Re: Forthcoming FlexibleMapAccessor refactor

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: Forthcoming FlexibleMapAccessor refactor

Adam Heath-2
Adrian Crum wrote:

> I have refactored the OFBiz FlexibleMapAccessor so that it
> does some of the same magic as FlexibleStringExpander
> (compact objects, reusable null object, factory method,
> expression cache).
>
> The problem is, I have re-introduced a number of type-safety
> compiler warnings. I'm not a generics guru and I don't know
> how to fix them. In addition, code that calls the factory
> method generates type-safety warnings too.
>
> If it isn't too much to ask, could you take a look at it
> once I get it committed, and then do your generics magic on
> it? I'll just commit the FlexibleMapAccessor class first,
> and wait before converting any peripheral code over to the factory method.

This is an open-ended response to this community.  I watch my compile
output, and when I see messages saying to turn on -X:deprecation, in
components I've already converted, I just silently fix the new warnings
that occur.  I completely understand that generics take time to learn,
and am more than willing to convert for other people.

I may not fix warnings when they are first introduced.  I may not even
compile each new changeset as they are committed.  As is well known, we
all vary in the time we give to processing ofbiz, so if I don't add
generics to something you've done, don't fret, it'll happen eventually.

I will eventually also produce a best practices doc for doing this to
ofbiz.  I probably have enough experience with doing this for this
project, to write it now, but was sorta waiting until I got the entire
codebase converted.  I've been slowly working with the components, which
is taking a bit longer than expected.

As a bonus, the following url is quite excellent, in helping to answer
estoeric questions about java generics; it kinda requires a basic
understanding of them first, however.  It also comes in a 4.5M pdf.

http://www.angelikalanger.com/GenericsFAQ/JavaGenericsFAQ.html
Reply | Threaded
Open this post in threaded view
|

Re: Forthcoming FlexibleMapAccessor refactor

Adrian Crum
Adam,

Thanks for the reply! I will spend some time reading that material and
try to solve it myself.

-Adrian

Adam Heath wrote:

> Adrian Crum wrote:
>> I have refactored the OFBiz FlexibleMapAccessor so that it
>> does some of the same magic as FlexibleStringExpander
>> (compact objects, reusable null object, factory method,
>> expression cache).
>>
>> The problem is, I have re-introduced a number of type-safety
>> compiler warnings. I'm not a generics guru and I don't know
>> how to fix them. In addition, code that calls the factory
>> method generates type-safety warnings too.
>>
>> If it isn't too much to ask, could you take a look at it
>> once I get it committed, and then do your generics magic on
>> it? I'll just commit the FlexibleMapAccessor class first,
>> and wait before converting any peripheral code over to the factory method.
>
> This is an open-ended response to this community.  I watch my compile
> output, and when I see messages saying to turn on -X:deprecation, in
> components I've already converted, I just silently fix the new warnings
> that occur.  I completely understand that generics take time to learn,
> and am more than willing to convert for other people.
>
> I may not fix warnings when they are first introduced.  I may not even
> compile each new changeset as they are committed.  As is well known, we
> all vary in the time we give to processing ofbiz, so if I don't add
> generics to something you've done, don't fret, it'll happen eventually.
>
> I will eventually also produce a best practices doc for doing this to
> ofbiz.  I probably have enough experience with doing this for this
> project, to write it now, but was sorta waiting until I got the entire
> codebase converted.  I've been slowly working with the components, which
> is taking a bit longer than expected.
>
> As a bonus, the following url is quite excellent, in helping to answer
> estoeric questions about java generics; it kinda requires a basic
> understanding of them first, however.  It also comes in a 4.5M pdf.
>
> http://www.angelikalanger.com/GenericsFAQ/JavaGenericsFAQ.html
>
Reply | Threaded
Open this post in threaded view
|

Re: Forthcoming FlexibleMapAccessor refactor

Adam Heath-2
Adrian Crum wrote:
> Adam,
>
> Thanks for the reply! I will spend some time reading that material and
> try to solve it myself.

I also started doing stuff on a much smaller project, where it is easier
to keep more in the head at once.

Then, after a few months there, I started doing ofbiz wholesale, and
that is what made me get commit privs again.  I didn't tell anyone I was
doing it, and did a whole bunch first, which then just helped me to
learn how to do it even better.
Reply | Threaded
Open this post in threaded view
|

Re: Forthcoming FlexibleMapAccessor refactor

Adrian Crum-2
Adam,

Thanks for encouraging me to learn this stuff. I figured out most of it. I still had to suppress warnings on the factory method though - since I want to return an untyped instance but I need to create a typed instance to return.

-Adrian


--- On Thu, 12/11/08, Adam Heath <[hidden email]> wrote:

> From: Adam Heath <[hidden email]>
> Subject: Re: Forthcoming FlexibleMapAccessor refactor
> To: [hidden email]
> Date: Thursday, December 11, 2008, 9:16 AM
> Adrian Crum wrote:
> > Adam,
> >
> > Thanks for the reply! I will spend some time reading
> that material and
> > try to solve it myself.
>
> I also started doing stuff on a much smaller project, where
> it is easier
> to keep more in the head at once.
>
> Then, after a few months there, I started doing ofbiz
> wholesale, and
> that is what made me get commit privs again.  I didn't
> tell anyone I was
> doing it, and did a whole bunch first, which then just
> helped me to
> learn how to do it even better.