Here is a draft which works fine for normal mount points:
http://cr.opensolaris.org/~pavelf/6778894-v2/
It uses sh(1) pattern matching features:
sh(1)
case word in [ pattern [ | pattern ] ) list ;; ] ... esac
A case command executes the list associated with the
first pattern that matches word. The form of the pat-
terns is the same as that used for file-name generation
(see File Name Generation section), except that a slash,
a leading dot, or a dot immediately following a slash
need not be matched explicitly.
=================
This solution is vulnerable to e.g. * expansion:
$ mkdir /\*
$ mount server:/export /\*
$ umountall -l
Nothing will be unmounted since every mountpoint will match against '/*'
and will be considered as a NFS mountpoint.
I do not expect that sysadmins use often '*' in names for mountpoints,
but such solution is not acceptable.
Any suggestions how to implement a robust grep in a shell?
--Pavel
Nicolas Williams wrote:
> On Tue, Dec 16, 2008 at 09:31:14PM +0100, Pavel Filipensky wrote:
>
>> If you provide a path to df it will use realpath() -> resolvepath(2) :
>>
>> $ truss df -n /WWW
>> resolvepath("/WWW", "/WWW", 1024) = 4
>>
>
> Ah!
>
>
>> But the point was to specify the path (mountpoint) to df, without it we
>> do not get any advantage from using df.
>> So we must parse the /etc/mnttab anyway.
>>
>
> OK. Glad I asked!
>
> Nico
>
