Closed GoogleCodeExporter closed 8 years ago
That's odd, thanks for reporting!
One weakness of the crawler is Javascript support. If these links are
dynamically added to the page, than that might be the cause. Based on your
description, I think this is not the case.
Can you perhaps send me a debug log ? (make debug) Or could you give some
larger page snippets ?
Original comment by niels.he...@gmail.com
on 19 Jul 2012 at 12:18
kindly find the file attached.
index.html has three links link1.php?id=1, link1.php?id=2, link1.php?id=3.
link1.php includes three different file on the basis parameter id, it includes
file1.php, file2.php, file3.php respectivily.
In this scenario contents of file2.php and file3.php will not crawled, as
skipfish crawler will ignore link1.php?id=2, link1.php?id=3
Original comment by anuragno...@gmail.com
on 20 Jul 2012 at 6:04
Attachments:
Thanks for the files! Can you perhaps give me the debug output as well ?
It might be the case that the parameter is marked as bogus. This can
incorrectly happen when the different parameter values result in "hardly
noticeable" difference. By design we allow small difference to occur while
still marking a page as similar / same.
(not saying this isn't something that can't be fixed. So debug output would
help)
Original comment by niels.he...@gmail.com
on 21 Jul 2012 at 4:11
kindly find the debug output:
== PIVOT DEBUG ==
== Pivot [root] [0] ==
Type : PIVOT_ROOT
State : PSTATE_DONE
Flags : linked 2, case 0/0, fuzz_par -1, ips 0, sigs 0, reqs 0, desc 1/4
== Pivot http://skipfish.test.web/ [0] ==
Type : PIVOT_SERV
State : PSTATE_CHILD_INJECT
Flags : linked 2, case 0/1, fuzz_par -1, ips 0, sigs 0, reqs 0, desc 2/3
Target : http://skipfish.test.web/ (200)
MIME : text/html -> - [UTF-8:-]
-> Issue : type 20201, extra 'no distinctive 404 behavior detected', URL: / (200)
== Pivot link1.php [0] ==
Type : PIVOT_FILE
State : PSTATE_CHILD_INJECT
Flags : linked 2, case 0/0, fuzz_par -1, ips 0, sigs 0, reqs 0, desc 1/1
Target : http://skipfish.test.web/link1.php (200)
MIME : text/html -> - [UTF-8:-]
-> Issue : type 10901, extra '(null)', URL: /link1.php (200)
== Pivot id [0] ==
Type : PIVOT_PARAM
State : PSTATE_PAR_INJECT
Flags : linked 2, case 0/0, fuzz_par 1, ips 0, sigs 1, reqs 0, desc 0/0
Target : http://skipfish.test.web/link1.php?id=1 (200)
MIME : text/html -> - [UTF-8:-]
Try : 1, 2, 3
== Pivot abc.php [0] ==
Type : PIVOT_PATHINFO
State : PSTATE_CHILD_INJECT
Flags : linked 2, case 0/0, fuzz_par -1, ips 0, sigs 0, reqs 0, desc 0/0
Target : http://skipfish.test.web/abc.php/ (200)
MIME : text/html -> - [UTF-8:-]
== END OF DUMP ==
Original comment by anuragno...@gmail.com
on 23 Jul 2012 at 6:06
could you please guide me, how i can fix this.
Original comment by anuragno...@gmail.com
on 24 Jul 2012 at 5:18
Is there any other information required from my side?
Original comment by anuragno...@gmail.com
on 27 Jul 2012 at 4:42
Ah sorry for the delay. I'm currently traveling but will check back next week.
If you by any chance could send me the full debug.log file (feel free to just
email it) than that would be great!
Original comment by niels.he...@gmail.com
on 28 Jul 2012 at 1:14
Without the full debug log I'm unable to look into this. I'm aiming to release
the next version next week and it has crawler changes which also affect the
"duplicate" marking of parameters.
If you:
$ grep "link" outputdir/pivots.txt
Can you give me the output ? Can you also just mail me the "debug.log" file ?
(e.g. "make debug; ./skipifish [...options..] 2> debug.log ")
Original comment by niels.he...@gmail.com
on 7 Aug 2012 at 11:46
With 2.08b out, I'm assuming this is fixed. Can you please verify ?
Cheers,
Niels
Original comment by niels.he...@gmail.com
on 1 Sep 2012 at 6:56
Hi,
Sorry for the delay response.
It is not fixed , i have attached debug.txt,pivot.txt please look into it.
thanks
Original comment by anuragno...@gmail.com
on 10 Sep 2012 at 7:08
Attachments:
Fixed -> New. I'll review the logs, thanks for attaching them!
Original comment by niels.he...@gmail.com
on 12 Sep 2012 at 3:21
Problem reproduced and it's due to us not scraping the response due to
similarities with previous requests.
Can you try the attached patch ? It should fix the issue.
Thanks!
Niels
Original comment by niels.he...@gmail.com
on 13 Sep 2012 at 9:32
Attachments:
It is now fixed , thanks a lot for your support.
One more question is it possible to just only crawl using skipfish (not doing
any vulnerability test), as my requirement is that i need to know the total URL
in a website with all the external links.
Original comment by anuragno...@gmail.com
on 13 Sep 2012 at 11:49
Using --no-checks , all injection tests should be disabled. Passive tests are
still performed on the responses so you'll still get some issues reported
(mostly low risk, low hanging fruit).
The resulting report directory will contain a pivots.txt file and you can start
future scans with:
./skipfish [..flag..] @path/to/pivots.txt
Unfortunately, I actually noticed today that there is a small bug in this
functionality . In the test case you created for this bug description, the
final pivot.txt file will have links1.php=1 and not links1.php?id=2 (because in
memory, this is one "pivot" with two values for the parameter id). However, it
will contain the asdf.php and abc.php (or whatever they were called ;p).
I'll look into the above mentioned bug.
Hope this helps and again thanks for reporting!! (and your patience)
Niels
Original comment by niels.he...@gmail.com
on 13 Sep 2012 at 8:01
I think it is perfect , links1.php?id=2 is not a new link, hence not a new
pivot , it is now able to find the new links attached with different parameters
so working perfectly. As per my knowledge it is not a bug.
thanks.
Original comment by anuragno...@gmail.com
on 14 Sep 2012 at 5:26
This is also fixed in 2.10b which is now in SVN. If you have time, could you
please test this ?
1) svn checkout http://skipfish.googlecode.com/svn/trunk/ skipfish-read-only
2) cd skipfish-read-only ; make; ./skipfish [....]
Cheers!
Niels
Original comment by niels.he...@gmail.com
on 23 Nov 2012 at 8:51
Original issue reported on code.google.com by
anuragno...@gmail.com
on 19 Jul 2012 at 11:23