pytest-xfiles 0.2.0

File formats supported

The sample above shows use of YAML formatted data. Out of the box, JSON format is also supported via function_json, module_json and package_json fixtures.

More formats can be supported (see below).

The function, the module and the package

The file tests/shakespeare/__init__.py defines a test package.

The file tests/shakespeare/test_romeoyjulieta is a test module

The def test_romeo as in:

def test_romeo(function_yaml):
    print(function_yaml)

defines a test function.

Currently there is no notion of test class as I did not need it. It may appear later on.

Names of data files

Data file names are derived from related object (package, module, function) and use format specific extensions (.json, .yaml, special ._x_, other can be added).

Package data file has name __init__.py with extension changed to format specific one, e.g. __init__.json.

Module data file has name such as test_romeoyjulieta.py with extension changed to format specific one, e.g. test_romeoyjulieta.json.

In case of test function, test function name must be added. To make files more readable, the test_ part of the function is removed. For test_romeo function can be e.g. test_romeoyjulieta.romeo.json.

The files are typically in the same directory as relevant python test suite code.

What are the ._x_ files?

There are special fixtures function_xfile, module_xfile and package_xfile, which only return path to a file with extension ._x_ (and do not attempt to load the content).

The ._x_ files are used as base for implementing other fixtures, e.g. as for JSON:

@pytest.fixture(scope="function")
def function_json(function_xfile):
    path = function_xfile.with_suffix(".json")
    with path.open(encoding="utf-8") as f:
        import json
        return json.load(f)

As shown, the function_json simply takes the ._x_ file path, replaces the extension with it’s own .json and returns data loaded from such file.

Adding support for other data formats (e.g. CSV)

Following the function_json example above, we may load data from any other data file, e.g. for .csv:

from csv import reader

import pytest


@pytest.fixture(scope="function")
def function_csv(function_xfile):
    path = function_xfile.with_suffix(".csv")
    with path.open(encoding="utf-8") as f:
        return list(reader(f))


def test_codes(function_csv):
    print(function_csv)

Creating fixtures based on provided data

It is easy to take any of data availalbe and use it to create object of your preference. E.g. assuming that the package_yaml returns information about author in form of dictionary with keys “name” and “surname”, one can create fixture classy_author returning specific class instance. Put following into conftest.py:

class Author(object):
    def __init__(self, name, surname):
        self.name = name
        self.surname = surname

    @property
    def full_name(self):
        return "{self.name} {self.surname}".format(self=self)


@pytest.fixture(scope="module")
def classy_author(package_yaml):
    return Author(package_yaml["name"], package_yaml["surname"])

and use it from test test_classy_author.py:

def test_custom_fixture(classy_author):
    print(classy_author.full_name)

Fixtures provided

Here is summary of fixtures provided. In all cases we assume we have tests/sub/test_thing.py with a test function test_fun and all required data files are available.

Note on YAML package and YAML version

Switched from pyyaml to ruamel.yaml as it supports YAML version 1.2.

If you require YAML files using version 1.1, use % YAML 1.1 in your YAML file.