A web service managing Robot Framework tasks/tests.
This web service shall start Robot Framework tasks/tests and return and cache the according reports.
You can run the image and map your test cases into the webservice with a volume :
docker run --rm --publish 5003:5003 \
--volume <host directory of test cases>:/robot/tests \
--env SUITE_FOLDER=tests \
ghcr.io/marketsquare/robotframework-webservice:master
You can also run the image and map your test cases and your variable files (separated by spaces) into the webservice with volumes :
docker run --rm --publish 5003:5003 \
--volume <host directory of test cases>:/robot/tests \
--volume <host directory of variable files>:/robot/variables \
--env SUITE_FOLDER=tests \
--env "VARIABLE_FILES=variables/variables.py variables/variables2.py" \
ghcr.io/marketsquare/robotframework-webservice:master
Almost as Docker, but you might need to attach the webservice to the host network:
podman run --network host -v ./examples:/robot/tasks --env SUITE_FOLDER=tasks rf-webservice:latest
pip install robotframework-webservice
and execute from command line:
python -m RobotFrameworkService.main -p 5003 -t path_to_my_taskfolder
There are 2 types of endpoints:
- Execution
- Reporting
Endpoints that trigger execution of a robot task/test, for instance:
Call robot task/test:
http://localhost:5003/robotframework/run/mytask
Call robot task/test with variables:
http://localhost:5003/robotframework/run/mytask?myVariable1=42&anotherVariable=Mustermann
Response contains a header field x-request-id
that can be used to retrieve logs and reports of this execution asynchronously (see reporting endpoints)
There are endpoints for synchronous and asynchronous request:
# connection remains open for duration of my task/test
http://localhost:5003/robotframework/run/mytask
# connection closes immediately - result must be requested with the x-request-id
http://localhost:5003/robotframework/run/mytask/async
There is also the all-in-one endpoint POST http://localhost:5003/robotframework/run
which trigger execution of a robot files, test, task or suite.
It can be customized with options in JSON payload. All available options are documented in Swagger schema and examples.
Response contains a header field x-request-id
that can be used to retrieve logs and reports.
By default, execution is asynchronous, but it can be changed with sync option.
There is no limitation on executed Robot processes! It is easy to push the webservice in DOS with too many requests at once
curl -X 'POST' \
'http://localhost:5003/robotframework/run' \
-H 'accept: application/json' \
-H 'Content-Type: application/json' \
-d '{
"paths": [
"examples"
],
"test": "Demonstration Test"
}'
curl -X 'POST' \
'http://localhost:5003/robotframework/run' \
-H 'accept: application/json' \
-H 'Content-Type: application/json' \
-d '{
"paths": [
"examples"
],
"task": "Demonstration Task"
}'
curl -X 'POST' \
'http://localhost:5003/robotframework/run' \
-H 'accept: application/json' \
-H 'Content-Type: application/json' \
-d '{
"paths": [
"examples"
],
"task": "Task with more variables",
"variables": {
"firstname": "Max",
"lastname": "Mustermann"
}
}'
Endpoints that provide log.html
and report.html
for a specific task execution. You require the x-request-id
from a previous response that triggered the execution.
The web service starts automatically with uvicorn inside. Simply call:
python -m RobotFrameworkService.main
You can check available options with
python -m RobotFrameworkService.main --help
python -m RobotFrameworkService.main -p 5003 -t path_to_my_taskfolder
You can provide variable files that are passed to all robot suites on execution:
python -m RobotFrameworkService.main -p 5003 -t path_to_my_taskfolder --variablefiles config/env/test.py
You can start RobotFrameworkService with bare WSGI servers:
uvicorn RobotFrameworkService.main:app --port 5003
Or start web service with other WSGI server, i.e waitress:
waitress-serve --port 5003 RotbotFrameworkService.main:app
Swagger-UI is available under http://localhost:5003/docs
This project contains some tasks, tests and variables for demonstration. They are located in examples
folder. You may add
your own task/test suites in that directory, if you like.
Tasks may contain spaces, URL must not. Luckily, Robot Framework supports CamelCase as well as snake_case syntax.
Meaning: "Another Task" can be trigger in url with parameter AnotherTask
or another_task