BUG: fix Series repr when name is tuple holding non string-type

[lodagro]

Not sure if this is in line with latest unicode work, @y-p?

[ghost]

on git master the test fails with

File "src/pandas/pandas/core/format.py", line 88, in _get_footer
series_name = "('%s')" % "', '".join(self.series.name)
TypeError: sequence item 0: expected string, int found

the test in b837101 already passes with (PR #2005).
as a side note, the PR excises stringify and stringify_seq in favor of a new pprint_thing(),
which should handle more corner-cases and is clearer about unicode/str issues.
a lot of that PR is migrating things over to using that, so it'd be nice if they weren't reintroduced by
new patches if possible.

[ghost]

I vote to keep the test of course.

[lodagro]

So what do we do? Is it ok that you add the test in #2005?

[ghost]

Done. 11dff0d
note that I rewrite the PR as i go along to keep things tidy, so the SHA may change.

[lodagro]

Ok, closing this one as it is redundant now.
Hopefully #2005 gets merged in soon, which will also close #2051.

[wesm]

Thanks. I'll review and merge #2005 as soon as I can