slice content checker

slice content checker

Checker for checking that the contents of two slices are the same (using DeepEquals) without respect for order

https://code.launchpad.net/~natefinch/juju-core/004-slicechecker/+merge/181096

(do not edit description out of merge proposal)

[natefinch, 2013-08-20 16:40:10]

Please take a look.

[natefinch, 2013-08-20 18:04:28]

Please take a look.

[natefinch, 2013-08-20 18:47:22]

Please take a look.

[rog, 2013-08-21 08:05:28]

LGTM with one last thought

https://codereview.appspot.com/13126043/diff/1/testing/checkers/checker.go
File testing/checkers/checker.go (right):

https://codereview.appspot.com/13126043/diff/1/testing/checkers/checker.go#ne...
testing/checkers/checker.go:170: return false, fmt.Sprintf("SameContents expects
the obtained value to be a slice, got %q",
i think i'd use %v and tob itself - nice to see the actual type.

[jameinel, 2013-08-21 10:33:46]

I have a bunch of comments on a different proposal where this showed up:
https://codereview.appspot.com/13082044/diff/1/testing/checkers/checker.go

[rog, 2013-08-21 12:45:26]

On 21 August 2013 11:33,  <john.meinel@canonical.com> wrote:
> I have a bunch of comments on a different proposal where this showed up:
> https://codereview.appspot.com/13082044/diff/1/testing/checkers/checker.go

I think you might find that the latest version has some changes that address
your concerns.

[jameinel, 2013-08-21 12:48:16]

LGTM, you addressed my comments in the other code, and you can remove the doc
comment since it doesn't apply anymore.

https://codereview.appspot.com/13126043/diff/7001/testing/checkers/checker.go
File testing/checkers/checker.go (right):

https://codereview.appspot.com/13126043/diff/7001/testing/checkers/checker.go...
testing/checkers/checker.go:157: // use it on very large slices
It looks like you actually fixed this to use 2 maps and compare them, so I think
you can remove this comment.

[gz, 2013-08-21 14:15:50]

LGTM.

https://codereview.appspot.com/13126043/diff/7001/testing/checkers/checker.go
File testing/checkers/checker.go (right):

https://codereview.appspot.com/13126043/diff/7001/testing/checkers/checker.go...
testing/checkers/checker.go:197: mexp := make(map[interface{}]int)
Shouldn't these both have the slice Len as the second param? Again, less for
optimisation, more for the why-not factor.