Pandas getting sqlalchemy datetime type ignores None timezones

When converting to datetime64, pandas does:

        if col_type == 'datetime64': 
            try:
                tz = col.tzinfo
                return DateTime(timezone=True)
            except:
                return DateTime
        if col_type == 'timedelta64':

2 problems 1. The exception catch is way to broad. I think this was probably supposed to catch an AttributeException (For col with no tzinfo attribute at all) 2. columns can have a tzinfo =None, in which case a Datetime(timezone=False) should be created instead.

Comment From: jorisvandenbossche

Did you have an issue with this with an actual case? Because I don't think this code is working at all, as columns (series) have never a tzinfo attribute.

I also mentioned this in #9086, the issue for timezone support in the sql functions.

Comment From: mangecoeur

This didn't trigger an actual bug, its more that i was looking through other issues and i noticed this bit of code didnt make sense....

Comment From: jorisvandenbossche

It indeed does not make much sense .. I am going to close this, as there is already #9086 to track this.