On 05/14/2011 01:29 AM, Jan Zeleny wrote:
> Hi,
> I'm sending the first version of the patch solving issues with nested
> groups without GID. The patch is definitely not final, let's say it's
> the first draft of final patch. It depends on Jakub's recent series of
> rfc2307bis patches and it won't be final until they are pushed.
>
> At this moment I'd like to hear your opinion on the concept of the patch.
>
> Thanks in advance
> Jan
Instead of modifying the global search filter to not include GID and
using the generic sdap_nested_group_lookup_group(), why not create a
separate function with a special non-posix search filter as a fallback
if neither user nor posix-group filters match?
Frankly I don't like the idea of creating new functions with almost the same
code. The file has 5.5k lines as it is and I don't want to add more just to
separate couple lines of code.
I haven't done a full review but two other comments -
I think that a sysdb_attrs_get_bool function would be handy instead of
handling the TRUE/FALSE strings ourselves.
I was considering this too, thanks for this comment :-)
The SYSDB_POSIX attribute should be named differently to indicate it
is
a bool, "isPosix" perhaps. "posixGroup" is commonly used as
objectlass.
Good thought. I'll change it.
Jan