to_excel() fails if unicode sheet name is used

[lbeltrame]

Like the subject says, the check in DataFrame.to_excel() checks for an instance of str, rather than basestring, hence unicode names skip the if check and are treated like files, which causes an exception later on in the code. Changing isistance() to check for basestring rather than str fixes the issue.

[lbeltrame]

It looks a little more complicated than expected because in py3 isinstance(foo, str) is fine, but not in py2.

[mforelldavis]

In py2 basestring is the superclass for str and unicode so you can use it to catch both; in py3 we don't have the basestring or unicode types anymore. I would try this:

try: basestring
except NameError: basestring = str #py3
if isinstance(excel_writer, basestring):
...

[changhiskhan]

2to3 should convert basestring to str though no?

[mforelldavis]

Yes, it will, but don't you want this snippet to work regardless of whether it's being run with 2x or 3x?

[wesm]

Everything checking out over here on Python 2 and 3. Closing the issue