On 06/03/2010 15:42, Vincent van Ravesteijn wrote:
Abdelrazak Younes schreef:
Hello,
I am a bit puzzled about Cursor::getStatus(). This always apply to
the next inset which is not exactly the inset at the cursor. In case
of a math inset in a paragraph in a table, cursor will ask the the
paragraph what is the inset a the next position... this is all quite
weird...
There are LFUNs like LFUN_INSET_MODIFY that have the AtPoint flag.
This means that they either operate on the current inset or the next
one. Compare with the old LFUN_NEXT_INSET_MODIFY which we removed.
Tabular-feature didn't have this flag, but inset-modify does. That's
why we have those new problems with math and tables.
I think it was wrong to use inset-modify as a replacement of
tabular-feature. This because it does not change some setting of the
inset itself, but settings that are related to cursor positions within
this inset.
Does this help you ?
Yes, thanks. In effect, all this means that InsetTableCell should be
treated as real insets and we must transfer some code from InsetTabular
to InsetTableCell. While you are maybe right (and you should have told
me ;-)) that the replacing tabular-feature with inset-modify was a too
big move at this point, I still believe this move is correct. This table
code needed the implosion :-)
I am right now trying to fix the long tabular dialog and it is still
difficult to grasp after my last cleaning up :-(
Abdel.
