refactor(generic): add a seed to the Shuffle method

[mateusfdl]

What?

Why?

How?

[codecov-commenter]

Codecov Report

Base: 98.18% // Head: 98.19% // Increases project coverage by +0.00% 🎉

Coverage data is based on head (64cc264) compared to base (60d95f9).
Patch coverage: 100.00% of modified lines in pull request are covered.

Additional details and impacted files

@@           Coverage Diff           @@
##             main      #91   +/-   ##
=======================================
  Coverage   98.18%   98.19%           
=======================================
  Files           7        7           
  Lines        1103     1106    +3     
=======================================
+ Hits         1083     1086    +3     
  Misses         14       14           
  Partials        6        6           

Impacted Files	Coverage Δ
generic.go	100.00% <100.00%> (ø)

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[joaofnds]

I think we should either add a new method that accepts a seed or leave it to the user to set the seed on rand directly, like this:

rand.Seed(time.Now().UnixNano())
collections.Shuffle(slice)

[mateusfdl]

I think we should either add a new method that accepts a seed or leave it to the user to set the seed on rand directly, like this:

rand.Seed(time.Now().UnixNano())
collections.Shuffle(slice)

I strongly disagree. This approach couples the client to generate their seed through the rand package. We added a layer of abstraction on top of rand.Shuffle, passing a callback through the collection.Shuffle we are not telling the client the details of the Shuffle implementation, and it allows them to create their seed mechanism

[joaofnds]

This is the way Go developers are used to doing things with rand. It is, IMO, the approach that causes the least surprise to Go developers.

[mateusfdl]

My first thought was: it ain't the "go way". On the other hand, I realized this approach forces the client to set it explicitly; otherwise, it will use the default seed