GitLab

- More tests
- a few bugfixes and refactors
- renamed methods to be private/protected
- documentation updates

Also replaced 'action' with 'command' (the first argument)

And added a number of tests

I know it's a different ticket but it seemed to easy not to do here

Also fixed a coupld of distcheck issues with .pyc files

But essentially, this is a full rewrite of the tool, since to handle it nicely I had to get rid of most of the functions and make a somewhat nicer class of it.

Also added a -q option while i was at it

Doesn't test anything useful yet, mostly unit testing framework

Don't include example.zone test data file in dist tarball.
It is generated from an .in file via ./configure.

This was noticed by emhn during Debian packaging.
(distclean created a source different than extract tarball.)

I didn't get okay, but I mentioned my patch on jabber.
I tested on one system. This seems to be trivial.

This had a quick review via jabber.

They were missing in some files that were longer than a few lines.
Not real important, but to be consistent.

used reorder_message_file.py

Some rewording to be clear about difference of bind10 and b10-init.

Remove some trailing whitespace.

Say the resolver is experimental. Also introduce resolver.

In manpages, improve various SEE ALSO.

Fix a wrong man page reference.

This was reviewed via jabber.

This is an ad-hoc workaround for the failures in stats tests. It's okayed to be
committed on jabber. Note that this workaround should be removed after a
fundamental fix for stats tests (#1668) is done. An investigation task for the
fix will be worked on #2689.  For the details of the failures, see #2666 and
http://git.bind10.isc.org/~tester/builder//BIND10-valgrind/20130212205704-CentOS5-x86_64-GCC/logs/unittests.out

Do a hack to make the tests run with separate build and source
directories. The generated file ended up in a different directory than
the rest of the modules and the __init__.py. It therefore didn't work.
We use the same trick as log_messages, a forwarder module and addition
to PYTHON_PATH to find it.

Only the real file is installed.

The selectSubnet() function offers two ways to find a subnet for a
particular interface: using the name of the interface that the subnet is
configured for; and using the packet source address. Also, the redundant
warning is removed if suitable subnet was not found as this warning is
already issued elsewhere.

... by using the reverse of the name as the key.

% seems to (possibly) cause an unexpected side effect in an SQL statement.

sqlite3 data source currently doesn't handle escaled labels 100% correctly,
especially in identifying the "previous name" for DNSSEC negative proofs.
since this is not the point of these tests, it should be better to work around
it by the use of non escaped labels.

Before, the query tests used the old datasrc::InMemoryClient wrapped
around a ZoneFinder, from where it got its zone data (the old in-memory
design). This is no longer available.

The mock zone finder also rigs some names to do some special things. We
cannot support this with the new datasrc::memory::InMemoryClient. So we
implement a mock client in this commit.

Note that because we no longer have the old RBTree, and the new
DomainTree is not useful as it returns const nodes, we use a std::map
instead. This is slightly slower when no exact match occurs as far as
these unittests are concerned, but they are tests anyway and there
aren't many zones in the map, so we can ignore this slight inefficiency.

Explicit allocation failures cause the return of packet to the client
indicating that the address allocation has been refused.  Also added
was a "catch" for other exceptions - this causes the packet from the
client that generated the exception to be ignored.

This is not allowed by our coding style. Use the correct message ID
instead (which was previously unused).