Add Ruby to Travis testing

- Added RVM-based Ruby test driver that tests MRI and JRuby.
- Fixed JRuby compilation (at least in my current setup): force source
  version to 1.6 (Java 6) to allow generics and annotations.
- Modify the skipped JRuby JSON tests so that the exit code is 0 (skip()
  results in a failing exit code from `rake test`). An upcoming PR
  should fix JSON under JRuby in general soon.

Dockerfile for building protoc on Centos 6.6

Uncomment conflict fields from pkg-config files.

The referenced bug was fixed in 2007 and has been released in pkg-config-0.23
(16.Jan.2008). The fixed version is widely available.

adding globals() to exec call for python3 compatibility

Handle 0-length byte buffers in micro and nano protobufs

Fix java compilation issues when processing large .proto files

Run conformance tests against Java for the Travis build.

Change-Id: I8ef8664f7facf86028be3f4f0d5d2efc8a685d6d

Objective C compiler plugin and runtime

Cleanup setup.py

- Comply with flake8, except for indentation width, which is 2.

- Move human-centric metadata to top of 'setup()' call.

- Add Trove classifiers for supported Python versions.

- Use 'find_packages()' + MANIFEST.in to avoid errors in listing
  modules and packages.

Closes #131 and #161.

Adds conformance test support for Java.

Change-Id: I845ee1ab1005d25c8d77a8c2ed801c0f7b7c847b

ruby: allow a message field to be unset

ruby: return nil for out of bounds

ruby arrays don't throw an exception; they return nil.  Lets do the
same!

this fix also includes the ability to use negative array indicies

ruby: Add size to RepeatField

starting to make `RepeatedField` quack like an array

additional changes:
* make sure gemspec gets all ruby code files
* add homepage in gem spec removes one of the warnings, and the gem spec authors are pushing
everyone to include a homepage in the gem
* remove excess whitespace in test suite to bring formatting inline with the rest of the file

ruby: build cleanups

* update docs to simplify build steps
* Gemfile.lock seemed to have an older version specified
* do not check in the pkg dir

Generate pbconfig.h in generate_descriptor_proto.sh

Sync nanoproto from AOSP

Include well-known type protos in "make install".

This is less ideal from a dex count perspective because it requires a
new variable for each message, and because most apps have proguard
rules that will ensure that CREATOR classes are retained.

However, it is required to be able to use nano protos inside of AIDL
files, as the autogenerated AIDL code fails to compile otherwise. This
is a substantial benefit as it allows for backwards-compatible
parameters and return types in AIDL methods along the lines of
safeparcel.

Bug: 19084705
Change-Id: I66a2c0424b96cf8ff6b631b186cc4f9407dfc1f4

When a MessageNano containing a String is serialized into a buffer that
is too small to contain it, and the buffer's boundary happens to be
where the string field's length delimiting varint is serialized,
and the string's length and 3*length have the same length when
encoded as a varint, an IllegalArgumentException is thrown rather than
an OutOfSpaceException.

Github issue: https://github.com/google/protobuf/issues/292

Change-Id: If478d68cf15bfd0662252d008e42b2bf1ff1c75e

It turns out dex (apparently) was inlining these protected final
methods from ExtendableMessageNano into every message class. Removing
these methods from the base class and inlining their code reduces
the method count by 2 methods / message when the store_unknown_fields
option is on.

Change-Id: I0aa09f2016d39939c4c8b8219601793b8fab301f

I wasn't able to get the clear() method to inline into the
constructor when optimizations are on in proguard. As a result,
every message has an extra superfluous kept method assuming the
app never uses clear() directly.

There are a couple of instances where setting this option false is
necessary in order to get code dexing successfully without hitting
the method limit, e.g. https://goto.google.com/tltzq

In this example, I tried turning on the method/inlining/unique and
method/inlining/short optimizations before resorting to adding the
generate_clear option, but the method count did not decrease. The
clear() methods were contributing over a thousand extra methods.

Change-Id: If6a9651d6a59cdf70b1040d8248779710ac73105

@IntDef is a support library annotation which allows build tools to
determine the valid set of values for a given integer field when that
field is intended to be restricted like an enum. This avoids the
overhead of enums while still allowing for compile-time type checking
in most circumstances.

Change-Id: Iee02e0b49a8e069f6456572f538e0a0d301fdfd5

Change-Id: Ie2a9e36276ac35e10b3f8d379b5742d50a0374e9