commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Rahul Akolkar (JIRA)" <j...@apache.org>
Subject [jira] Commented: (SCXML-27) Improve possibility to extending (subclassing) the default semantics
Date Tue, 03 Oct 2006 17:35:20 GMT
    [ http://issues.apache.org/jira/browse/SCXML-27?page=comments#action_12439571 ] 
            
Rahul Akolkar commented on SCXML-27:
------------------------------------

Let me state my assertions behind resolving this as won't fix:

a) The class in question has non-trivial functionality
b) We tried to make it pluggable, exactly for these kind of usecases
c) Custom semantics is considered advanced API for the library
d) The existing method signatures and call stack cannot be compromised
e) Because of (a) and (d), some (or many) cases might require subclasses to indulge in copious
copy-paste to find the correct insertion points
f) I personally expect advanced users to be somewhat more tolerant than casual users
g) Making a particular change that makes sense for any one usecase does not solve this issue
eternally

To reopen this ticket, therefore, we need a complete proposal stating how:

a) Subclassing convenience has improved for many (if not all) usecases (preferred way to illustrate
is test cases)
b) All existing API contracts are preserved
c) The changes (additions, really) constitute the minimal essential set

This is not to say your suggestion may not have merit for your usecase, it does, but these
are the (more important) broader constraints.

> Improve possibility to extending (subclassing) the default semantics
> --------------------------------------------------------------------
>
>                 Key: SCXML-27
>                 URL: http://issues.apache.org/jira/browse/SCXML-27
>             Project: Commons SCXML
>          Issue Type: Improvement
>    Affects Versions: 0.5
>            Reporter: Hallvard Trætteberg
>
> I've extended the default SCXML semantics (subclassed SCXMLSemanticsImpl) with a more
flexible event matching method. However, this required copying a lot of code from SCXMLSemanticsImpl,
just to replace the event matching method. The problem was that the eventMatch method takes
String and Set arguments, with no way to get at the Transition who's event is to be tested.
I copied the relevant code and replaced eventMatch(String,Set) with eventMatch(Transition,Set),
to make it more flexible. I ask that this change is made to SCXMLSemanticsImpl, to better
support overriding.
> The flexibility I've added is to use regular expressions for matching, and since this
happens often, I wanted to cache the compiled pattern. The natural place to do this is the
Transition, so I subclassed the Transition class, included a field for holding the compiled
pattern and made sure to use this subclass whenever a regexp was needed. Then in the modified
eventMatch code, I check whether the Transition is of this subclass and used the cache to
improve performance.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       

---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org


Mime
View raw message