Allure pytest integration

2.15.3 · active · verified Thu Apr 09

Allure pytest integration provides a way to integrate the Allure Test Reporting Framework with pytest tests. It functions as a pytest plugin, automatically collecting test results and generating comprehensive, interactive reports. The library is actively maintained as part of the `allure-python` monorepo, with frequent minor releases, typically on a monthly or bi-monthly basis.

Warnings

breaking Older versions of `allure-pytest` (pre-2.14.3) may not correctly handle `allure.title` for fixtures when used with `pytest` version 8.4 or later.
Fix: Upgrade to `allure-pytest` version 2.14.3 or higher to ensure fixture titles are correctly displayed in reports.
gotcha If you were previously using `pytest-allure-adaptor`, it must be uninstalled before installing `allure-pytest` due to conflicting implementations. The `allure-pytest` setup process will explicitly check for this conflict.
Fix: Run `pip uninstall pytest-allure-adaptor` before `pip install allure-pytest`.
breaking As of `allure-python` 2.14.0, official support for Python 3.7 has been dropped. Users on Python 3.7 will not receive updates or fixes for newer `allure-pytest` versions.
Fix: Upgrade your Python environment to 3.8 or newer.
gotcha The Allure Report command-line tool (required to generate and serve HTML reports) is a separate installation from `allure-pytest` and requires a Java Runtime Environment (JRE) to function.
Fix: Download and install the Allure command-line tool (e.g., via Homebrew on macOS, or manually) and ensure Java is installed and accessible in your PATH.
gotcha In versions prior to 2.15.0, reports for tests defined within nested classes might have incorrect `fullName`, `historyId`, `testCaseId`, and `subSuite` values, leading to improper grouping and history tracking.
Fix: Upgrade to `allure-pytest` version 2.15.0 or higher for accurate reporting of nested test classes.
gotcha As of version 2.14.3, built-in (reserved) `pytest` markers are no longer converted into Allure tags. If your reporting workflow relied on these markers appearing as tags, this behavior has changed.
Fix: Use explicit `allure.tag()` or `@allure.tag()` decorators for custom tags, or consider custom hooks if specific `pytest` markers still need to be mapped to Allure tags.

Install

pip install allure-pytest Install allure-pytest

Imports

allure
```
import allure
```
Primary import for Allure decorators and dynamic API functions.
LabelType
```
from allure_commons.types import LabelType
```
Used for defining standard Allure labels programmatically.

Quickstart

This quickstart demonstrates a basic pytest test file with Allure decorators. To generate the Allure report, first run pytest with the `--alluredir` option to create XML results. Then, use the `allure serve` command (from the separately installed Allure command-line tool) to process these XML files and open an interactive HTML report in your browser. The Allure command-line tool requires Java.

# test_example.py
import allure
import pytest

@allure.epic("Web Testing")
@allure.feature("Login Functionality")
def test_successful_login():
    allure.dynamic.story("User can log in with valid credentials")
    with allure.step("Enter username and password"):
        print("Entering credentials...")
    with allure.step("Click login button"):
        print("Clicking login...")
    assert True

@allure.epic("API Testing")
@allure.feature("User Management")
def test_get_user_profile():
    allure.dynamic.story("Retrieve user data via API")
    with allure.step("Send GET request to user endpoint"):
        print("Sending request...")
    with allure.step("Verify response status code"):
        print("Checking status...")
    with allure.step("Validate user data in response"):
        print("Validating data...")
    assert True

# To run tests and generate Allure results:
# pytest --alluredir=allure-results

# To serve the Allure report (requires Allure command-line tool):
# allure serve allure-results

view raw JSON →