pytest-fluent 0.4.0

A pytest plugin in order to provide logs via fluentd

pytest-fluent

pytest-fluent is a Python package in order to extend pytest using Fluentd for logging test results.

Description

pytest is one of the most powerful testing suites with a lot of functionality and many plugins. Fluentd is a well-established log collector which is used in many modern web architectures. So, why not putting both worlds together in order to gain real-time log access to your distributed test runs?

pytest-fluent enables you to forward your test data immediately to your preferred log-sink from any node spread around your infrastructure. The streamed data are available for each pytest stage and formatted in JSON.

Each pytest session gets an unique identifier (UID) assigned as well as each executed test case. With these UIDs, you can easily filter sessions and testcase data entries, for instance in your favorite database.

Installation

The package is available at pypi.org and can be installed by typing

pip install pytest-fluent

Usage

pytest-fluent-logging forwards meta data from pytest to Fluentd for further processing. The meta data are

• unique session ID
• unique test ID
• status of the session respectively test case
• test case name
• test results
• record_property entries
• custom testcase information
• custom session information
• timestamps

Furthermore, the Python logging instance can be extended in order to forward test case runtime logging:

from logging import getLogger

def test_my_runtime_log():
    value = 1
    getLogger().info("Setting value to %s", value)
    assert value == 1

or:

from logging import getLogger

def test_my_runtime_log():
    value = 1
    getLogger('fluent').info("Setting value to %s", value)
    assert value == 1

Fixtures

In order to create your own logger, request the fixture get_logger as follows:

def test_my_runtime_log(get_logger):
    logger = get_logger('my.Logger')
    value = 1
    logger.info("Setting value to %s", value)
    assert value == 1

If you want to get the current UIDs, use the session_uid and test_uid fixtures as follows:

def test_unique_identifier(get_logger, session_uid, test_uid):
    logger = get_logger('fluent')
    logger.info("Session ID: %s", session_uid)
    logger.info("Test ID: %s", test_uid)
    value = 1
    assert value == 1

Callbacks

If you want to add custom data to the datasets of the pytest_sessionstart and pytest_runtest_logstart stages, decorate your callback functions with the following decorators:

from pytest_fluent import (
    additional_session_information_callback,
    additional_test_information_callback
)

@additional_session_information_callback
def provide_more_session_information() -> dict:
    return {
        "more": "session information"
    }

@additional_test_information_callback
def provide_more_test_information() -> dict:
    return {
        "more": "test information"
    }

pytest CLI extensions

The pytest CLI application can be called with the following arguments in order to configure fluent-logging.

argument	description	default
--session-uuid	Use a custom externally created UUID, e.g. link a CI job with the pytest session.
--fluentd-host	Fluentd host address. If not provided, a local Fluentd instance will be called.
--fluentd-port	Fluent host port	24224
--fluentd-tag	Set a custom Fluentd tag	'test'
--fluentd-label	Set a custom Fluentd label	'pytest'
--fluentd-timestamp	Specify a Fluentd timestamp	None
--extend-logging	Extend the Python logging with a Fluent handler	False
--add-docstrings	Add test docstrings to testcase call messages
--stage-settings	Use custom stage settings file. See documentation

Ini Configuration Support

Default values of the CLI arguments for a project could also be defined in one of the following ini configuration files:

1. pytest.ini: Arguments are defined under the pytest section in the file. This file takes precedence over all other configuration files even if it's empty.

[pytest]
addopts= --session-uuid="ac2f7600-a079-46cf-a7e0-6408b166364c" --fluentd-port=24224  --fluentd-host=localhost --fluentd-tag='dummytest' --fluentd-label='pytest' --extend-logging

2. pyproject.toml: are considered for configuration when they contain a tool.pytest.ini_options section is available

[tool.pytest.ini_options]
addopts="--fluentd-port=24224 --fluentd-host=localhost --fluentd-tag='test' --fluentd-label='pytest' --extend-logging"

3. tox.ini: can also be used to hold pytest configuration if they have a [pytest] section.

[pytest]
addopts= --fluentd-port=24224 --fluentd-host=localhost --fluentd-tag='test' --fluentd-label='pytest'

If the same option is specified in both CLI and ini file, then CLI option would have higher priority and override the ini file values.

What data are sent?

pytest-fluent sends any information, e.g. stage information or logging from a test case, as a single chunk. For instance, the data collection from test_addoptions.py test looks as following

[
    {
        "status": "start",
        "stage": "session",
        "sessionId": "d8f01de3-8416-4801-9406-0ea3d5cfe3c0"
    },
    {
        "status": "start",
        "stage": "testcase",
        "sessionId": "d8f01de3-8416-4801-9406-0ea3d5cfe3c0",
        "testId": "6b444275-4450-4eff-b5d9-8355f0f99ab0",
        "name": "test_fluentd_logged_parameters.py::test_base"
    },
    {
        "type": "logging",
        "host": "myComputer",
        "where": "test_fluentd_logged_parameters.test_base",
        "level": "INFO",
        "stack_trace": "None",
        "message": "Logged from test_base",
        "sessionId": "d8f01de3-8416-4801-9406-0ea3d5cfe3c0",
        "testId": "6b444275-4450-4eff-b5d9-8355f0f99ab0",
        "stage": "testcase"
    },
    {
        "type": "logging",
        "host": "myComputer",
        "where": "test_fluentd_logged_parameters.test_base",
        "level": "INFO",
        "stack_trace": "None",
        "message": "Logged from test_base",
        "sessionId": "d8f01de3-8416-4801-9406-0ea3d5cfe3c0",
        "testId": "6b444275-4450-4eff-b5d9-8355f0f99ab0",
        "stage": "testcase"
    },
    {
        "name": "test_fluentd_logged_parameters.py::test_base",
        "outcome": "passed",
        "duration": 0.0013457999999999526,
        "markers": {
            "test_base": 1,
            "test_fluentd_logged_parameters.py": 1,
            "test_fluentd_logged_parameters0": 1
        },
        "stage": "testcase",
        "when": "call",
        "sessionId": "d8f01de3-8416-4801-9406-0ea3d5cfe3c0",
        "testId": "6b444275-4450-4eff-b5d9-8355f0f99ab0"
    },
    {
        "status": "finish",
        "stage": "testcase",
        "sessionId": "d8f01de3-8416-4801-9406-0ea3d5cfe3c0",
        "testId": "6b444275-4450-4eff-b5d9-8355f0f99ab0",
        "name": "test_fluentd_logged_parameters.py::test_base"
    },
    {
        "status": "finish",
        "duration": 0.00297708511352539,
        "stage": "session",
        "sessionId": "d8f01de3-8416-4801-9406-0ea3d5cfe3c0"
    }
]

whereat each object in the array is sent independently via Fluentd.

Specifying a timestamp

Timestamps are added to the information if the --fluentd-timestamp option is enabled:

[pytest]
addopts= --session-uuid="ac2f7600-a079-46cf-a7e0-6408b166364c" --fluentd-port=24224  --fluentd-host=localhost --fluentd-tag='dummytest' --fluentd-label='pytest' --fluentd-timestamp='@timestamp' --extend-logging

The timestamp is added to each message. The value is in ISO 8601 format. A sample of the data collection from test_addoptions.py (as above) would look as below:

[
    {
        "status": "start",
        "stage": "session",
        "sessionId": "d8f01de3-8416-4801-9406-0ea3d5cfe3c0",
        "@timestamp": "2022-12-25T03:00:00.000000Z"
    },
    {
        "status": "start",
        "stage": "testcase",
        "sessionId": "d8f01de3-8416-4801-9406-0ea3d5cfe3c0",
        "testId": "6b444275-4450-4eff-b5d9-8355f0f99ab0",
        "name": "test_fluentd_logged_parameters.py::test_base",
        "@timestamp": "2022-12-25T03:00:00.100000Z"
    }
]

Changelog

The changelog.

Contributing

We welcome any contributions, enhancements, and bug-fixes. Open an issue on Github and submit a pull request.