dupl things two switch/case clauses are the same when they differ in variable type creation. Mention the actual behaviour in README. about dupl

[mibk]

This is just an issue related to the method dupl uses for clone detection. It's a "false positive" match. The algorithm works only with token types, not their values. From certain angle I would say it is a duplicate because if both cases are that similar, their behaviour could be probably abstracted to an interface.

Ad 2. issue: I think it works as expected. You can check #2 where we have discussed the desired output format.

Did I miss something?

[oryband]

actually, this switch case is for initializing different structs which implement the same interface, so there isn't any further abstraction i can do in this case. if you say this is a FP than ok, but I think the chance for FPs should be mentioned in the README or something.

[oryband]

it also means i should unfortunately take this tool with a grain of salt, since not all outputs are TP :\

[mibk]

That's just the way it is. You are right it should be more emphasized in README. It only contains one sentence about it (It ignores values of AST nodes, it cares just about the types.) so far. The reason is English is my second language. So I find it quite uncomfortable to write long passages of text.

dupl is a tool, it is not a source of truth. It should provide a user information about possible clones that would have been otherwise missed by the user.

[dmitshur]

Yeah, dupl is not a tool whose output can be run in CI and non-zero output can be considered a failure, the way, for example, vet is. It's more like golint, its output are hints, some of which may be helpful, some others may be noise.

On another note, in your original code, couldn't you factor out:

dec := json.NewDecoder(bytes.NewReader(b))
dec.UseNumber()

It seems that part is duplicated without necessity. It case it happens to lead to more readable, better code to factor it out, then dupl has been helpful even in this relatively false positive-y case.

[oryband]

I agree, and maybe you should quote from here in the README.

[oryband]

thanks for the quick response :)
hurray for awesome maintainers 👍

[mibk]

You're welcome :-)