Raymond Feng wrote:
Hi, Frank.
Your proposal sounds promising, I cannot wait to move to SDO 3 :-).
For SDO 2.1, if it's not constrained by the spec, adding some metadata
to the generated interface/class would be helpful.
Thanks,
Raymond
----- Original Message ----- From: "Frank Budinsky (JIRA)"
<tuscany-dev@ws.apache.org>
To: <[EMAIL PROTECTED]>
Sent: Tuesday, February 13, 2007 4:29 PM
Subject: [jira] Commented: (TUSCANY-1110) Improve the performance of
TypeHelperImpl.getType(Class)
[
https://issues.apache.org/jira/browse/TUSCANY-1110?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12472943
]
Frank Budinsky commented on TUSCANY-1110:
-----------------------------------------
Yang, can you explain more about your idea?
I also have a couple of ideas for how to speed up
TypeHelper.getType(Class):
1) when we move a Java5 dependency, we should generate an annotation
in the interface, which we could use at runtime to find the Type. The
exact format of the annotation depends on an SDO 3 feature to support
Java metadata annotations, which the SDO collaboration is currently
considering.
2) before Java5, we could possibly do something like this:
step 1) determine the class name from the provided Class.
step 2) mangle the name to determine the impl class name (e.g.,
org.example.Foo -> org.example.impl.FooImpl)
step 3) create an instance and then call getType() - or better
yet, we could change the generator pattern to generate a static
method that returns the type - and then call it.
I'm sure there are also other possible ways to do this, but the
question is what's the priority for this? Does anyone know if the
current performance of this method is a concern that needs immediate
attention?
Improve the performance of TypeHelperImpl.getType(Class)
--------------------------------------------------------
Key: TUSCANY-1110
URL: https://issues.apache.org/jira/browse/TUSCANY-1110
Project: Tuscany
Issue Type: Improvement
Components: Java SDO Implementation
Affects Versions: Java-SDO-Mx
Reporter: Raymond Feng
Fix For: Java-SDO-Mx
In the SDO databinding for SCA, we need to introspect a java
class/interface to figure out the corresponding SDO type. Looking
into the code, there is a TODO comment:
//TODO more efficient implementation ... this is a really bad one!
Do you plan to provide a more efficient impl :-)?
Thanks,
Raymond
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
I'm not 100% sure of what's being proposed here but I think Jax-B has
something like this already. If I'm right maybe looking at that and as
best as can be done to align it might be something worthwhile ?
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
