Welcome to the Test training

Demos

Unit performance tests using timeit

For the purpose of this demo we will compare several sort algorithms using timeit

Prerequisites:

• Python installed

Timeit

timeit will measure the time spent to execute some statements. We will use repeat method from timeit

times = repeat(setup=setup_code, stmt=stmt, repeat=1, number=1)

Timeit measure the time spent to run stmt executed number times.
repeat method runs it repeat times and return an array of measured time.

Measuring performances

perfo/perfo.py contains the code to run and measure the sorting algorithm.
The sorting algorithm are defined in simple/sort.py. You can define your own and test it

Work to do

• Define some test cases
  • array is already sorted (1, 3, 5, 9, 11)
  • array is already sorted but the min value is at the end (3, 5, 9, 11, 1)
  • array is sorted but in reversed order (11, 9, 5, 3, 1)
• For each of those determine which algorithm is the better for 10_000 values

Integration performance tests using locust

For the purpose of this demo we will test a simple REST API hosted locally in a docker image

Prerequisites:

• Docker installed https://docs.docker.com/docker-for-windows/install/
• Check if docker is properly installed

docker run "hello-world"

• Retrieve swaggerapi/petstore docker image

docker pull swaggerapi/petstore

• Run the container and check it is correctly serving the API at http://localhost:8080

docker run -d -e SWAGGER_HOST=http://localhost -e SWAGGER_BASE_PATH=/v2 -p 8080:8080 swaggerapi/petstore

Locust

Open locust_files\my_locust_file.py
Run locust
locust -f locust_files/my_locust_file.py --host http://localhost:8080

Optionaly an actual online rest API can be tested
https://petstore.swagger.io/

---

Pytest tutorial

Prerequisites

• Work in virtual environment

  python - m venv .venv
  .venv\scripts\activate
  

  or (depending on the sytem) source .venv/bin/activate
• Install the requirements

pip install requirements.txt

• Install the development requirements

pip install requirements_dev.txt

Work to do

Run the tests

pytest tests

1st exercise

Open tests\stringlib_test.py and analyse the code
Follow the instructions
Once all done check the test's coverage using this command line

pytest tests --cov=stringlib --cov-report html

Check the coverage and if needed add tests to reach 100% coverage

2nd exercise

Use Test Driven Development to code those requirements.
TDD principles:

1. Write the test
2. Run the test -> it must fail
3. Code the minimum feature to make the test pass
   3.1 Check the produced code is fully covered by test
4. Move on to the next requirement
5. At one stage refactor your code to make it simpler/nicer, to remove duplication

Requirements:

• Easy one to start : Write a function that returns "Hello world"
• Create a calculator (a class or a module or a file) with those methods:
  • add to compute the sum of two values
  • substr to compute the substraction of two values
  • multiply to compute the multiplication of two values
  • divide to compute the division of one value by the other
  • average to compute the average of a list of values

Ask if requirements are not clear enough
Test happy path, limit cases and expected error cases.
Use parametrized test if possible
Compute coverage and add tests to reach 100% coverage\

Once done, refactor you code to have:

• substr calling add
• divide calling multiply
• average calling statistics.mean

Integration tests tutorial

Prerequisites

• Install postman https://www.postman.com/downloads/
• Download swagger file from https://petstore.swagger.io/

Work to do

Open Postman Import the swagger file (API definition file)

Launch Find Pets By status Update it to test the result of the request In test tabs add :

pm.test("Status test", function () {
    pm.response.to.have.status(200);
});

Do the same for the Add a new pet query
Duplicate this query and create a negative test (wrong id) with expected 500 code as answer

Work to do:

1. Get a valid pet ID using Find Pets by status and use it in Find pet by ID, add a test on 200 status
2. Using a really big ID add a negative test on Find pet by ID and test 404 status
3. Using Add a new pet create a well known Pet and test Find pet by ID with this ID and add a test checking the 'name' attribute
4. Create a test suite with
   • a test creating one Pet setting the ID as global variable and testing 200
   • a test getting the Pet using the ID of the previous one and testing the name
   • a test deleting the Pet using the ID and testing the result

Open a runner tab, launch the full test suite and check the results
It can also be done without the GUI, in command line using newman, example using the docker image of newman
docker run -v /mnt/d/Projects/test_training/postman:/etc/newman -t postman/newman:ubuntu run "tests.json"

System test or end to end (e2e) test with selenium

Prerequisites:

• install selenium libraries pip install selenium

• install selenium web driver https://www.selenium.dev/documentation/en/selenium_installation/installing_webdriver_binaries/ for chrome https://sites.google.com/a/chromium.org/chromedriver/getting-started

• install selenium ide https://www.selenium.dev/selenium-ide/

Selenium test with IDE

Launch selenium IDE and start recording a search on python.org (PEP 604) Save the test
Play the test

Selenium test from python

Open and run selenium_test.py
Doc: https://selenium-python.readthedocs.io/)
Demo of ide code generation

Work to do: test on keyce-it.fr Check the title of the pages with the ide

• Accueil
• Le campus
• Formations
• Contacts Generate python and run from python and change the code with contains (substr in str)