Gennet/Kurtosis info into WLS with master changes

[AlbertoSoutullo]

Another big PR, changes:

• Using new nwaku image that uses base64 payloads
• Changed wls.star to use base64 payloads
• Analisys branch is already updated with latest master branch atm, so it was easier to do this refactor as it was very behind compared with master (half of the changes are because of this).
• Now gennet topology and service information (like ip, ports and so on) is passed to WLS.
• Refactored WLS module in several files to easier readability, maintainability and testing (can still be improved).
• (bug) Nodes now correctly use --port-shift value when preparing ports in Kurtosis
• (bug) Nodes now correctly use --port-shift value when preparing CMD in Kurtosis
• (bug) Now WLS is correctly taking its configuration from config.json instead wls.yaml
• (bug) Fixed mismatching nodes with topics. Topics were being selected from nodes that were not correspondant.
• (bug) Arguments in WLS were not parsed, it was getting always default values.
• (bug) Parameters in parser.add_argument were wrong (regarding action=store_true) making argparser not working correctly if arguments were passed.
• (bug) uniform mode in wls was not working correctly in WLS. Fixed.
• Cleaned unnused variables.
• Added tests for WLS in python and Starlark (related with Test python modules #45 and Gennet: tests #68 ). Gennet is still missing tests.
• Fixed old tests, now multinode approach is tested correctly.

Remember that in order to test the tests in starlark, you have to change the name of tests.star to main.star and do kurtosis run . while this process is not automatized with CI.

In order to run python tests, you can go to the desired directory (as wls-module) and run python -m unittest discover .

Closes #49 , closes #88 , closes #94 ,

[0xFugue]

Glad we all switched to dispatchers! The WSL injection seems like a busy loop: if _time_to_send_next_message(...) is false, may be it is better to sleep for the remaining time as opposed to spin in while True. But LGTM otherwise.

[AlbertoSoutullo]

Glad we all switched to dispatchers! The WSL injection seems like a busy loop: if _time_to_send_next_message(...) is false, may be it is better to sleep for the remaining time as opposed to spin in while True. But LGTM otherwise.

This is indeed super good feedback and you are right. I would change it but we are planning to change this to non-blocking message inyection so it will be changed anyway.