STYLE_GUIDE.md

Coding guidelines and standards

Coding Style

C++

C++ code should conform to Google C++ Style Guide.

Python

Python code should conform to PEP8.

Running Python and Bazel Style Checks

Style checks for Python and Bazel can be run with the following commands (docker has to be available):

$ bash -x -e .travis/lint.sh

In case there are any Bazel style errors, the following command could be invoked to fix and Bazel style issues:

$ docker run -i -t --rm -v $PWD:/v -w /v --net=host golang:1.12 bash -x -e -c 'go get github.com/bazelbuild/buildtools/buildifier && buildifier $(find . -type f \( -name WORKSPACE -or -name BUILD -or -name *.BUILD \))'

After the command is run, any Bazel files with style issues will have been modified and corrected.

Naming Conventions

Any new operation that is introduced in Tensorflow I/O project should be prefixed with a pre-defined text. This is done to ensure that different Tensorflow projects (for e.g Tensorflow I/O) declare their operation names that are globally unique, across the entire Tensorflow ecosystem.

For more details on defining custom operations, please refer this tensorflow RFC.

Depending on the module's programming language (C++ or python), pre-defined prefix text may vary. This is detailed below.

C++

Any new operation that is introduced in C++ module should be prefixed with text “IO>”.

for example:

A new operation named “ReadAvro” should be registered as “IO>ReadAvro”.

Regular way:

REGISTER_OP("ReadAvro")
                    ...
                    ...
                    ...
     });

Suggested way:

REGISTER_OP("IO>ReadAvro")
                    ...
                    ...
                    ...
     });

Please note in suggested way, how the operation name is prefixed with "IO>".

Python

Any new operation that is introduced in python module and that which corresponds to the operation in C++ module, should be prefixed with "io_". This ensures that C++ operation binding is mapped correctly.

for example:

The python equivalent declaration of above operation named “IO>ReadAvro” would be "io_read_avro".

def read_avro(filename, schema, column, **kwargs):
  """read_avro"""
  ...
  ...
  ...
  return core_ops.io_read_avro(...);