Re: Case sensitivity

From: "Stephan Szabo" <sszabo(at)kick(dot)com>
To: "Brian Piatkus" <brian(at)brianp(dot)demon(dot)co(dot)uk>
Cc: <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Case sensitivity
Date: 2000-07-05 21:09:00
Message-ID: 049201bfe6c5$3fec4da0$0c64010a@kick.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

You haven't given the hba.conf entry you are using, so I'm assuming
you're using ident with the sameuser usermap.

You can create mixed case names by double quoting the name
(as mentioned by Tom Lane)

Or, if you're using both unix and NT machines so the names come
out differently for the same users, you might be able to get away
with just using a usermap file. If you really didn't feel like making
and keep the usermap up to date, you might want to look at how
the sameuser usermap is defined and make an equivalent
that did what you wanted (or do case insensitive matches),
probably something like [not at all tested or even tried to compile]:

else if (strcmp(usermap_name, "caseless_sameuser")) == 0)
{
if (strcasecmp(ident_username, pguser) == 0)
*checks_out_p = true;
else
*check_out_p = false;
}

I think the changes below are a bit drastic and make sure that
noone will ever be able to use mixed case names with it and
also are assuming that the usernames can't contain any characters
that might get translated by the |= that aren't A-Z.

----- Original Message -----
From: "Brian Piatkus" <brian(at)brianp(dot)demon(dot)co(dot)uk>
To: <pgsql-hackers(at)postgresql(dot)org>
Sent: Tuesday, July 04, 2000 1:37 PM
Subject: [HACKERS] Case sensitivity

> There seems to be a general issue here with usernames.
>
> PG creates only lower case usernames. I am happy that if usEr can't type,
it's
> his problem but I have a definite problem with Identd authentication from
an NT
> client (or ident server whichever you choose ) which insists on
pretty-printing
> the identd response.
>
> I have therefore hacked my code in auth.c. This seems to be of no other
> consequence but I am not a C programmer nor did I have time to check it
out
> elsewhere.
>
> I may be masking another, more general, issue but this does for me ! Can
we get
> it or equivalent in future releases please ?
>
> Regards.
>
> BTW The actaul code works - this is a cut&paste.
>
>
*---------------------------------------------------------------------------
> Talk to the ident server on the remote host and find out who owns the
> connection described by "port". Then look in the usermap file under
> the usermap *auth_arg and see if that user is equivalent to
> Postgres user *user.
>
> Return STATUS_OK if yes.
> --------------------------------------------------------------------------
-*/
> bool checks_out;
> bool ident_failed;
>
> /* We were unable to get ident to give us a username */
> char ident_username[IDENT_USERNAME_MAX + 1];
> !!!!!!!!! char* c = ident username;
>
>
>
> /* The username returned by ident */
>
> ident(raddr->sin_addr, laddr->sin_addr,
> raddr->sin_port, laddr->sin_port,
> &ident_failed, ident_username);
>
> if (ident_failed)
> return STATUS_ERROR;
>
> ||||||| while (*c) { *c++ |= 0x20;}
>
> verify_against_usermap(postgres_username, ident_username, auth_arg,
> &checks_out);
>
> return checks_out ? STATUS_OK : STATUS_ERROR;
> }
>

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2000-07-05 21:18:49 Re: 2nd update on TOAST
Previous Message Jan Wieck 2000-07-05 21:04:45 Re: [HACKERS] Re: Revised Copyright: is this morepalatable?