Hi Ruchith,

So this doesn't mean that there are two trees - one for AXIOM (the
whole cached tree), and one for DOM ? That is, if i request a specific
element as Document , only it will be converted as DOM, and if i
request a document as an OMElement, it will be presented as such (all
of this using one object structure) ? Thanks very much.

Regards,
Angel

On 3/27/07, Ruchith Fernando <[EMAIL PROTECTED]> wrote:
When the requests hit the security handlers we *convert* the LLOM
axiom structure to a DOOM axiom structure. And then we set the DOOM
envelope as the envelope of the message context.

At this point there is only *one* object structure and this object
structure supports both AXIOM and DOM interface. Hence we can perform
security operations on this structure using wss4j and xml-security
libraries. And the rest of axis2 can access the *same* object
structure via the AXIOM interfaces :-)

Thanks,
Ruchith

On 3/27/07, Davanum Srinivas <[EMAIL PROTECTED]> wrote:
> Ruchith,
>
> Please explain how we *don't* create 2 separate om trees (one for just
> om and one for DOOM)...or do we? :)
>
> -- dims
>
> On 3/27/07, Ruchith Fernando <[EMAIL PROTECTED]> wrote:
> > Yes .. right now Rampart simply parses the whole DOM tree.
> >
> > IMHO 99% of the time the SOAP Body will be signed and in those
> > instances we will have to anyway read parse the complete envelope in
> > verifying the signature. How do you suggest we can improve performance
> > in such cases?
> >
> > Thanks,
> > Ruchith
> >
> > On 3/27/07, Angel Todorov <[EMAIL PROTECTED]> wrote:
> > > Hi Ruchith,
> > >
> > > I have one concern regarding DOM <-> AXIOM conversion with respect to 
Rampart.
> > >
> > > If i want to read a signature from the SOAP Header (suppose the
> > > signature is generated from some client already and the SOAP request -
> > > sent to the server) - does rampart parse the whole message into DOM
> > > format into memory, or only the header parts that contain the
> > > signature are parsed into a DOM tree ? I am asking this because of the
> > > performance implications this may have.
> > >
> > > Thanks very much.
> > >
> > > Regards,
> > > Angel
> > >
> >
> >
> > --
> > www.ruchith.org
> > www.wso2.org
> >
>
>
> --
> Davanum Srinivas :: http://wso2.org/ :: Oxygen for Web Services Developers
>


--
www.ruchith.org
www.wso2.org

Reply via email to