Setup CI for Tensorflow I/O

[yongtang]

This issue tracks the efforts and progresses to setup CI for TensorFlow I/O. We plan to use Google CI so coordination with TensorFlow's CI might be needed.

[jsimsa]

@martinwicke seeing as the I/O modules will likely not require GPUs for tests, do you think it would make more sense for the project to use Travis CI?

[martinwicke]

I think that would be a fine choice, provided it can be built within the time that Travis gives you.

[yongtang]

I think it makes sense if Travis CI could complete the build in time. Though additionally we may always want to have at least one CI test running on Google CI (and match the same configurations as TensorFlow repo itself), just to make sure TensorFlow I/O is compatible with TensorFlow release.

[yongtang]

I created a PR with Travis CI added: #16

The configuration just uses the default Ubuntu 14.04 with bazel 0.17.2

So far it looks ok. I think we could use it for now, though at some point I think an additional test from Google CI that matches the TensorFlow repos build environment would be great (so that tensorflow-io always matches tensorflow binary build).

[yongtang]

This issue has been addressed in PR #16.