Required technologies / libraries
- Java
- Maven
- Cucumber
- Selenium Webdriver
- Rest Assured
- Docker
Task 1.
✅ GUI test that will perform a search for a package on the InPost website by its number and check if it has a status as expected. List of packages with statuses:
- no: 520113014230722029585646, expected status: Delivered
- no: 520107010499997005638120, expected status: Passed for delivery
- No: 523000016696115042036670, expected status: The label was cancelled
- No: 520000011395200025754311, expected status: Delivered
Task 2.
✅ API test that will perform a Parcel Lockers search for a city (several cities) and save the data of the returned Parcel Lockers (name, postal code, coordinates) to the file 'parcellockers.{city}.json'.
Guidelines:
- ✅ publish the repository with the solved tasks on github.
- ✅ run tests from a Docker image
- ✅ when running tests it should be possible to indicate whether you want to run only GUI , API or all tests
- ✅ test results should produce html report
- ✅ GUI test report should contain screenshot in case of unsuccessful test result
- ✅ (*) simulation of running tests on few environments
- ✅ (*) docker-compose.yml
Deadline: usually 3 working days is enough, please let me know if you need more time.
In case of any questions regarding the task, do not hesitate to reach out to us. However please note that all needed informations are in the task and if you'd have any assumptions, feel free to document them.
You may send us your solution as a .zip or a link to your repository.
API tests are in inpost-api-tests
module and GUI tests are in inpost-gui-tests
module. In inpost-common
there is common code - it should already be built and installed locally, before running tests:
cd ./inpost-common
mvn clean install
cd ..
All tests, on default environment:
mvn clean test --fail-at-end
To run tests on environemnt add:
-D environment=prod
-D environment=sandbox
-D environment=sandboxpl
To run API tests only, add:
-D cucumber.filter.tags="@api"
or:
cd ./inpost-api-tests
mvn clean test
To run UI tests only, add:
-D cucumber.filter.tags="@gui"
or:
cd ./inpost-gui-tests
mvn clean test
To change browser, add:
-D browser=chrome
-D browser=firefox
For example, API tests on prod
environment:
mvn clean test -D environment=prod -D cucumber.filter.tags="@api" --fail-at-end
mvn clean test -D environment=prod -D cucumber.filter.tags="@gui" -D browser=chrome --fail-at-end
docker pull selenium/standalone-firefox
docker run -d -p 4444:4444 -p 7900:7900 --shm-size="2g" selenium/standalone-firefox
Check the grid at http://localhost:4444/ui/, see the browser action http://localhost:7900.
Run tests, for example:
mvn clean test -D environment=prod -D cucumber.filter.tags="@gui" -D grid=http://localhost:4444 --fail-at-end
docker compose -f docker-compose.yml up
Check the grid - same as above - at http://localhost:4444/ui/, see the browser action http://localhost:7900.
Run tests - same as above:
mvn clean test -D environment=prod -D grid=http://localhost:4444 --fail-at-end
To stop:
Ctrl+C
docker compose -f docker-compose.yml down
Build the image:
docker build -t inpota .
In the meantime, run the grid:
docker compose -f docker-compose.yml up
Run the container:
docker kill inpota
docker run --rm -t -d --name inpota inpota
Then run tests of your choice, for example:
docker exec inpota bash -c "./mvnw clean test --fail-at-end -D grid=http://host.docker.internal:4444"
Or run tests, collect results, open reports.
To clean up after runs:
docker compose -f docker-compose.yml down
docker kill inpota
open inpost-ui-tests/target/cucumber.html
open inpost-api-tests/target/cucumber.html
mkdir ./results/api/prod
docker cp inpota:inpost-api-tests/target ./results/prod/api
mkdir ./results/gui/prod
docker cp inpota:inpost-api-tests/target ./results/prod/gui
You can run docker_tests_run.sh
script which runs tests, collects them locally and opens reports. For example:
chmod u+x docker_tests_run.sh
./docker_tests_run.sh prod
./docker_tests_run.sh sandbox @api
./docker_tests_run.sh sandboxpl @gui
User is in charge of inputting city name in scenario. If they provide it somehow incorrectly, that's how it will be executed within API.
Ad. Task 1. Test data and desired effect don't match - left failed tests.
Ad. Task 2. Some data fails tests because it is bad - left failed tests. This is not deterministic - there are times when that bad data is not present on environment (even on prod) 🙃