Too many false positives

[ArneBachmannDLR]

Great project! I found at least one line of code that could be improved (moved loop to a dictionary expression); but I see, however, 99.5% false positives, e.g.:

• loop-try-except-usage can only be moved outside the loop, if the except clause contains a break. Otherwise I want to keep processing the loop
• loop-invariant-statement is usually not at all invariant but depends on a loop variable
• use-tuple-over-list is most often typed as a List[...] and therefore cannot and should not hold a tuple, even if initiated empty

[adamzev]

I enjoyed the talk and checking out the tool!

I ran into some issues with Pandas treating lists differently than tuples (when using them as keys).

I also hit a lot of false positives with loop-invariant-statements (unless I'm missing something).

I tried to simplify out an example of something it'd hit a false positive on:

keys = []
for record in ({'a': 4}, {'b': 8}):
    keys.append(tuple(k for k in record if k in ('b', 'c')))

print(keys)

I think it was saying k in ('b', 'c') was loop invariant.