* A few more cases for binary conformance tests.

* over-encoded varints (encoded in more bytes than are necessary).
* truncated varints (>32 bits for 32-bit types).

* Fixed Python decoding bug with 32-bit varints.

* Fixed 1L -> 1LL for 32-bit platforms.

This is basic and more tests will be added over time.

* Factored Conformance test messages into shared test schema.

* Updated benchmarks to use new proto3 message locations.

* Fixed include path.

* Conformance: fixed include of Python test messages.

* Make maven in Rakefile use --batch-mode.

* Revert changes to benchmarks.

On second thought I think a separate schema for
CPU benchmarking makes sense.

* Try regenerating C# protos for new test protos.

* Removed benchmark messages from test proto.

* Added Jon Skeet's fixes for C#.

* Removed duplicate/old test messages C# file.

* C# fixes for test schema move.

* Fixed C# to use the correct TestAllTypes message.

* Fixes for Objective C test schema move.

* Added missing EXTRA_DIST file.

Add note about JSON tests maybe being wrong.

- Add note about the JSON test maybe not being correct yet.
- Add test to checks the generated names for double underscores to be sure they
  are what is expected.

- String behaviors
- Testing only valid JSON is allowed

Add more JSON tests around underscores

Incase the different runtimes have different generation/implementations for
for the different primitive field types, cover a larger range of the proto
types in the oneof zero tests.

* Add conformance test for zero fields in oneofs.
* Add failures to the "expected" files.

can reasonably be expected to be interoperable.

(The fact that not all Int64/Uint64 values *can* be exactly represented in
IEEE-754 and thus interoperably amongst JSON implementations is precisely the
reason for wrapping them when we serialize... it shouldn't be a surprise that
we can't round-trip in unwrapped form for all values.)

The failure lists were falling out of date because the
tests would pass even if people forgot to remove failures
from the list.

- Update all the new runner cases to handle a skip result as a skip and not a failure.
- Make ObjC use the new serialize_error result if it fails to generate the data.

commit e841bac4 seems to have mis-merged the
change to reporting skipped tests.

I didn't check if there were other merge issues.

Change-Id: Ibb3fe6f919cc7ca0df91da5e1697ba33d259e433

This was enabled by the recent open-sourcing of JSON
support and MessageDifferencer.

MessageDifferencer allows the conformance suite to expand
because it allows us to write tests for payloads that parse
successfully.  To verify the testee's output payload, we
need to parse it back into a message and compare the message
instances.  Comparing output bytes vs. a golden message is
*not* valid, because protobufs do not have a canonical
encoding (especially in the presence of maps, which have
no prescribed serialization order).

We only add one small JSON test for now, but with the
framework in place we now have the foundation to dramatically
expand the coverage of the conformance test suite.

Also added the ability for the testee to skip tests that
exercise features that are unimplemented.  This allows
Java (which currently has no JSON support) to skip tests
involving JSON.

Change-Id: I697b4363da432b61ae3b638b4287c4cda1af4deb

Change-Id: I382dcda97fa123a6da4ff5faad5d7ece95853f33

This allows us to enable conformance tests even when we know
that some tests are failing and need to be fixed.

Change-Id: I372f43663008747db6f2b2cf06e6ffa4c6d85b2d

This is necessary for testing on iOS.

Change-Id: I54ec1e3aa2e9fbfff9a6cd6580920a6a62069b63

Change-Id: Ib75664194491643f8e4f1503a2ed942a2d1e1655