Update Python SDK so FeatureSet can import Schema from Tensorflow metadata

[davidheryanto]

What this PR does / why we need it:
This PR extends FeatureSet, Entity and Feature class in Python SDK so they can contain constraints defined in the Schema from Tensorflow metadata:
https://github.com/tensorflow/metadata/blob/ddf582f66eeeddb862de6d53c3e03d6eed1c04a6/tensorflow_metadata/proto/v0/schema.proto

These contraints presence_constraints, shape_type and domain_info can be used for validation of feature values and presence in Feast. Typical usage when one is already using Tensorflow data validation is as follows:

import tensorflow_data_validation as tfdv

# Use tensorflow_data_validation to generate initial schema for validation
train_stats = tfdv.generate_statistics_from_csv(data_location="/data/train.csv")
schema = tfdv.infer_schema(statistics=train_stats)

# Create a new FeatureSet or retrieve an existing FeatureSet in Feast
feature_set = FeatureSet(
            name="myfeatureset",
            entities=[Entity(name="id", dtype=ValueType.INT64),],
            features=[
                Feature(name="feature1", dtype=ValueType.STRING),
                Feature(name="feature2", dtype=ValueType.INT), ...
            ],
        )
# Update the entities and features with constraints defined in the schema
feature_set.update_schema(schema)

Which issue(s) this PR fixes:

Related to #172

Does this PR introduce a user-facing change?:

NONE

[feast-ci-bot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: davidheryanto

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [davidheryanto]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[woop]

Other than the above, it looks good as a first cut.

[davidheryanto]

/hold
Need to resolve comments and add method to export schema from feature set.

[davidheryanto]

/hold cancel

[woop]

@davidheryanto this PR contains the same code as #449

[davidheryanto]

Yes because the end to end tests were written in Python and it depends on the Python SDK,
so for the end to end tests I have to include changes in the Python SDK too.

[zhilingc]

/lgtm