phobos commit, revision 2013


user: dsimcha

msg:
Make SortedRange work with const/immutable arrays.

http://www.dsource.org/projects/phobos/changeset/2013

On 9/16/2010 5:32 PM, dsource.org wrote:
> phobos commit, revision 2013
> 
> 
> user: dsimcha
> 
> msg:
> Make SortedRange work with const/immutable arrays.
> 
> http://www.dsource.org/projects/phobos/changeset/2013
> 

The phobos tests started failing after this submit.

>From http://d.puremagic.com/test-results/test_data.ghtml?dataid=3112:
  Testing generated/linux/debug/unittest/std/range
  std/range.d(1154): Error: can only initialize const member _input inside
constructor

1154 is the assignment, but the problem was likely triggered by adding the const unit test at line 5288.

For some reason, it doesn't look like the win32 build failed.  The test output is very different between windows and linux.  My guess is that windows just isn't running most of the tests or something.

(The linux32 failures from 9:30am through 5:00pm was due to the phobos library
output directory changes (posix -> linux).. so those can be ignored.)

Later,
Brad

  Yea, I really haven't the slightest clue why the tests don't fail on
Windows, unless the debug code just doesn't run.  They don't fail when I
run the unittests from CodeBlocks on the individual file either, hence
my completely missing this and checking in this code.

I've checked in a (hopefully) fix that I tested by commenting out the debug line and making that code always run.

On 9/16/2010 11:56 PM, Brad Roberts wrote:
> On 9/16/2010 5:32 PM, dsource.org wrote:
>> phobos commit, revision 2013
>>
>>
>> user: dsimcha
>>
>> msg:
>> Make SortedRange work with const/immutable arrays.
>>
>> http://www.dsource.org/projects/phobos/changeset/2013
>>
> The phobos tests started failing after this submit.
>
> > From http://d.puremagic.com/test-results/test_data.ghtml?dataid=3112:
>    Testing generated/linux/debug/unittest/std/range
>    std/range.d(1154): Error: can only initialize const member _input inside
> constructor
>
> 1154 is the assignment, but the problem was likely triggered by adding the const unit test at line 5288.
>
> For some reason, it doesn't look like the win32 build failed.  The test output is very different between windows and linux.  My guess is that windows just isn't running most of the tests or something.
>
> (The linux32 failures from 9:30am through 5:00pm was due to the phobos library
> output directory changes (posix ->  linux).. so those can be ignored.)
>
> Later,
> Brad
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos
>