PyTest Notes — Akash

Naming Conventions

PyTest auto-discovers tests only when they follow these exact naming rules.

Entity	Convention
File / Module	Case insensitive — must match `test_.py` or `_test.py`
Class Name	Must start with exactly `Test`
Method Name	Must start with exactly `test`

💡

PyTest scans directories recursively for files matching the pattern. No registration needed — just follow the naming rules.

Execution Methods

Run entire directories, single modules, or filter by markers. Flags control verbosity and parallelism.

pytest -s -v .\API-Automation\Execute all modules in the directory
pytest -s -v .\API-Automation\pytest_test.pyExecute one specific module

Flag	Purpose
-s	Show `print()` statements from test methods
-v	Provide detailed / verbose output
-rs	Show details for skipped test cases
-p no:warnings	Suppress all warnings
-n=3	Run 3 test cases in parallel (requires pytest-xdist)

Fixture Functions

Fixtures provide setup and teardown logic. The yield keyword splits setup (before) from teardown (after).

🐍 Python

import pytest

@pytest.fixture(autouse=True, scope="module")
def fixture_function():
    print()
    print("Setup Method")     # runs before tests
    yield                       # ← setup | teardown boundary
    print()
    print("Teardown Method")  # runs after tests

📌

yield boundaryAll statements before yield act as setup and run at the start of the defined scope. All statements after yield act as teardown and run at the end of the scope.

💡

@pytest.fixture(autouse=True) — when used, the fixture will be applied automatically to every test without manually passing it as an argument.

Fixture Scope

Controls how often a fixture is set up and torn down. Use broader scopes to avoid repeated expensive operations.

Scope	Lifecycle
function	Default — runs before and after every test case
class	Runs once before the class; tears down after all class tests complete
module	Runs once before the module; tears down after all module tests complete
package	Runs once per package directory
session	Runs once per entire test session

🐍 Python

@pytest.fixture(scope="class")
def db_connection():
    conn = create_connection()
    yield conn
    conn.close()

conftest.py

A special file that holds shared fixtures available across multiple modules.

conftest.py

Fixtures defined here are available to all functions, classes, and modules in the same directory and below.

Module-level fixture

Fixtures defined inside a regular module (test_*.py) can only be used within that specific file.

📂

Place conftest.py at the root of your test directory for project-wide fixtures, or in sub-directories for scoped sharing.

Markers

Markers control test behavior: skipping, grouping, ordering, dependencies, and parametrization.

💡

Custom markers must be registered in pytest.ini to suppress unknown-marker warnings. Add them under [pytest] markers =.

pytest.ini

[pytest]
markers =
    sanity
    regression
    zem
    ask

Skipping & XFail

Three ways to skip tests, plus a marker for expected failures.

Decorator / Call	Behavior
@pytest.mark.skip(reason=…)	Unconditionally skip — always skipped regardless of environment
@pytest.mark.skipif(condition, reason=…)	Skip only when the given condition evaluates to `True`
pytest.skip()	Skip programmatically at runtime inside a test or fixture
@pytest.mark.xfail(reason=…)	Mark a test as expected to fail — highlighted specially in reports

🐍 Python — Skip example

import pytest
import sys

@pytest.fixture
def usersession():
    session_active = False
    if not session_active:
        pytest.skip("User Session is not active")

class Test:
    @pytest.mark.skipif(sys.platform == "win32", reason="Not applicable for Windows")
    def test_Method1(self):
        print("Method One")

    def test_Method2(self, usersession):
        print("Method Two")

⚠️

Use -rs flag when running pytest to see the skip reasons in the output summary.

Ordering

Control test execution order using the pytest-ordering package. Register custom markers in pytest.ini.

📦

Install with: pip install pytest-ordering

🐍 Python

@pytest.mark.run(order=1)
def test_first():
    ...

@pytest.mark.first
def test_also_first():
    ...

📌

Dependency — use pytest-dependency to make a test depend on another passing first.
@pytest.mark.dependency(depends=["Test::test_launch"]) — format is ClassName::TestMethod.

Grouping with Markers

Tag tests with custom markers and run only specific groups using -m.

🐍 Python

@pytest.mark.sanity
def test_login():
    ...

@pytest.mark.regression
def test_checkout():
    ...

pytest -m "sanity and regression" .\API-Automation\Run tests that have both tags
pytest -m "sanity or regression" .\API-Automation\Run tests that have either tag
pytest -s -v -m "not sanity" .\API-Automation\Run all tests except sanity-tagged ones

Parallel Execution

Run multiple tests simultaneously using pytest-xdist to significantly cut down total suite time.

📦

Install with: pip install pytest-xdist

pytest -s -v -n=2 .\API-Automation\Run 2 test cases in parallel
pytest -s -v -n=auto .\API-Automation\Auto-detect CPU count and use all cores

⚠️

Tests running in parallel must be independent — no shared state or order-dependent logic. Fixtures with scope="session" may behave differently under -n.

Parametrization

Used for data-driven testing when input quantity is small. Each parameter set becomes a separate test run.

🐍 Python

import pytest

class Test:
    @pytest.mark.parametrize('num1, num2, sum', [(1, 1, 2), (2, 3, 5)])
    def test_addition(self, num1, num2, sum):
        assert num1 + num2 == sum

    def test_sum(self, num1, num2, sum):
        assert num1 + num2 == sum

📌

For larger datasets, consider using CSV/JSON files with a fixture that reads and yields data, keeping the test method clean.

Hooks

Hooks let you capture command-line arguments and inject environment metadata into HTML reports.

🐍 Python — Command-line hook

import pytest

def pytest_addoption(parser):
    parser.addoption(
        '--browser',
        action='store',
        default="edge",
        help="Browser to run test"
    )

@pytest.fixture
def browser(request):
    return request.config.getoption("--browser")

pytest -s -v .\API-Automation\ --browser=chromePass browser value from command line

🐍 Python — HTML Report metadata hook

import pytest
from pytest_metadata.plugin import metadata_key

def pytest_addoption(parser):
    parser.addoption('--browser', action='store', default="edge", help="Browser to run test")

@pytest.fixture
def browser(request):
    return request.config.getoption("--browser")

@pytest.mark.optionalhooks
def pytest_metadata(metadata):
    metadata.pop("Python", None)

@pytest.mark.optionalhooks
def pytest_configure(config):
    config.stash[metadata_key]["Project Name"] = "Example"

📌

pytest_metadata and pytest_configure hooks allow you to customize the environment section of HTML reports — great for CI/CD pipelines where you need traceability.

PyTest Quick Reference

Naming Conventions

Execution Methods

Fixture Functions

Fixture Scope

conftest.py

conftest.py

Module-level fixture

Markers

Skipping & XFail

Ordering

Grouping with Markers

Parallel Execution

Parametrization

Hooks