https://bugs.freedesktop.org/show_bug.cgi?id=58838

--- Comment #2 from Owen Genat <[email protected]> ---
Created attachment 78588
  --> https://bugs.freedesktop.org/attachment.cgi?id=78588&action=edit
Example spreadsheet showing comparisons

I am not sure if this constitutes a confirmation of this bug or not. I agree
with mariosv that the EXACT() function is the preferred method for string
comparison. It would seem to me that comparisons via raw equality (equals sign)
of the empty string "" with CHAR() and UNICHAR() output are, for some reason,
variable i.e., dependent upon the character in question being compared.

In the attached spreadsheet there are columns for comparison of CHAR() output
for characters 0 through 256 (one extra to be certain) against a NULL cell, an
empty string cell, and a cell that has been edited (for curiosity sake). This
is repeated again for UNICHAR() output for the same range. A comparison of the
results of these tests is then made i.e., NULL result for CHAR() test against
NULL result for UNICHAR() test.

I have conducted these tests under Ubuntu 10.04 x86_64 running TDF LO v.3.5.7.2
and TDF LO v4.0.2.2. There is no difference in output between the LO versions.
As can be seen in the attachment most empty string tests return a FALSE value
that is the same as that returned for a NULL value. There are some characters
though (0-8, 14-31, 127, 249-251, 254-255 for the CHAR() function and 0-8,
14-31, 127-132, 134-159 for the UNICHAR() function) that return a TRUE value. I
have not tested higher Unicode values.

I am not sure if these characters constitute an edge case, but the behaviour
would certainly appear to be unusual or inconsistent.

-- 
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Libreoffice-bugs mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/libreoffice-bugs

Reply via email to