RE: export of Calc functions to OOXML

2014-07-31 Thread Winfried Donkers
Hi Eike, (re-introducing the mailing list I left out in my previous reply) >> FormulaGrammar::isOOXML( meGrammar ) only returns true when saving as >> OOXML, and FormulaGrammar::isODFF( meGrammar ) only returns true when saving >> as ODFF. >> Is that correct? > Yes, that's correct. Also note th

Re: export of Calc functions to OOXML

2014-07-30 Thread Eike Rathke
Hi Winfried, On Wednesday, 2014-07-30 08:10:06 +0200, Winfried Donkers wrote: > I note that the PODF-handling is always done when saving, regardless > of the file type being saved (e.g. ods, xlsx). What do you mean? It is only done for .ods but not for .xlsx > The result is that e.g. the formul

RE: export of Calc functions to OOXML

2014-07-29 Thread Winfried Donkers
Hi Eike, > There are already entry points that do something similar for the old non-ODFF > (aka PODF) OOo-XML format, i.e. in formula/source/core/api/FormulaCompiler.cxx > FormulaCompiler::CreateStringFromTokenArray() and in > formula/source/core/api/token.cxx all functions with *Rewrite*() >

Re: export of Calc functions to OOXML

2014-07-21 Thread Eike Rathke
Hi Kohei, On Monday, 2014-07-21 09:13:32 -0400, Kohei Yoshida wrote: > > BUT, I think I might know why it doesn't get called anymore ... the > > OOXML formula string is now directly generated from the token array, > > which bypasses all mechanisms we had in place for the Excel binary > > export a

Re: export of Calc functions to OOXML

2014-07-21 Thread Kohei Yoshida
On Fri, 2014-07-18 at 17:54 +0200, Eike Rathke wrote: > BUT, I think I might know why it doesn't get called anymore ... the > OOXML formula string is now directly generated from the token array, > which bypasses all mechanisms we had in place for the Excel binary > export and the earlier OOXML exp

RE: export of Calc functions to OOXML

2014-07-21 Thread Winfried Donkers
Hi Eike, >> It seems to me that re-implementing the OOXML export functionality is rather >> a complex business, i.e. not something suitable for me. Shall I create a bug >> report for this? > [ ... ] > If you want to take a stab at this just tell me, otherwise I'll do. > A bug report might be g

Re: export of Calc functions to OOXML

2014-07-21 Thread Eike Rathke
Hi Winfried, On Monday, 2014-07-21 11:39:30 +0200, Winfried Donkers wrote: > Neither XclExpFmlaCompImpl::AppendTrailingParam(), nor > XclExpFmlaCompImpl::FinishFunction() gets hit when set as a breakpoint in gdb. That confirms my assumption that it is due to the change that formula expressions

RE: export of Calc functions to OOXML

2014-07-21 Thread Winfried Donkers
Hi Eike, >> The Calc function GAMMADIST has an optional 4th argument (cumulative), >> whereas this argument is mandatory for Excel. >> When saving a document with function GAMMADIST as xls, this argument is >> added if necessary (in /sc/source/filter/excel/xeformula.cxx). >> When saving this doc

Re: export of Calc functions to OOXML

2014-07-18 Thread Eike Rathke
Hi Winfried, On Tuesday, 2014-07-15 12:28:35 +0200, Winfried Donkers wrote: > The Calc function GAMMADIST has an optional 4th argument (cumulative), > whereas this argument is mandatory for Excel. > When saving a document with function GAMMADIST as xls, this argument is added > if necessary (in

export of Calc functions to OOXML

2014-07-15 Thread Winfried Donkers
Hi, The Calc function GAMMADIST has an optional 4th argument (cumulative), whereas this argument is mandatory for Excel. When saving a document with function GAMMADIST as xls, this argument is added if necessary (in /sc/source/filter/excel/xeformula.cxx). When saving this document as xlsx, this