[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <3F92038E.1020803@jackhammer.org>
From: pdt at jackhammer.org (Paul Tinsley)
Subject: Question: is this exploitable?
"Escaping quote characters might work OK in MySQL, but it is at best
only a database-dependent solution."
Nobody said anything about simply quoting a string, if you read the
description I posted of quote, it does more than that. The function
that we are talking about IS part of DBI, not some crazy cooked up thing
that was written just for MySQL. I never quoted the DBD::mysql
documentation, all of that came directly from DBI. If the driver writer
implements all the calls DBI documents that are available, this should
work fine. If not, it's a problem with the driver, not with the user.
"You'd have to write an entirely different mechanism to untaint data
bound for Oracle...and another one for other different database
implementations."
That is DBD's job...
"For one, they keep you in a database-independent environment (which
makes sense, since you're using DBI)."
Good thing he was suggesting to use part of DBI.
As for which is the better of the two ways, there was no argument
there. I was simply answering your question as to how it protected from
SQL injection.
Thanks,
Paul Tinsley
P.S. - If you wish to further debate it, I suggest we take it off list,
we have definitely gone off topic at this point.
Jonathan A. Zdziarski wrote:
>Escaping quote characters might work OK in MySQL, but it is at best only
>a database-dependent solution. Take a look at Oracle, instead of
>double-quotes, single-quotes are used. And instead of being escaped,
>they are simply doubled (e.g. ' becomes ''). You'd have to write an
>entirely different mechanism to untaint data bound for Oracle...and
>another one for other different database implementations.
>
>This is why placeholders are a better solution. For one, they keep you
>in a database-independent environment (which makes sense, since you're
>using DBI). For another, they insure you don't have to worry about
>accidentally missing the escaping of some data.
>
>On Sat, 2003-10-18 at 22:36, Paul Tinsley wrote:
>
>
>>I don't believe this is a true statement.
>>
>>
>
>
>
>
Powered by blists - more mailing lists