Re: Autocorrect functionality

On 26/10/22 09:24, Dag wrote:
> On onsdag den 19. oktober 2022 17.42.29 CEST, Volker Krause wrote:
>> On Mittwoch, 19. Oktober 2022 01:05:19 CEST Albert Astals Cid wrote:
>>> Sorry for the cross-posting.
>>> Please CC me, i'm on none of the two lists
>>>
>>>
>>> Both pimcommon and calligra have Autocorrect functionality, but PIM one
>>> broke at some point and I guess no one realized?
>>
>> [...]
>>
>>> Solutions:
>>>   A) Make PIM look for files also in GenericDataDir/calligra and live with
>>> the dependency to calligra for the feature B) Accept that the feature in
>>> PIM was broken for a long time and just remove it C) Make a library that
>>> both PIM and Calligra use
>>>
>>> Ideally we would go with C but we all know it's not like we have lots of
>>> people with time available to jump and work on this...
>>
>> Right, that's ultimately what I'd like to see as well, all the auto-
>> correction, grammar checking and translation features moving
>> into Frameworks,
>> alongside spell-checking. Those are all features you would like to have
>> everywhere you write more than a few words of text IMHO.
>>
>> But lacking time for this, it's not more than a wishlist item
>> unfortunately...
> I think there are two separate issues here:
> 1) The data, where to host/maintain/install, and
> 2) code duplication. To reduce the amount of work we could postpone this
>     until needs get pressing or sombody finds the time.
> 
> The data we have in calligra atm is not very comprehensive. This makes it
> pretty useless if it is not updated.
> The data in libreoffice is pretty good though, so using that data would be
> nice.
> But how to do that? Could we:
> . Merge current LO data with calligra data,
>      A job to do now, and data must be maintained two places.
> . Copy current data from LO into our repo keeping data format,
>      When data is updated in LO, somebody needs to make a new copy.
>      Install both calligra data and LO data.
> . Copy data automatically from LO at release time keeping data format,
>      Install both calligra data and LO data.
>      
> I don't particulary like any of these, but trying to be practical...
> 

I am not familiar with the AutoCorrect data in Calligra and LO, but sounds to me like LO are keeping 
their data updated/maintained? if so, could we change our parsing to parse/consume their data 
directly? (i.e. like using myspell dictionaries).

Regards,
Ahmad Samir