| RE: [stp-dev] STP project on Google Summer of Code 2009 |
WS-Policy instance(s)
will be extracted and saved in configured destination. Consequentially user will
like to validate and edit saved policy instance with Policy Editor. What happens
if policy has non-standard assertions (not WS-compliant) and Policy Editor has
not appropriate XML-schema in catalog?
Will framework provide only simple text based editor; warn user; ask for
appropriate schema; try to generate schema using reverse engineering
approach?
I do not completely understand how handler chain process meta information regarding relationships between different assertions: it seems that every handler “know” only about own assertion and extract it. But in this case, how the whole policy instance structure with alternatives will be composed? How the Policy model will be built and serialized?
What is the advantage
to use Axiom as XML parser in the project?
I think it makes sense to adapt described way (JSP-Servlet-Web Service-Policy Bilder) to Eclipse standard. Pragmatic approach is using functionality of Eclipse import/export wizards for that. For async and event based scenarious (like extracting policy from SOAP message) Eclipse WTP monitor can be evaluated.
I am missing in design
kind of abstract interface for all
sources and destinations. It makes sense to provide easily extensible
framework potentially supporting a new implementations for
sources/destinations(FileSystem, DB, Queue,
etc) and formats (even not XML).