python-ongtrum 0.1.4

Ongtrum

Ongtrum is a Python test runner designed for speed

Installation: pip install python-ongtrum

Why Ongtrum is Fast ?

Scan

Recursively scans for test files in the project directory with a Cython-optimized filesystem scanner

Parse

Each file is parsed with a Cython-optimized AST parser

Execute

For each file, all its test classes and methods are executed in one go using Python’s exec
To reduce overhead, files are processed in batches, keeping worker processes alive throughout the run

Optional Parallelism

Tests can be run across multiple processes with --workers

Benchmark

Execution Commands:

• Unittests: python -m unittest discover <Project Dir>
  
• PyTest: python -m pytest <Project Dir> -s -q
  
• Ongtrum: python -m ongtrum.py' -q -p <Project Dir>
  

Test Files: 5440
Test Class: 10880

• Unittests: 12.462s
  
• PyTest: 27.217s
  
• Ongtrum: 5.950s (vs Unittests 2 × Faster, vs PyTest 4.5 × Faster)
  

Features

Test Suites

Organize tests into named suites using the @suites decorator

from ongtrum.annotation import suites


class TestSuite:
    @suites(['suite_one', 'suite_two'])
    def test_dummy_1(self):
        assert 1 == 1

Test Parameters

Ongtrum allows you to define multiple sets of input values for a test method using the @parameters decorator
Each dictionary represents a different test case

from ongtrum.annotation import parameters


class TestParameter:
    @parameters([{'a': 1, 'b': 2}, {'a': 10, 'b': 20}])
    def test_add(self, a, b):
        assert a in [1, 10]
        assert b in [2, 20]

Test Preps

Ongtrum allows you to define reusable preparation functions (“preps”) in dedicated prep files,
which are automatically loaded for your tests <>

Scopes:

• Session - Run once per test session - Useful for expensive setup shared by all tests - Cached for the session
  
• class - Run once per test class - Shared among all methods in that class - Cached for the class
  
• method - Run for each test method individually - Re-run for every method call
  

Configuration File

By default, Ongtrum looks for an ongtrum.yaml file:

• At the project root when running a directory
  
• In the parent directory when running a single test file
  

Alternatively, you can specify a config file explicitly using ongtrum -c path/to/ongtrum.yaml

Configuration Example

prep_files:
  - preps/session_preps.py
  - preps/class_preps.py
  - preps/method_preps.py

Declaring a Prep

from ongtrum.annotation import prep


@prep(scope='session')
def session_prep():
    return 'Session Prep'


@prep(scope='class')
def class_prep():
    return 'Class Prep'


@prep(scope='method')
def method_prep():
    return 'Method Prep'

Applying a Prep

from ongtrum.annotation import preps


@preps('class_prep')
class TestExample:

    @preps('method_prep')
    def test_something(self, method_prep):
        assert self.class_prep == 'Class Prep'
        assert method_prep == 'Method Prep'

Test Filter

You can run a subset of tests using the -f / --filter option
Filter format: <File>.<Class>.<Method>
You can also use wildcards (*) for flexible matching

Examples:

ongtrum -p tests/test_example.py -f test_example
ongtrum -p tests/test_example.py -f test_example.TestClass
ongtrum -p tests/test_example.py -f test_example.TestClass.test_method

ongtrum -p tests/test_example.py -f *.TestClass
ongtrum -p tests/test_example.py -f *.*.test_method