Closed GoogleCodeExporter closed 9 years ago
I will investigate your work and let you know what I come up with - we are
happy to
increase speed and memory efficiency. But, before we would integrate such a
patch we
have to threat model this approach and review the code to see if we're
confident it
provides the same level of security.
Original comment by arshan.d...@gmail.com
on 7 Jul 2008 at 2:10
Sounds good. I have basically taken the existing code and translated it from
DOM to SAX, so instead of
removing nodes from the tree, I am simply not copying events from the input
stream to the output stream.
If you have any questions, feel free to ask, I am also on the mailing list,
this might provide a better forum for
discussion coding questions.
I am attaching a complete integration of my code into AntiSamy: the DOMScanner
will use the SAX filter as
well and there is now SAXScanner that can be used directly by AntiSamy (but you
will not be able to access the
DOM result any longer(
Original comment by larstemp...@gmail.com
on 7 Jul 2008 at 2:21
Attachments:
If there is anything I can do to help you with the review, please let me know.
I would like to use AntiSamy for the
Apache Sling project, but depend on the SAX-implementation, because I want to
add additional SAX filters, for
microformat extraction, link count, spam protection, etc.
Original comment by larstemp...@gmail.com
on 22 Jul 2008 at 10:10
Original comment by arshan.d...@gmail.com
on 19 Nov 2008 at 7:15
Original comment by arshan.d...@gmail.com
on 3 Dec 2008 at 3:33
Has this been added to the trunk and/or a release yet?
Original comment by manosbat...@gmail.com
on 12 Feb 2009 at 4:26
lars, can you email me so we can work on this together? I would like to add
this by
version 1.4 (1.3 will be out this week hopefully). The patch needs some updates
and I
am in a good position to make this switch soon.
Original comment by arshan.d...@gmail.com
on 17 Mar 2009 at 2:34
Changing label to enhancement instead of defect as that more accurately
characterizes
this change.
Original comment by li.jaso...@gmail.com
on 17 Mar 2009 at 4:00
I am very interested in seeing this integrated. Performance and memory
efficiency is
extremely important. Is there any word on this / 1.4?
Original comment by jason.cl...@gmail.com
on 1 Feb 2010 at 7:30
I've analyzed this patch, and I am all for making the change to SAX, but I
would like
it to be slightly more organized. I'm not suggesting the current patch isn't
good, I
just don't know what an organized SAX implementation would look like.
I wouldn't expect this in 1.4, sadly.
Original comment by arshan.d...@gmail.com
on 8 Mar 2010 at 6:13
Lars, you're missing some code from this patch, most notably changes to
AntiSamy.java
(and possibly AntiSamyDOMScanner, I can't tell). I'm integrating this code next
week,
hook me up!
Original comment by arshan.d...@gmail.com
on 21 May 2010 at 7:54
After some work this was added to the baseline. It will be in the next version
(1.4.1) as an optional scanning method in opposition to DOM-based. Scans will
still
default to DOM, but an extra method signature will allow users to choose
SAX-based
instead.
Original comment by arshan.d...@gmail.com
on 1 Jun 2010 at 4:55
Original issue reported on code.google.com by
larstemp...@gmail.com
on 7 Jul 2008 at 11:25Attachments: