when ports are changed, set both egress and untagged... a switch
reset/sets the untagged bit upon egress change...

The Lenovo's SNMP get is broken w/ multiple OIDs, so add a hack
to detect that the switch is a Lenovo, and don't use getmany (well
call getmany_real w/ one OID at a time).

to 10 seconds to deal w/ slow switches, but also puts in a proper
state machine for creating vlans.  This deals w/ the fact that
this switch can not create the vlan and activate the vlan in a
single command...

more weirdness on how to get the current set of vlans

This can be used to set the STP priority for the root switch..

Also, fixup and clean up the tests so that the pass...  minor edits
to support my custom config, and drop a couple tests as those are
tested in another test...

it's own..  also make sure getegress fully works.. more tests need
to be written..

subdivide the query into smaller chunks..   perfect example of how
the pysnmp hlapi isn't that high level...

we can split it, do so..  This happened w/ a switch when using SNMPv3..
need to write a test case for this...

apply them..

test cases..

so that we don't have to make the bits the same length as the switch..

so we need to import and use that one...

catch more cases for only tagged ports..

compare the switch's port list against the definition to make sure
that there is a complete definition...

Add the ability to skip the switch tests that can take a while to
complete...