Updating Two Fer Analyzer to use our new analyzer convention

[manumafe98]

The goal is to update the analyzer to how we are currently implementing them

Todo:

• If we agree to change the format suggestion to informative update the design.md
• Also update comment on website-copy
• Fix conflicts after the merge of Adding analyzer for log-levels concept exercise #136
• Merge Adding analyzer comments for Log Levels website-copy#2328

[sanderploegsma]

I would like to note the following though:

1. I noticed you added a lot of extra smoke tests. This is nice, but since we have most of the approaches covered in unit tests already there is technically no need to add a smoke test for each approach as well. I try to aim for 2-3 smoke tests per exercise.
2. I'd like to point out that each exercise analyzer is free to choose the best approach regarding implementation. The fact that most analyzers apply the visitor pattern doesn't mean that all of them have to. There may be exercises where walking the AST using the .walk method is more beneficial, or perhaps a new approach works well for some. As long as they satisfy the existing Analyzer interface and produce correct results, I don't think we should enforce a convention. To that end, I would like to suggest that we don't refactor the HammingAnalyzer, simply to show that there are more ways to implement new analyzers.