when presented with duplicate columns.  The behavior of emitting a
warning and replacing the old column with the same name still
remains to some degree; the replacement in particular is to maintain
backwards compatibility.  However, the replaced column still remains
associated with the ``c`` collection now in a collection ``._all_columns``,
which is used by constructs such as aliases and unions, to deal with
the set of columns in ``c`` more towards what is actually in the
list of columns rather than the unique set of key names.  This helps
with situations where SELECT statements with same-named columns
are used in unions and such, so that the union can match the columns
up positionally and also there's some chance of :meth:`.FromClause.corresponding_column`
still being usable here (it can now return a column that is only
in selectable.c._all_columns and not otherwise named).
The new collection is underscored as we still need to decide where this
list might end up.   Theoretically it
would become the result of iter(selectable.c), however this would mean
that the length of the iteration would no longer match the length of
keys(), and that behavior needs to be checked out.
fixes #2974
- add a bunch more tests for ColumnCollection

implemented right before the release of 0.9.3 affected the case where
a UNION contained nested joins in it.   "Join rewriting" is a feature
with a wide range of possibilities and is the first intricate
"SQL rewriting" feature we've introduced in years, so we're sort of
going through a lot of iterations with it (not unlike eager loading
back in the 0.2/0.3 series, polymorphic loading in 0.4/0.5). We should
be there soon so thanks for bearing with us :).
fixes #2969   re: #2967
- solve the issue of join rewriting inspecting various types of
from objects without using isinstance(), by adding some new
underscored inspection flags to the FromClause hierarchy.

oriented row lookups were not matching up to the ad-hoc :class:`.ColumnClause`
objects that :class:`.TextAsFrom` generates, thereby making it not
usable as a target in :meth:`.Query.from_statement`.  Also fixed
:meth:`.Query.from_statement` mechanics to not mistake a :class:`.TextAsFrom`
for a :class:`.Select` construct.  This bug is also an 0.9 regression
as the :meth:`.Text.columns` method is called to accommodate the
:paramref:`.text.typemap` argument. [ticket:2932]

- get PG dialect to work around "no nonexistent binds" rule for now,
though we might want to reconsider this behavior

More fixes for cross references and reducing warnings (3rd wave)