dataframe and ischar

[CdeMills]

Hello,

I have a question about ischar applied to dataframe() objects. "dataframes"
try to mimic arrays, in particular when selecting a sub-part where all data
have the same type. In this context, using ischar returns true.

But this annoys me in the context of strcmp and strncmp: those functions
test their arguments with ischar() and iscellstr() and use slightly
different codes in both situations. If I use "ischar" on a dataframe
containing only strings, this succeeds, but the code in do_strcmp_fun fails.
OTOH, with
tt={"one", "two"; "three", "four"}; strcmp("one", tt)
the code uses the branch associated to the "is_cellstr" path. This means
that a cell array of strings is not considered as a char. 

My question is: what's the philosophy of ischar with respect to an object
which tries to mimic a string array but whose engine behind uses cellstr ?
Should ischar return true for singleton element and false for more complex
arrangements ?

TIA

Pascal



--
View this message in context: 
http://octave.1599824.n4.nabble.com/dataframe-and-ischar-tp4668638.html
Sent from the Octave - Maintainers mailing list archive at Nabble.com.