[gtkextra] Undesirable behavior in "changed" signal ?

[Victor M. <ko...@em...>]

Hi Adrian

I have a problem with the behaviour of "changed" signal.

I'm working  on 2 projects :

1) a mysql table edit interface who needs a behaviour like this:
when i click on  other cell (DEACTIVATING current cell) i want a signal to be 
emitted in case of changing the cell content so i can update mysql table.
if cell text is not changed i don't want any signal 

2) A gtksheet cell pattern matching using regexp .
Here i need this behaviour:
   a signal to be emitted when I press a key in cell(ALMOST like change signal
does).
  
The problem is that the "changed" signal combines these 2 behaviour types.
I mean it's emittd when i press a key in cell and at the DEACTIVATION of the
cell is   emitted again , also nothing is changed from the last key press.

I think these 2 behaviour types should not be treated by the same signal.
They are different. 
On the other side , i don't understand why you created "cleared" signal,
which is in fact a particular case of "changed" signal:

    void changed(Widget,row,col){
          gchar *str;

          str=gtk_sheet_get_cell_text(widget,row,col)
           if( !str) { // cleared cell }
      }

I'm sending this mail today , because i didnt know until tonight that there is 
a " changed" signal. I wrote the gtkextra manual for so many weeks 
and i didnt know this signal  exists :-))

In mysql interface i implemented the behaviour 1) , but it's not  a trivial task.
I think in many table editing programs i'll need behaviour 1)


Could you enlight me , please?

Thanks very much

PS : Could you not CC me anymore , please?
       I'm already subscribed to the list and all youi mails arrive twice.
-- 
_$%&^%$&*#)(*$)#*%$*#&%*$#)#*$)#
RUK;
#!/bin/ssh
#The Unix Guru's View of Sex
unzip ; strip ; touch ; grep ; finger ; mount ; fsck; more ; yes ; umount ; sleep

-------
Mailuri? Gramada! La www.emoka.ro iti faci moka
un cont de mail cu 12 MB spatiu de stocare!