Explicitly trusting the content-type header value and the file
extension when they match alone is not a good idea.
Here's a process I have seen done in the past that seems to work fairly
effectively.
1) Is the content-type header value a valid file-type your expecting to
receive (HTML/XML/GIF)? If not, fail.
2) Does the content-type header value and file extension match with a
mime-type entry? If not, fail.
3) Is the file within certain size constraints? If not, fail.
4) Does the file have proper format according to what its claiming to
be? If not, fail.
5) If is an HTML file, then run it through some security filtering
libraries.
* Hopefully I got the steps right. Someone might want to double check
the logic flow *
With respect to the question, step 4 is where the details really
matter. While you could use the unix 'file' command (as suggest by
another poster) to determine actual file type, I would prefer another
approach.
Use the content-type header value that the files claims to be and parse
based on that premise. For instance if the file claimed to be GIF when
it hits step 4, run it through an image parser and see if there are any
errors. Usually when I see files uploaded via web interface, the
expect type of file is fairly limited for the most part. Normally maybe
a few types of text files (HTML, CSV, XML), pictures (GIF, JPG, PNG),
possibly mp3's, etc. I would handle each type of file on a case-by-case
basis.
I don't know exactly how the unix 'file' command works, but I believe
its going to make its best guess based on certain identifiable format
indicators. And if this is all your looking for, its a great util.
Personally I think its better to know if a file would actually parse
rather than just appears to be something it might not be.
Just preference between the different methods.
jeremiah-
On Sunday, January 9, 2005, at 01:22 PM, Ofer Shezaf wrote:
>
> Hi Jeremiah,
>
> I was researching lately the issue of ensuring that files (uploaded and
> downloaded) are of the right type.
>
> Do you think that matching extension and content type header would be
> enough? If no, are you aware of any technology to determine a file type
> according to its content?
>
> ~ Ofer
>
> Ofer Shezaf
> CTO, Breach Security
>
> Tel: +972.9.956.0036 ext.212
> Cell: +972.54.443.1119
> ofers_at_breach.com
> http://www.breach.com
>
>> -----Original Message-----
>> From: Jeremiah Grossman [mailto:jeremiah_at_whitehatsec.com]
>> Sent: Saturday, January 08, 2005 3:44 AM
>> To: Alfred Hitchcock
>> Cc: webappsec_at_securityfocus.com
>> Subject: Re: Content monitorting in Application Security
>>
>> Sounds like common web site functionality and the resulting security
>> challenge.
>>
>> Here are techniques that may help...
>>
>> 1) When receiving an uploaded file of any kind, use various parser
>> libraries to sanity check the actual format of data. Ensuring the file
>> being uploaded is what it claims to be. With the incoming file
>> extension and content type header in agreement. jpeg's should be
>> formatted like jpegs, mp3's like mp3's, html like html and so on.
>>
>> 2) If you plan on handling files beyond plain text, such as zips and
>> exe's, you may consider using some type of A/V product as well. A nice
>> security add-on that can be useful depending on the situation.
>>
>> 3) This following method is strictly about XSS and HTML/JavaScript
>> content.
>>
>> While its fairly easy to filter all HTML tags from a file to prevent
>> XSS, its exponentially harder to separate HTML from executable
>> client-side code (JavaScript). Especially when the HTML is freeform
> and
>> most tags need to be supported on the web site. I've long said its a
>> slippery slope to support use-submitted HTML, but sometimes it can't
> be
>> helped.
>>
>> There are a few things than can do help mitigate the risk of the
>> uploaded files.
>>
>> a. Filter out potentially malicious HTML tags or only allows a
>> strict
>> set of safe HTML tags.
>> b. Filter out potentially malicious tag attributes or only
> allows a
>> strict set of safe tag attributes.
>>
>> * The either or is a give and take of security vs.
>> functionality/ease-of-use.
>>
>> Depending on the programming language you are using, there might
> be
>> some libraries available that could help make this process easier. I
>> haven't used them, but I noticed there are libraries available for
> Perl.
>>
>> http://cpan.uwinnipeg.ca/dist/HTML-StripScripts
>> http://cpan.uwinnipeg.ca/dist/HTML-Scrubber-StripScripts
>>
>> There might be some available if you use some other language.
>>
>>
>> best of luck!
>>
>>
>> jeremiah-
>>
>>
>>
>>
>>
>>
>> On Friday, January 7, 2005, at 04:55 AM, Alfred Hitchcock wrote:
>>>
>>> Hi All,
>>> I have a major doubt it would be of great help if anybody can
> provide
>>> solution to this.
>>> I have a web page which allows to upload files such as jpeg and html
>>> files.
>>> Is there any mechanisms which can detect malicious html files. E.g.
> if
>>> a html page has got a malicious java script such as alert('xss')
> then
>>> how can we check these things. One more point to be noted here is
> that
>>> uploading of file can be done by any user.
>>>
>
Received on Jan 13 2005
Intro
