CASSGO-77: Add golangci-lint integration

[dkropachev]

golangci-lint is standard de facto in golang community.
Using will help with:

• Having code uniformity
• Enforcing certain linting rules
• Prevent bugs
• Prevent unused variables, increasing driver efficiency
• Easy and speedup reviewing proceccess by detecting problems early

 
govet/fieldalignment stands out from rest of the linters, it will help to define structures in a way that will make them occupy less memory and provide better productivty as result.
 

[joao-r-reis]

I don't think it's a good idea to say something like "X is standard de facto in golang community" when advocating for a change because a lot of the times when I search online to determine whether this is true or not I'll find a lot of discussion around it and not really any consensus.

In this particular case from what I'm reading it looks like golangci-lint is just a way of running multiple linters so you could just add golanci-lint and run the go vet linter through golanci-lint. Instead what this PR is doing is adding a bunch of more linters instead of just go vet in addition to adding golangci-lint integration.

Personally I'm fine with using golangci-lint but I think the current list of linters that you have is way too large and as a result the diff is huge. This will cause all existing open PRs to have giant conflicts and personally I don't think it's worth the hassle. Also it will probably make it very frustrating to contribute to the driver because how harsh the linter checks are. Maybe you can try running golangci-lint with the default linters only and see if it's a good compromise.

[dkropachev]

I don't think it's a good idea to say something like "X is standard de facto in golang community" when advocating for a change because a lot of the times when I search online to determine whether this is true or not I'll find a lot of discussion around it and not really any consensus.

In this particular case from what I'm reading it looks like golangci-lint is just a way of running multiple linters so you could just add golanci-lint and run the go vet linter through golanci-lint. Instead what this PR is doing is adding a bunch of more linters instead of just go vet in addition to adding golangci-lint integration.

Personally I'm fine with using golangci-lint but I think the current list of linters that you have is way too large and as a result the diff is huge. This will cause all existing open PRs to have giant conflicts and personally I don't think it's worth the hassle. Also it will probably make it very frustrating to contribute to the driver because how harsh the linter checks are. Maybe you can try running golangci-lint with the default linters only and see if it's a good compromise.

What we can do here, is to replace govet with golangci with minimal changes, and later on add more linters in a separate PRs, advocating for each of them separately.

[joao-r-reis]

What we can do here, is to replace govet with golangci with minimal changes, and later on add more linters in a separate PRs, advocating for each of them separately.

Yeah that is a good approach 👍

[dkropachev]

What we can do here, is to replace govet with golangci with minimal changes, and later on add more linters in a separate PRs, advocating for each of them separately.

Yeah that is a good approach 👍

Done, please take a look.

[joao-r-reis]

👍 can you squash commits and format the commit message like this? Then I'll merge