On Tue, Sep 09, 2008 at 11:33:08PM +0200, Abdelrazak Younes wrote:
> Vincent van Ravesteijn - TNW wrote:
>> For the moment, forget this cur.setSelection() part. It is still not
>> correct probably. Although I don't encounter the problem anymore, it
>> might not be correct to call it in cur.setSelection(...). Maybe it
>> breaks down something else.
>>
>> cs 26146 did solve a same problem, but then the selection is removed
>> when the mouse is released. Iin this case however, the mouse isn't
>> released yet, so the empty selections still exist. Maybe we should add
>> it to line 1234 in Text3.cpp ?
>>
>
> I am not sure I understand. I'll let JMarc handle this part.
>
>> Consider it work in progress..
>>
> Still some comments:
>
> is_beg_pit and is_end_pit: please rename those to something more
> meaningful related to the selection state. IIUC, maybe 'is_sel_begin'
> and 'is_sel_end'?
>
> + char_type const bb[] = { row.sel_beg == 0 && is_beg_pit ? 1 : 0 +
> + row.sel_end == par_->size() && is_end_pit ? 2 : 0 +
> + row.sel_end == row.endpos() && sel_end_pos == row.endpos() ? 4
> : 0 };
>
Incitentalluy, since this is only a single char this might as well look
like
char_type const bb = row.sel_beg == 0 && is_beg_pit ? 1 : 0
+ row.sel_end == par_->size() && is_end_pit ? 2 : 0
+ row.sel_end == row.endpos() && sel_end_pos == row.endpos() ? 4 : 0;
.... foo(&bb, 1);
Look already a bot less convoluted.
char_type bb = (row.sel_beg == 0 && is_beg_pit);
bb |= (row.sel_end == par_->size() && is_end_pit) << 1;
bb |= (row.sel_end == row.endpos() && sel_end_pos == row.endpos()) << 2;
Hm. Not much difference.
char_type bb = 0;
if (row.sel_beg == 0 && is_beg_pit)
bb += 1;
if (row.sel_end == par_->size() && is_end_pit)
bb += 2;
if (row.sel_end == row.endpos() && sel_end_pos == row.endpos())
bb += 4;
might make the intention mor obvious, even if it's a bit longer.
Andre'
