[PATCH] nss-srv-tests: Add regression test for option re_expression

[Lukas Slebodnik]

ehlo,

Attached patches are unit test for ticket
https://fedorahosted.org/sssd/ticket/2487

The ./nss-srv-tests passed with sssd-1.12.1.

LS

[Jakub Hrozek]

On Tue, Nov 18, 2014 at 11:23:47AM +0100, Lukas Slebodnik wrote:
>  ehlo,
>  
>  Attached patches are unit test for ticket
>  https://fedorahosted.org/sssd/ticket/2487
>  
>  The ./nss-srv-tests passed with sssd-1.12.1.
>  
>  LS 
>  From ee8a9a9d65f1ca71d945c95eb664d1ff7f04cfc5 Mon Sep 17 00:00:00
> 2001
>  From: Lukas Slebodnik <lslebodn(a)redhat.com&gt;
>  Date: Tue, 18 Nov 2014 11:09:51 +0100
>  Subject: [PATCH 1/2] TESTS: Add function to purge optionf from confdb 
Isn't it better to always set up and teardown the whole db instead of
cherry-picking the right parameters?

[Lukas Slebodnik]

On (20/11/14 14:01), Jakub Hrozek wrote:
> On Tue, Nov 18, 2014 at 11:23:47AM +0100, Lukas Slebodnik wrote:
> > ehlo,
> > 
> > Attached patches are unit test for ticket
> > https://fedorahosted.org/sssd/ticket/2487
> > 
> > The ./nss-srv-tests passed with sssd-1.12.1.
> > 
> > LS
>
> > From ee8a9a9d65f1ca71d945c95eb664d1ff7f04cfc5 Mon Sep 17 00:00:00 2001
> > From: Lukas Slebodnik <lslebodn(a)redhat.com&gt;
> > Date: Tue, 18 Nov 2014 11:09:51 +0100
> > Subject: [PATCH 1/2] TESTS: Add function to purge optionf from confdb
>
> Isn't it better to always set up and teardown the whole db instead of
> cherry-picking the right parameters? 
The purpose is to remove just some options from
confdb and do not tuch sysdb.

rebased patches on top of the tatest prebzina's changes.

LS

[Jakub Hrozek]

On Wed, Jan 14, 2015 at 03:54:40PM +0100, Lukas Slebodnik wrote:
>  On (20/11/14 14:01), Jakub Hrozek wrote:
>  >On Tue, Nov 18, 2014 at 11:23:47AM +0100, Lukas Slebodnik wrote:
>  >> ehlo,
>  >> 
>  >> Attached patches are unit test for ticket
>  >> https://fedorahosted.org/sssd/ticket/2487
>  >> 
>  >> The ./nss-srv-tests passed with sssd-1.12.1.
>  >> 
>  >> LS
>  >
>  >> From ee8a9a9d65f1ca71d945c95eb664d1ff7f04cfc5 Mon Sep 17 00:00:00 2001
>  >> From: Lukas Slebodnik <lslebodn(a)redhat.com&gt;
>  >> Date: Tue, 18 Nov 2014 11:09:51 +0100
>  >> Subject: [PATCH 1/2] TESTS: Add function to purge optionf from confdb
>  >
>  >Isn't it better to always set up and teardown the whole db instead of
>  >cherry-picking the right parameters?
>  The purpose is to remove just some options from confdb and do not tuch sysdb.
>   
I know what is the purpose, but it's wrong :-)

It's better if the tests are isolated and don't depend on one another.

[Jakub Hrozek]

On Wed, Jan 14, 2015 at 04:47:10PM +0100, Jakub Hrozek wrote:
>  On Wed, Jan 14, 2015 at 03:54:40PM +0100, Lukas Slebodnik wrote:
>  > On (20/11/14 14:01), Jakub Hrozek wrote:
>  > >On Tue, Nov 18, 2014 at 11:23:47AM +0100, Lukas Slebodnik wrote:
>  > >> ehlo,
>  > >> 
>  > >> Attached patches are unit test for ticket
>  > >> https://fedorahosted.org/sssd/ticket/2487
>  > >> 
>  > >> The ./nss-srv-tests passed with sssd-1.12.1.
>  > >> 
>  > >> LS
>  > >
>  > >> From ee8a9a9d65f1ca71d945c95eb664d1ff7f04cfc5 Mon Sep 17 00:00:00 2001
>  > >> From: Lukas Slebodnik <lslebodn(a)redhat.com&gt;
>  > >> Date: Tue, 18 Nov 2014 11:09:51 +0100
>  > >> Subject: [PATCH 1/2] TESTS: Add function to purge optionf from confdb
>  > >
>  > >Isn't it better to always set up and teardown the whole db instead of
>  > >cherry-picking the right parameters?
>  > The purpose is to remove just some options from confdb and do not tuch sysdb.
>  > 
>  
>  I know what is the purpose, but it's wrong :-)
>  
>  It's better if the tests are isolated and don't depend on one another. 
To be more specific -- I would prefer if mock_confdb_domain() first
deleted the whole confdb and then only added the specified params.

Then, if we need per-test parameters, the test itself calls
confdb_add_whatever()