On 03/03/2011 06:37 PM, Jochen Bern wrote:
> On 03/03/2011 05:54 PM, Andreas Ericsson wrote:
>> I think that would be quite stupid, as it would prevent users from
>> using shell variables in their commands, or multiple dollar-signs
>> for that matter.
>
> Not quite, the handling of escaped dollar signs ("$$") is in a separate
> branch immediately preceding the "unknown macro name" branch. Hence, the
> properly escaped notation ("$$FOO$$BAR" in Nagios, becomes "$FOO$BAR" on
> shell level) would still be available.
>
Yes, but how about a command_line such as this:
$USER1$/check_something '$pas$word'
$pas$ is not a valid macro, so by the reasoning of the original poster
we should remove it, but that would mean there can *never* be two
dollar-signs that aren't part of macros inside anything where macros
get interpreted, and that would be a real limitation put in place to
prevent the minor discomfort of having to redo a botched job.
I just flat out refuse to accept patches with those effects.
> (I admit I wasn't aware that it's the *shell* which reduces "$FOO$" to
> "$", rather than Nagios.)
>
> On one hand, trusting Nagios to leave "$FOO$" be (rather than properly
> configuring "$$FOO$$") is not supported.
It is. If the dollar-signs are within single-quotes the shell won't even
touch them and '$FOO$' gets passed verbatim to whatever command one's
trying to run.
--
Andreas Ericsson [email protected]
OP5 AB www.op5.se
Tel: +46 8-230225 Fax: +46 8-230231
Considering the successes of the wars on alcohol, poverty, drugs and
terror, I think we should give some serious thought to declaring war
on peace.
This post was automatically imported from historical nagios-devel mailing list archives
Original poster: [email protected]