minimal_ set of features the current API lacks to


enable


that? You see: I am not talking about a grammar checker API, BUT API


extensions


that would allow external prototyping of a grammar checker API. [I realise


the


answer to this question might be somewhat labourious and become a new topic


of


discussion.]

Cheers,

Jorge M. Pelizzoni.


It might be instructive to see the Python grammar checker if you remember anything about it Christian...


If I wanted to write a grammar checker, the first thing that I would do would be to take a look at the code that implements the spelling code in OOo to see how they label words and such. I see the issue broken down into a few pieces...

* How to obtain notification that new stuff has been entered. This is likely to mean a listener of some sort, especially if written using the API and is not integrated into OOo using an API designed just for this.

* How to read the text in the area of interest.

* How to mark things in the text so that they are noticeable as problematic.

Jorge, it sounds to me like you have certainly solved the problem of looking at the text and figuring out grammatically what needs to be done in an efficient manner. The part that is likely to slow you down is watching the input and dealing with the different input types such as tables and such...

--
Andrew Pitonyak
My Macro Document: http://www.pitonyak.org/AndrewMacro.sxw
My Macro Book: http://www.hentzenwerke.com/catalog/oome.htm
Free Info:  http://www.pitonyak.org/oo.php


--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]



Reply via email to