pytest-glow-report

Features

• Automatic report generation when the installed pytest plugin is enabled
• One result per pytest test, including setup, call, and teardown failures
• A glow-report wrapper for Python's built-in unittest runner
• HTML and JSON output, plus SQLite-backed history for the last 10 runs in the same report directory
• Result search and pass/fail/skip filters in the enhanced HTML view
• Custom report title, environment metadata, and logo hooks for pytest
• Test steps, logs, and embedded PNG/JPEG screenshots
• Dark and light themes

Internet access is required for the report styles and interactive controls: Tailwind CSS and Alpine.js are loaded from public CDNs when the report is opened.

Installation

Install the package:

pip install pytest-glow-report

pytest itself is an optional dependency. To install it with the plugin:

pip install "pytest-glow-report[pytest]"

Python 3.8 or newer and pytest 6.0 or newer are supported.

pytest

The plugin is registered through pytest's pytest11 entry point and is enabled by default after installation. Run pytest normally:

pytest

This writes:

reports/
├── report.html
├── report.json
└── history.sqlite

Use another output directory or disable generation for a run:

pytest --report-dir=my_reports
pytest --no-glow-report

--report-dir and --no-glow-report are pytest options; they do not configure the unittest wrapper.

unittest

Run unittest discovery through the installed CLI:

glow-report run -- unittest discover -s tests

The wrapper uses BeautifulTestRunner, preserves unittest's process exit status, and writes to reports/. It records passes, failures, errors, skips, subtests, expected failures, and unexpected successes. Expected failures are represented as skipped results; unexpected successes are represented as failures.

Configuration

Environment variables

The pytest integration reads these variables:

POSIX shell example:

export GLOW_REPORT_TITLE="E-Commerce Regression"
export GLOW_TEST_TYPE="Regression"
export GLOW_BROWSER="Chrome 120"
export GLOW_ENVIRONMENT="Staging"
pytest

PowerShell example:

$env:GLOW_TEST_TYPE = "Smoke"
$env:GLOW_BROWSER = "Edge 120"
$env:GLOW_DEVICE = "Windows 11"
pytest

pytest customization hooks

Add a logo URL or image data URI in conftest.py:

import pytest


@pytest.hookimpl(optionalhook=True)
def pytest_html_logo():
    return "https://example.com/logo.png"

Add environment metadata:

import os

import pytest


@pytest.hookimpl(optionalhook=True)
def pytest_html_environment():
    return {
        "Browser": "Chrome 120",
        "Environment": os.environ.get("ENV", "local"),
        "Build": os.environ.get("CI_BUILD_ID", "dev"),
        "Branch": os.environ.get("GIT_BRANCH", "main"),
    }

The report always includes the Python version and platform. Hook values are merged into the Environment section.

Test-context API

The context API records data only while a test is running under the pytest plugin or BeautifulTestRunner. Calls made outside an active test still print their console message but are not attached to a report result.

Steps

Use report.step(title) with synchronous or asynchronous functions. The report records the step name, duration, and passed/failed status, then re-raises any exception.

from beautiful_report import report


@report.step("Opening login page")
def open_login(driver):
    driver.get("/login")


@report.step("Loading account")
async def load_account(client):
    return await client.get("/account")

Logs

from beautiful_report import report

report.log("User created successfully")

Screenshots

Attach an existing PNG or JPEG file:

from beautiful_report import report

report.screenshot("login page", path="screenshots/login.png")

Or capture from a synchronous Selenium- or Playwright-style object:

report.screenshot("after submit", driver=driver)

The driver must provide either get_screenshot_as_png() or a synchronous screenshot() method that returns bytes. Screenshots are embedded as data URIs in both report.html and report.json; large screenshots therefore increase both artifact sizes.

Output and history behavior

• report.html is the human-readable report.
• report.json contains the same summary, test results, environment metadata, history snapshot, and pytest session exit status when applicable.
• history.sqlite stores aggregate run counts and duration. The HTML and JSON outputs show up to the 10 most recent rows from that database.
• Reusing a report directory updates its history. Deleting or changing the directory starts a new history database.
• Each run replaces report.html and report.json in the selected directory.

Generated reports can contain test names, tracebacks, captured output, custom logs, metadata, and screenshots. Treat them as potentially sensitive CI artifacts and review them before sharing.

Development

pip install -e ".[dev]"
pytest tests/
ruff check .
mypy src/
python -m build

See CONTRIBUTING.md for contribution guidelines.

Author

Dhiraj Das · Follow @dkkkkd on X

License

MIT License. See LICENSE.