OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

office-formula message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [office-formula] ISERROR(undefinedname) behavior


Eike Rathke:
> There is a special case with the ISERROR function when used with an
> undefined name, as in =ISERROR(undefinedname)
> 
> Excel and Gnumeric return TRUE, while Calc returns a #NAME? error
> because the error is already propagated at compile time when
> encountering 'undefinedname' and the formula is never interpreted.
> Kspread 1.5.0 returns FALSE, which obviously is wrong.
> 
> Do we want to require that ISERROR still returns TRUE in this case?

I think we should.  I would expect that an undefined error name would be mapped to SOME error in the given implementation.  (It has to map to SOMETHING, right?)  No matter what it gets mapped to, it would be mapped to an error, so ISERROR( some-error ) should be true.

In theory, we could declare the whole thing to be a syntax error if it's an unrecognized error name.  But that would extremely discourage any effort by any implementation to have more specific error names; in practice, you'd have to have a fixed set of error names that could never be (practically) extended.  Excel is already limited that way; there's no need for ODF to be so limited.

--- David A. Wheeler


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]