To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=45843


User jodygoldberg changed the following:

                  What    |Old value                 |New value
================================================================================
                    Status|NEW                       |STARTED
--------------------------------------------------------------------------------




------- Additional comments from [EMAIL PROTECTED] Mon Feb 27 07:43:05 -0800 
2006 -------
re: CONTINUE handling.
blah my other tests were working smoothly I'll examine this now to find the
misunderstanding.

> different summary function or "displayed value" is selected
The XL variant does not support custom agregation routines.  There seems to have
been some attempt but it does not work as far as I can tell.  By definition it
returns the sum.  As an extension we could consider supporting using the field
aggregator, or allowing user specified aggregation in the target field
specifier, but that would be an extension.

> Getpivotdata should be limited to values that are actually part of the
> DataPilot table
This is an interesting point for debate.  XL apparently has this restriction.  A
filter that has no visible subtotal will return a REF, but that seems like a
pointless limitation.

> Re: Refreshing after XL import.
I realize it is a kludge.  We're being hit with several layers of problem, that
was a solution that will solve some of them at the expense of others.
1) Having GetPivotData depend on ScDPOutput means that it is pulling values from
the _generated_ result.  On xls import we do not appear to be generating based
on the cached data and instead use the in sheet.  Throw in a few other bugs and
the DPOutput has a tenuous link to what is on the screen until we ::Output.  The
re are two potential solutions for GETPIVOTDATA.  Either pull the data off the
sheet, which seems impossibly kludgy (there is no layout information), or ensure
that the DPOutput is generated using the cached values with a matching
aggregation approach to Ms XL.  The latter seems best, but will take some time. 
 

2) GetPivotData is being calculated before the pivot is loaded and generating a
#REF.  We need to trigger a recalc for things that use them.  I looked at two
mechanisms.
    a) Keep track of all formulas that use GETPIVOTDATA on xls import and recalc
them after loading the pivots.  This became unyeildy as all of the potential
callers were considered (cells, validation, names).
    b) mark the content of the pivot table as being dirty to force all things
that reference it to recalc.  This is less clean, but much simpler.  It is also
a useful side effect of regeneration.

3) It was the easiest way to do what RefreshAfterLoad + SetButtons does for the
ods importer. Although it does lose us the autoformating.

Solving these problems will take some time.  For now, I'm advocating the 1 line
sledge hammer of ScDPObject::Output.

---------------------------------------------------------------------
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to