On Wed, 2011-06-29 at 22:24 +0200, Markus Mohrhard wrote: > Kohei, I think, as this is mostly your area, you might want to have a > closer look at the whole code around the offset problem.
So, everything comes down to what PivotField::nCol is expected to contain; is it a (always 0-based) dimension index, or a column index whose smallest index may not be always 0. The column offset variable was there because PivotField::nCol was expected to contain a column index rather than a dimension index if the data source was a sheet, while for the other data source types it was expected to contain a dimension index. First things first, Bubli's fix appears to work just fine. This is all good. Now, that would also imply that in the latest pivot table code, PivotField::nCol always contain a dimension index even for the internal sheet data source type. I remember I fixed some bugs some time ago in this area because nCol didn't contain a column index, so I know for a fact that this was not always the case. Maybe my re-work of the dialog changed this, my cache table re-work did this. I don't know for sure. But I have to keep this fact in mind going forward, in case I come across any old code that may still expect PivotField::nCol to contain a column index. But that's fine since I personally prefer it always containing a dimension index anyway. :-) Regards, Kohei -- Kohei Yoshida, LibreOffice hacker, Calc <kyosh...@novell.com> _______________________________________________ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
