[Scons-users] Glob() exclude doesn't match when used inside VariantDir

Tue Sep 22 06:00:44 EDT 2015

I would argue it's more for the developer list now.

V/R,
William

On September 22, 2015, at 10:39 AM, Florian Miedniak <florian.miedniak at web.de> wrote:

Hi Bill,

I just created a pull request for the issue: https://bitbucket.org/scons/scons/pull-requests/252/fixed-3011-glob-called-with-exclude-didnt/diff

Notes:

- Is it okay to discuss this further on scons-user?

- [offtopic] I just did a python bootstrap.py and was overwhelmed how simply it is to "build" SCons! The people maintaining our development environment just couldn't believe getting a rpm with a local scons build including my changes for deployment ...

-Florian

Gesendet: Donnerstag, 17. September 2015 um 15:24 Uhr
Von: "Bill Deegan" <bill at baddogconsulting.com>
An: "SCons users mailing list" <scons-users at scons.org>
Betreff: Re: [Scons-users] Glob() exclude doesn't match when used inside VariantDir

Thanks!

-Bill

On Thu, Sep 17, 2015 at 3:40 AM, Florian Miedniak <florian.miedniak at web.de> wrote: 

Hi Bill,

I just created an issue in bug tracker (#3011) and will provide a pull request as you suggested for further discussion.

-Florian

Gesendet: Mittwoch, 16. September 2015 um 16:46 Uhr
Von: "Bill Deegan" <bill at baddogconsulting.com>
An: "SCons users mailing list" <scons-users at scons.org>
Betreff: Re: [Scons-users] Glob() exclude doesn't match when used inside VariantDir

Florian,

Good catch on the bug still being open. I've closed it. (It should have been as the feature is (in theory) complete)

The preferred method for sending a patch would be to fork the scons repo and send a pull request.

Any chance we can ask you to do that?

That way the version history would properly represent who made the fix.

-Bill 

On Wed, Sep 16, 2015 at 7:35 AM, Florian Miedniak <florian.miedniak at web.de> wrote: 

Hi,

first of all, I'm very glad scons now has support for exclude patterns for Glob()! Compared to the prior workarounds the solution IMO fits
perfectly in the scons way of declaring artifacts.

Nevertheless, I think I've found a bug or at least a from my point of view unexpected behaviour when using Glob-excludes together with variant
dirs: When Glob is called within a separate SConscript, it fails to properly apply exclude patterns because the returned list of files contains path information (see attached testcase for details).

I think I found the place in FS.py (SCons 2.3.6) where things go wrong. With attached patch, the problem went away for me. But I'm not yet sure, if I really found the root cause or just a workaround. In addition to that, I'm very uncertain about possible side-effects of the patch ...

So I'd very like someone to review both the testcase and, if it is considered valid, the proposed patch.

An additional note: While tracking down the problem, I found these issue 2545 in state NEW whereas Git commits fc2659e and 81e0a43 seem
to fix this issue. Does the issue still being in state NEW mean, that implementation of the exclude feature is considered still incomplete?

Regards,

-Florian

_______________________________________________
Scons-users mailing list
Scons-users at scons.org
https://pairlist4.pair.net/mailman/listinfo/scons-users

_______________________________________________ Scons-users mailing list Scons-users at scons.org https://pairlist4.pair.net/mailman/listinfo/scons-users

_______________________________________________
Scons-users mailing list
Scons-users at scons.org
https://pairlist4.pair.net/mailman/listinfo/scons-users

_______________________________________________ Scons-users mailing list Scons-users at scons.org https://pairlist4.pair.net/mailman/listinfo/scons-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist4.pair.net/pipermail/scons-users/attachments/20150922/76d7c690/attachment.html>