Description
Issue:
During the execution of rfbrowser tests in the provided example, it's apparent that with the given command input, you should be capable of running three tests concurrently, each in a distinct browser. This functionality operates correctly without the --argumentfile option, as expected. However, upon using it, the tests are instead executed sequentially, resulting in each browser being opened three times. For instance, if three processes are initiated, each one will execute Test 1. Once that's done, the same will repeat for Test 2... etc
Expected behaviour?:
When utilizing --testlevelsplit alongside --argumentfile, the expected behavior is for the execution to correctly manage three distinct browsers concurrently, as specified per test case.
cmd input:
pabot --testlevelsplit --processes 3 --argumentfile someargumentfile.robot samplerobotfile.robot
pip libraries and resources:
robotframework==7.0
robotframework-browser==18.1.0
robotframework-pabot==2.18.0
python version == 3.12.2
node version == v20.9.0
Sample code:
*** Settings ***
Library Browser
*** Test Cases ***
Test 1
[Tags] some_tag
[Template] Test templates for opening browser ${browser_name}
chromium
Test 2
[Tags] some_tag
[Template] Test templates for opening browser ${browser_name}
firefox
Test 3
[Tags] some_tag
[Template] Test templates for opening browser ${browser_name}
webkit
*** Keywords ***
Test templates for opening browser ${browser_name}
Open Browser https://www.google.com ${browser_name}
Go to https://www.amazon.com
Go to https://www.ebay.com
Go to https://www.facebook.com
Go to https://www.yahoo.com
Log ${browser_name}
Sample argument file
--include some_tag