Hi,
thanks for your response, I am able to reproduce it when I execute it locally:
1. yes, the working dir looks good
2. agree the whole confusion is caused by thinking the current project has modules, so then it makes sense that it ignores the sources of the parent module
3. based on the local reproduction it seems you are right it is looking for the source directory in the source directory
what is confusing to me is the reported key, it is not what is being set by me, but maybe this is some sonar generated key for the source folder? (com.bla.bla.bla:bla_bla -> lets call it com.a.b.c:bla_bla from now on where bla_bla is the source folder)
So I did a test and configured the sonar.sources to contain bla_bla2, and in this case I dont see the multi module error anymore, but the "folder does not exists", which can be because the check is earlier in the code. So I copied the sources to be in bla_bla2 and I got the multi module error again. Same with renaming it to blabla2
Then I thought maybe some sonar python setting is causing troubles so I removed those, but still the same.
At this point I decided to start to debug whats going on, and then I finally found whats causing the problem, it seems the sonarqube server returns with some unexpected properties for the project (maybe a misconfigured attempt to analyse before) my command line argument does not contain sonar.modules but if I set a break point at ProjectReactorBuilder#233 (5.6.3 sonarqube) I see its being set (or some other mechanism sets it not the server). Any idea what is going on? How can I fix it without deleting my other project from sonarqube server.
Thanks for your help
Cheers
Tamas