[ZODB-Dev] Fwd: [Zope3-dev] directory hierarchy proposal (Martijn Faassen)

[Casey Duncan]

On Wednesday 11 December 2002 08:04 am, Jeremy Hylton wrote:
> We also ran into another potential problem when trying to merge the
> BTrees into a single module namespace (where zope.btrees.OOBTree is
> a class).  There are a bunch of functions like union() that are
> specific to a particular flavor of btree.  It's not obvious what to do
> with them if they get merged.  Maybe oonunion, but the namespace gets
> pretty fully.

Perhaps they *should* be exposed as single functions which front for spec=
ific=20
implementations for each BTree type. That is the "union" function checks =
its=20
operand and calls the proper btree union code. OTOH making them methods o=
n=20
the objects would work too I think.

This also brings up another point that I recently encountered (vicariousl=
y=20
through Shane). It is currently quite difficult for an application to dis=
cern=20
whether an object is from the II, IO or OO families. Perhaps a common=20
abstract base class for each btree type would be helpful...

-Casey