If you have cloned this project already then you can skip this, otherwise you'll need to clone this repo using Git. If you do not know how to clone a GitHub repository, check out this help page from GitHub.
Follow the instructions to install Docker.
Follow the instructions to install Docker Compose.
If you have a fresh cloned repository ignore this step.
If you have run this repo without docker before please run to delete e.g. prebuilt .pyc
files:
$ make clean
$ docker-compose up -d selenium-chrome
It's necessary for reducing the amount of time that it takes for test collection to be specific about the set of tests you'd like to execute.
You do this by adding the test folder you want pytest to scan for tests. Typically, the folders follow the following format `/tests//. The following test folders exist for running tests with pytest:
- ./tests/webview/integration
- ./tests/webview/ui
- ./tests/webview/vendorbooks
- ./tests/archive/integration
- ./tests/legacy/ui
- ./tests/cops
- ./tests/neb/ui
An example on how to target the UI test for webview is below:
$ docker-compose exec selenium-chrome pytest tests/webview/ui
Note: The Run the tests using pytest section covers how to pass arguments to pytest in order to target specific tests
List the docker containers and find the one for selenium-chrome
$ docker container ls
A table will be displayed with column names. Find the one labeled PORTS
PORTS
4444/tcp, 0.0.0.0:32778->5900/tcp
Use a VNC application to connect to 0.0.0.0:32778
. The port number 32778
may be different.
The password for the VNC session is secret
.
Execute the tests as described above.
$ docker-compose exec selenium-chrome pytest
Switch over to the VNC window to see your tests running!
$ brew cask install chromedriver # macOS
$ make venv
$ source .venv/bin/activate
This utilizes pre-commit to format code using black and lint your code using flake8. This is IDE agnostic and runs only on checked in code before a commit.
$ make precommit
Dotenv is used by the framework to load environmental variables from a .env
file if it exists in the root of the project directory.
This is useful for loading environment variables that use usernames. To use a .env file copy the example and fill out the values.
$ cp .env.example .env
pip-tools are utilized for updating dependencies in the framework. Pip-tools provides a number of utilities to pip that are useful for cnx-automation. Cnx-automation installs dependencies from a number of our projects that may have conflicting versions. Pip-tools will warn when there is a conflict and can help with finding compatible versions between all the dependencies and sub-dependencies.
First, install pip-tools:
$ pip install pip-tools
To do updates to dependencies you'll first need to update the main dependency in ./requirements.in file.
When the dependency has been updated you can then generate the requirements.txt that will be used to install the dependencies.
$ pip-compile --output-file=requirements.txt
If you intend to run the legacy tests, you will need to set the LEGACY_USERNAME
and LEGACY_PASSWORD environment variables. You can either export them from your
shell profile or simply add them to a .env
file in the root dir of this repo.
When only using the pytest
command the default environment will set to QA.
To run against a different environment pass in a value for --webview_base_url
, --legacy_base_url
, --archive_base_url
:
Staging
$ pytest --webview_base_url https://staging.cnx.org --legacy_base_url https://legacy-staging.cnx.org --archive_base_url https://archive-staging.cnx.org tests/webview/ui
Production
$ pytest --webview_base_url https://cnx.org --legacy_base_url https://legacy.cnx.org --archive_base_url https://archive.cnx.org tests/webview/ui
To run a specific test or test module pass in a value for -k
:
$ pytest -k test_about tests/webview/ui
To run a specific project, pass in webview
, legacy
, or neb
for -m
:
$ pytest -m webview tests/webview
To run a more complicated example that runs a specific project and a specific test module in headless mode:
$ pytest -m webview -k test_home --headless tests/webview/ui
To run tests in parallel you can combine the above and use -n
option to specify the number of workers.
$ pytest -n 4 -m webview tests/webview/ui
To run smoke tests you can add the smoke
marker to the particular project that is being tested:
$ pytest -m "webview and smoke"
To run integration tests you'll want to add teh integration
marker and target the correct folder you want to scan for test:
$ pytest -m "webview and integration" tests/webview
The pytest plugin that we use for running tests has a number of advanced
command line options available. To see the options available, run
pytest --help
. The full documentation for the plugin can be found
here.
The TestRail integration is currently intended to be used during a local test run of the cnx-automation suite when the uploading of results to TestRail is desired.
Make a copy of of the testrail.example.cfg:
$ cp testrail.example.cfg testrail.cfg
Replace the example values with the appropriate values:
[API]
url = https://instance.testrail.net/
email = [email protected]
password = api_key
To run the tests only for webview and a specific set of tests:
$ pytest -m webview -k test_home --testrail --testrail-name release01 tests/
Consult the pytest-testrail project README.md
for more options
https://github.com/allankp/pytest-testrail
Use the markers.text_case
decorator with case number to upload the results to TestRail. More than one test case can be used by separating with a comma.
@markers.test_case('C10000', 'C10001')
def test_foo_uploads_bar:
You can find more in depth documentation in our Pytest Selenium Training.
This testing framework heavily relies on the PyPOM. The PyPOM library is the Python implementation of the PageObject design pattern.
The PageObject pattern creates a nice API abstraction around an HTML page allowing the test creator to focus on the intent of a test rather than decyphering HTML code. This design pattern makes the test framework much more maintainable as any code changes to the page can occur in the PageObject rather than within the test code.
According to Siman Stewart,
If you have WebDriver APIs in your test methods, You're Doing It Wrong.
The usage of pytest, pytest-selenium plugin, and the PageObject pattern allows for a succinct test structure like so:
from tests import markers
from pages.home import Home
@markers.webview
@markers.nondestructive
def test_nav_is_displayed(webview_base_url, selenium):
# GIVEN the main website URL and the Selenium driver
# WHEN The main website URL is fully loaded
page = Home(selenium, webview_base_url).open()
# THEN The navbar is displayed
assert page.header.is_nav_displayed
The inspiration for this framework is based on the Mozilla Addons Server Project and plenty of examples can be gleamed from their fantastic usage of the pattern.
i. go to ce-scripts repo and create virtual environment:
pyenv virtualenv 3.7.x virtualenv_name
activate it:
pyenv activate virtualenv_name
install requirements:
pip install -r requirements.txt
ii. cd python
run
gen_book_uris.py <archive_host> <cnx_uuid>
e.g. $ ./gen_book_uris.py archive-staging.cnx.org e42bd376-624b-4c0f-972f-e0c57998e765
iii. this will create url variations as per rex_redirects.py and stores it in output/<cnx_uuid.txt> file
e.g. output/e42bd376-624b-4c0f-972f-e0c57998e765.txt
iv. rex_redirects.py file can be found here: rex_redirects
vi. copy the created cnx_uuid.txt file into cnx-automation folder: fixtures/data/webview/
vii. in fixtures/webview.py add fixture to read and iterate through by tests:
tests/webview/integration/test_redirects_301.py
tests/webview/integration/test_rex_redirects.py