Re: [office-formula] Missing Excel Functions

[Eike Rathke <erack@sun.com>]

Hi David,

On Tuesday, 2008-12-09 23:22:35 -0500, David A. Wheeler wrote:

> > > > CELL - I found the note in the spec about platform independence.  Some of the values returned by the function are not platform specific.  Does it make sense to add it to the Large group with a note about some of the argument values?
> > > 
> > > It might; what do you have in mind?
> > 
> > CELL is problematic in itself. The definition of the "filename" info
> > type is most controversial, as one would expect it is defined to return
> > the file name, but that is not what Excel and OOo return. For legacy
> > reasons (once upon a time each sheet was a separate file) Excel includes
> > the sheet name, as does OOo. However, Excel does not do the IRI
> > notation. I currently don't see a real solution for this.
> 
> Perhaps we can define CELL for some values, and note that others are
> platform-specific.  I suspect, but do not know, that this is more of a debugging
> function than a function used extensively for interoperability.
> True? False?

Actually it's a legacy function that originally may had derived even
from Lotus, I presume. Some of its functionality is covered by other
functions, such as ROW() or COL(). Other is not, such as formatting
info, which is also problematic because usually changing cell formatting
should not change formula results at all.

I've seen CELL() in the wild being used to obtain the file name with
"filename" by stripping the part following the file name. So defining
"filename" to return only the file name without the sheet name might
even break some of those documents. On the other hand, to be fully
interoperable it would need another parameter to specify which
convention the returned string should follow, (a) Excel's notation plus
sheet name, or (b) IRI notation plus sheet name, or (c) IRI notation
without sheet name, and a proper default.

Which brings me to the point that SHEETNAME([Reference]) and FILENAME()
functions may be useful, which we don't have yet.

  Eike

-- 
Automatic string conversions considered dangerous. They are the GOTO statements
of spreadsheets.  --Robert Weir on the OpenDocument formula subcommittee's list.

PGP signature