ERR: concat of 2 Series categories does not preserve category dtype

[jreback]

xref #8640

As this is inconcistent with how DataFrame concat's work (its a different code-path), so needs updating

s = Series(list('abc'),dtype='category')
s2 = Series(list('abd'),dtype='category')

# this should raise
pd.concat([s,s2])

# this should be a category
pd.concat([s,s])

[springcoil]

Any remarks on how big of a job this is? Is it a good first PR because I might have a look then.

[jreback]

ahh, this is very straightforward actually, just a small change. have a look if you'd like