* Minor fixes.
* Improve coverage of QuickCheck generators.

Attachments

• testingimprovements.dpatch

The cabal file changes should be rejected, as they are just to keep 
haskell-src-exts happy and it's not clear that haskell-src-exts is doing 
the right thing here.

I've screened the rest.

OK for next 2 ones:

Thu Jan  6 13:57:04 WET 2011  Iago Abal <iago.abal@gmail.com>
  * Remove trailing space
Thu Jan  6 14:06:22 WET 2011  Iago Abal <iago.abal@gmail.com>
  * Clarify a commentary on commuteFiledir




Sun Feb  6 12:29:16 WET 2011  Iago Abal <iago.abal@gmail.com>
  * Add defaultToks to the export list of Darcs.Commands.Replace
  defaultToks will be used to create a QuickCheck generator for
TokReplace patches
  and it is desirable to reuse this definition to avoid code duplication.

OK since another (not yet reviewed) bundle by you imports defaultToks in
Darcs.Test.Patch.Prim.V1.



I'm trusting you and Ganesh on the next 4 patches:

Sun Feb  6 12:33:42 WET 2011  Iago Abal <iago.abal@gmail.com>
  * Fix bug in arbitraryTree causing it to break depth(result) <=
maxDepth postcondition

Sun Feb  6 13:22:24 WET 2011  Iago Abal <iago.abal@gmail.com>
  * Fix ArbitraryStateIn s (Tree p) instance to avoid generation of a
(non-senseful) initial sequence of empty trees

Sun Feb  6 13:31:39 WET 2011  Iago Abal <iago.abal@gmail.com>
  * Warn about the use of localIdentity for testing

Sun Feb  6 16:16:42 WET 2011  Iago Abal <iago.abal@gmail.com>
  * Fix arbitraryTree frequencies to significantly reduce empty trees


One remark on last patch: you should probably not leave the comment "--
The choice of NilTree was removed;" as it is, since after this patch, it
no longer makes sense.

Apart from the patch rejected by Ganesh I'm accepting the bundle, and
pushing it along with patch548, since compile+tests of darcs.net+these
bundles is fine.

On Tue, Feb 22, 2011 at 9:41 PM, Guillaume Hoffmann <bugs@darcs.net> wrote:

>
> Guillaume Hoffmann <guillaumh@gmail.com> added the comment:
>
> OK for next 2 ones:
>
> Thu Jan  6 13:57:04 WET 2011  Iago Abal <iago.abal@gmail.com>
>  * Remove trailing space
> Thu Jan  6 14:06:22 WET 2011  Iago Abal <iago.abal@gmail.com>
>  * Clarify a commentary on commuteFiledir
>
>
>
>
> Sun Feb  6 12:29:16 WET 2011  Iago Abal <iago.abal@gmail.com>
>  * Add defaultToks to the export list of Darcs.Commands.Replace
>  defaultToks will be used to create a QuickCheck generator for
> TokReplace patches
>  and it is desirable to reuse this definition to avoid code duplication.
>
> OK since another (not yet reviewed) bundle by you imports defaultToks in
> Darcs.Test.Patch.Prim.V1.
>
>
>
> I'm trusting you and Ganesh on the next 4 patches:
>
> Sun Feb  6 12:33:42 WET 2011  Iago Abal <iago.abal@gmail.com>
>  * Fix bug in arbitraryTree causing it to break depth(result) <=
> maxDepth postcondition
>
> Sun Feb  6 13:22:24 WET 2011  Iago Abal <iago.abal@gmail.com>
>  * Fix ArbitraryStateIn s (Tree p) instance to avoid generation of a
> (non-senseful) initial sequence of empty trees
>
> Sun Feb  6 13:31:39 WET 2011  Iago Abal <iago.abal@gmail.com>
>  * Warn about the use of localIdentity for testing
>
> Sun Feb  6 16:16:42 WET 2011  Iago Abal <iago.abal@gmail.com>
>  * Fix arbitraryTree frequencies to significantly reduce empty trees
>
> Just clarify that I'm not saying "I believe it reduces empty trees and
improve rate of commutable pairs", but I have measured that using QuickCheck
collect/classify.

>
> One remark on last patch: you should probably not leave the comment "--
> The choice of NilTree was removed;" as it is, since after this patch, it
> no longer makes sense.
>
Sorry, I will fix that comment in a future patch bundle.

>
> Apart from the patch rejected by Ganesh I'm accepting the bundle, and
> pushing it along with patch548, since compile+tests of darcs.net+these
> bundles is fine.
>
> __________________________________
> Darcs bug tracker <bugs@darcs.net>
> <http://bugs.darcs.net/patch539>
> __________________________________
>



-- 
Iago Abal Rivas

Attachments

• unnamed