fix: Teststarter false positives detection

[d3xter666]

fixes: #972

[coveralls]

coverage: 86.796% (+0.01%) from 86.782%
when pulling f158970 on fix-teststarter-false-positive
into 6635730 on main.

[matz3]

Fixing false positives causes more warnings in our fixtures? Is that really the expected outcome? 🤔

[codeworrior]

Fixing false positives causes more warnings in our fixtures? Is that really the expected outcome? 🤔

I think this is caused by the fact that now every script tag is reported when no script tag indicates a use of the test starter. In the fixture, 3 scripts are included. Not sure whether this approach is helpful for developers. I understand that we can't easily pick the one script that should be changed. But wouldn't it be sufficient to create a single message for the whole HTML file?

[matz3]

Any update here? I should have explained my doubts more clearly. What surprised me is that there are now more messages for files where a message was already reported. To me this is unexpected and adds more false positives, as the concept can only be applied once per file, so I would expect only one message per file. Using the first script tag for the message seems reasonable to me (or whatever the current behaviour was).

[d3xter666]

Any update here? I should have explained my doubts more clearly. What surprised me is that there are now more messages for files where a message was already reported. To me this is unexpected and adds more false positives, as the concept can only be applied once per file, so I would expect only one message per file. Using the first script tag for the message seems reasonable to me (or whatever the current behaviour was).

Sorry, I've missed your comments!
I have now update the script so that it only reports one issue per file and points to the first script tag on the page