[PATCH 6/8] urgent_replication: Use subunit reporting, remove allow_empty_output.

Jelmer Vernooij jelmer at samba.org
Mon Jan 26 19:27:50 MST 2015


On Mon, Jan 26, 2015 at 11:07:04AM +0100, Stefan (metze) Metzmacher wrote:
> Am 26.01.2015 um 10:50 schrieb Jelmer Vernooij:
> > I don't understand why we would be converting the URL to lowercase here. Your fix seems reasonable to me. Why do we need to still convert to lowercase in the paged searches case?
> 
> Before we had
> 
> if samdb_url.lower().startswith("ldap://"):
> 
> before.
> 
> See commit f1b3c4dd387ac04615c3443d5de1059d2dc59ba4
>   Author: Kamen Mazdrashki <kamenim at samba.org>
>   Date:   Sun Sep 26 02:13:39 2010 +0300
> 
>     s4-possibleinferiors.py: Fix usage of 'paged_search' module for
> remote LDB connections
> 
> 
> This means LDAP:// might also be supposed to work, but
> ldb_module_connect_backend() only handles 'ldap://'.
> 
> I'm also fine with removing only the "samdb_url = samdb_url.lower()" line
> and drop support for 'LDAP://'.

I think only handling ldap:// would be fine. I think it's actually
most important that we're consistent in what we support across the
Python and the C code.

startswith() is a terrible way of parsing URLs ;) The 'urlparse'
module (bundled with Python) in Python is probably better for parsing
the URL and extracting the scheme.

jelmer

-- 
Jelmer Vernooij <jelmer at samba.org> - https://jelmer.uk/


More information about the samba-technical mailing list