Regarding the HashMap, it would be much cheaper to use a Map decorator to wrap an existing instance of ParameterParser. Jakarta Commons Collections AbstractMapDecorator [1] would make this easy to implement.

[1] http://cvs.apache.org/viewcvs.cgi/jakarta-commons/collections/src/java/org/apache/commons/collections/decorators/AbstractMapDecorator.java?rev=1.3&content-type=text/vnd.viewcvs-markup

Eric Pugh wrote:
I just finished the commit.  Do you want to take the ball on trying to get
the hashmap out of the parameter parser?

I don't actually use intake, but it seems like many things rely on intake,
so I had to get it over to fulcrum before implementing it into turbine.  For
instance, the converting localization to fulcrum doesn't work without
converting to intake in fulcrum as well!

Eric


-----Original Message-----
From: Colin Chalmers [mailto:[EMAIL PROTECTED]
Sent: Friday, September 26, 2003 12:50 PM
To: Turbine Developers List
Subject: Re: cvs commit: jakarta-turbine-2/xdocs changes.xml


I have a first cut of Intake running as a Fulcrum Component.. However, I


don't have your changes yet. I would like to commit it

today for review.


Colin, since you seem to understand Intake a lot, would you

review the


changes I have made and I think that your refactoring for

2.4 should be made


in the fulcrum component, and the intake in 2.3 will be

deprecated. If you


are making things easier, then that will provide a clear incentive to
switch.


Cool, but I have to admit that the IntakeService still holds some grey areas for me.


The intake service in fulcrum leverages the localization service from
fulcrum as well!

The only difficulty I have is on the parameterparser and

valueparser. They


require a reference on Turbine 2.3.  Since I want to use the intake
component in a standalone use case, as well as using it with

Scarab which


uses Turbine 3.0.

So, what would be requried to ditch the

paremeterparser/valueparser? While


we could move them to fulcrum, in many ways they seem to be very http
request specific, and ought to not be used in Fulcrum.


Yeah as far as understand it Fulcrum shouldn't have any dependancies on HTTP stuff right? I think Parameterparser just contains a hashMap of the data we need, if we could extract the HashMap from Parameterparser and pass that to the Fulcrum version (of Intake) we could extract the required data from there?

Colin


Comments?

Eric




-----Original Message-----
From: Scott Eade [mailto:[EMAIL PROTECTED]
Sent: Friday, September 26, 2003 3:20 AM
To: Turbine Developers List
Subject: Re: cvs commit: jakarta-turbine-2/xdocs changes.xml


Colin Chalmers wrote:





I want to refactor some intake stuff to make it simplier to
understand, this will be going into the 2.4 release. Will



probably be




running a few things by you for feedback if you don't mind.



You seem




to be using the Intake service quite intensly?



Fine, but if you want me to try anything I will most likely

only have


applications running on 2.3.1-dev, not 2.4-dev.
If you search way back in the mail archives you can find
reference to an
extension to Intake that provided client side validation using
JavaScript.  JavaScript would be no substitute for validation on the
server, but it could improve the user experience.  It could be
interesting to integrate this if you have some time on your hands.




Let me know if 2.3.1 Intake is now performing as expected.



AFAIK the only outstanding problem relates to the use format of the values returned by DateString. I am avoiding this by not using DateString.

Thanks for your patch - it saved me digging deeper into

Intake than I


already have.



--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]



Reply via email to