This brings us more into line with other langauges, and makes it more
obvious when we are requiring protobuf generated code.

1. Added ruby22 and jruby tests to jenkins.
2. Added javascript tests to jenkins.
3. Added golang tests to jenkins.
4. Removed ruby19/ruby20 tests from travis. Support for ruby 2.0 has
   ended since 2016/02/24.
   https://www.ruby-lang.org/en/news/2016/02/24/support-plan-of-ruby-2-0-0-and-2-1/

Change-Id: Ie984b06772335352a4be7067ab2485f923875685

This fixes the test_acts_likes_an_array test in RepeatedFieldTest, which
checks that repeated fields respond to the same methods as regular Ruby
arrays. The bsearch_index and dig array methods seem to be new in Ruby
2.3 and so we should support those.

For JSON encoding we provide a new option to decide at
encode time whether to use camelCase or original proto field
names:

  json = MapMessage.encode_json(m, :preserve_proto_fieldnames => true)

The flags are:

  UPB_JSON_ACCEPT_LEGACY_FIELD_NAMES
  UPB_JSON_WRITE_LEGACY_FIELD_NAMES

The first just allows the parser to accept the old field names.
The second makes the printer print the old field names.

These flags are intended to be temporary, as a migration aid
for users.

The testee program of these two languages crashes on some test input.

Version number for beta languages is updated to v3.0.0-beta-2.
Version number for alpha languages is updated to v3.0.0-alpha-5.