[Catalyst] Hair-pulling over YAML config

Carl Vincent C.A.Vincent at newcastle.ac.uk
Tue Jan 30 09:58:00 GMT 2007


>Message: 10
>Date: Mon, 29 Jan 2007 19:49:29 +0100
>From: "A. Pagaltzis" <pagaltzis at gmx.de>
>Subject: [Catalyst] Re: Hair-pulling over YAML config
>To: The elegant MVC web framework <catalyst at lists.rawmode.org>
>Message-ID: <20070129184929.GF2896 at klangraum>
>Content-Type: text/plain; charset=utf-8
>
>* Joel Bernstein <joel at fysh.org> [2007-01-29 18:30]:
>> I think the difficulty here is that Config::Any::_load has the
>> potential if running without use_ext=>1 to generate a lot of
>> spurious warnings caused by all the plugins to which a given
>> config file is offered, but which can't parse the file. So any
>> patch would presumably only report warnings if use_ext=>1. But
>> that still doesn't seem ideal. Suggestions?
>
>Just warn if the file fails with *all* parsers. We’ve got a
>config file; we can reasonably assume that one of the parsers
>will be able to make sense of it. If none of them can, since
>it’s still a config file, it’s obviously a broken one.

That's exactly what we need - outputting warnings at on of the ConfigLoader levels loses the content of the error message. Could always add a "Keep Quiet" option...

Cheers

Carl

-- 
     -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
     Carl Vincent             http://www.netskills.ac.uk/ (URL)
     Systems Manager                       0191 222 5003 (voice)
     Netskills, Newcastle University       0191 222 5001  (fax)
     Training  —  Accreditation  -  Consultancy  —  Development


More information about the Catalyst mailing list