https://bugs.documentfoundation.org/show_bug.cgi?id=157089

--- Comment #8 from Eike Rathke <[email protected]> ---
(In reply to Heiko Tietze from comment #6)
> My first thought was that this is not defined in ODF but it is
No, it is not.

> 9.2.4 Column and Row Range Addresses
> Column and row addresses are cell range addresses that reference entire rows
> or columns.
> 
> The syntax of a row address is the same as a cell address, except the
> alphabetic values that indicate the column are omitted.
> 
> The syntax of a column address is the same as a cell address, except the
> numeric values that indicate the row are omitted.
Note that definition is in the global schema definition. That definition to me
is ambiguous and quite certainly did not want to say that one could mix the
notations of cell range addresses and entire column or row range addresses. It
likely is wrong worded and instead of
"a row address is the same as a cell address"
should read
"a row address is the same as a cell range address"
and then "except the alphabetic values that indicate the column are omitted"
actually would make sense. Similar for column address.

The ODFF formula specification is very explicit about that, see
5.8 References
https://docs.oasis-open.org/office/OpenDocument/v1.3/os/part4-formula/OpenDocument-v1.3-os-part4-formula.html#References

The syntax says you can have either
SheetLocatorOrEmpty '.' Column ':' '.' Column
or
SheetLocatorOrEmpty '.' Row ':' '.' Row
but no mix of
SheetLocatorOrEmpty '.' Column Row ':' '.' Column
or any such construct.


So if at all this would be an UI only feature, which when entered resolves to
the last row or column for the missing part. Note that an input of A:A3
wouldn't make sense and should resolve as A3:A and thus A3:A1048576, if not
error.

-- 
You are receiving this mail because:
You are the assignee for the bug.

Reply via email to