Re: [Boost-testing] regression.py/Boost.Build V2

From: Vladimir Prus (ghost_at_[hidden])
Date: 2006-02-17 02:45:16

David Abrahams wrote:

> Vladimir Prus <ghost_at_[hidden]> writes:
> 
>> Hello,
>> I've modified the regression.py script to work with Boost.Build V2, but
>> adding a --v2 option and adjusting the code to check for that option when
>> needed. That works for me, but I'd like to ask for help with two things.
>>
>> 1. Even though I've tried that without --v2 the script works, I'm wary
>> about committing a change that can possibly brake all regression tests.
>> Can somebody take the attached file and verify it does not break V1
>> regressions.
>>
>> 2. It would be great if somebody could double-check that "--v2" option
>> works. There's no need to test it with many toolsets -- just verify that
>> with one toolset and --v2 option, regression.py runs successfully.
> 
> I get:
> 
> python regression.py --v2 --toolsets=msvc-8.0 --runner=V2test

FWIW, this will run the tests and *upload results*. I did not think that's a
good idea, and that's why the page I've given mentions specific commands to
be run, which does not include upload.

But anyway,

>     commands[ command ]( **accept_args( args ) )
>   File "regression.py", line 730, in regression
>     setup( comment, toolsets, bjam_toolset, pjl_toolset, monitored, proxy,
>     [] )
> TypeError: setup() takes exactly 8 arguments (7 given)

This error is fixed now, and revised version is attached.

Should you decide to run the script start-from-finish, note that in the end
it will try to update itself from CVS, which might overwrite my
modification.

- Volodya