Cliff, Absolutely, that would be wonderful. I often tout that a benefit of ColdSpring is that it encourages you to write more testable code - and this would just make testing that code easier.
thanks, Dave On 7/2/06, Cliff Meyers <[EMAIL PROTECTED]> wrote:
One thing I mentioned to Chris was the idea of writing some abstract TestCase classes that autowire beans (by name, IIRC) into a test case based on what's defined in the XML files. This is fairly useful for writing integration tests without having to set up the bean factory yourself. They would be modeled closely after similar classes in the Java org.springframework.test package. If you guys think this would be useful, I'd be happy to put them together. -Cliff On 7/2/06, Chris Scott <[EMAIL PROTECTED]> wrote: > Wow, Dave, productive ride home! I'll have to give the > JavaSpringBeanFactoryAdapter a try right away. I spent my ride home > getting a $150 speeding ticket, being completely pissed, then went > right to sleep when I got home. I did however _think_ a lot about the > RegexPointcutAdvisor. I will build and commit soon... > > Chris > > Oh yeah, I thought it was a great conference as well. I was > particularly happy with the fact that the AOP BoF broke down into a > full on hacking session, with many laptops coding away, and some > serious progress made into cfcUnit tests being run by Ant in Eclipse. > Sweet! > > Chris Scott > [EMAIL PROTECTED] > http://cdscott.blogspot.com/ > http://www.coldspringframework.org/ > > > > > > > > > > > On Jul 2, 2006, at 10:02 AM, Dave Ross wrote: > > > So I think ColdSpring made a very good showing at the conference - > > even if just considering the number of excellent questions we were > > asked in the hallways, bar, etc. AOP was in full-force, with a lot of > > folks bouncing great ideas for usage of Chris and I. Not one "so all > > it does is logging?" was heard. > > > > So with that said, I thought I'd update this list on some of the cool > > new features we've discussed (and even implemented). > > > > 1) During one of Joe Rinehart's pre-conference talks, it was asked > > whether a regular java Spring BeanFactory could bet set as a parent of > > a ColdSpring BeanFactory. We actually didn't know for sure whether it > > would work - and I played with this the whole way home. Turned out we > > needed to write a little adapter class (you wrap your Java Spring > > BeanFactory in an instance of > > coldspring.beans.util.JavaSpringBeanFactoryAdapter and then pass that > > to a setParent() method on a regular ColdSpring BeanFactory). For > > reference: http://code.coldspringframework.org/browse/CSP-63 > > > > 2) AutoProxying - Those of you who saw and learned how ColdSpring's > > AOP framework is used may have drawn the conclusion that there's a lot > > of verbosity to the configuration of a single proxy (applying AOP to a > > single component takes a rather large definition in the xml). As you > > can imagine, this could become a serious pain if you wanted to apply > > advice across your entire codebase. What autoproxying does is allow > > you to configure an advisor that will not only match method > > signatures, but also the names/types of components, so that you can > > configure a single proxy to apply advice across an entire beanFactory. > > > > 3) Expanded pointcut model - Chris is also working on expanding the > > way that you can configure the mappedNames property of an advisor (the > > string pattern you use to map an advisor to methods being called). > > Right now you just are concerned with the name of the method being > > called - the expanded syntax would also allow you to match arguments - > > meaning if you only wanted to apply advice when a numeric value > > between 100 and 150 was passed to a certain method, you could express > > that via the expanded pointcut syntax. > > > > 4) Whatever you think is missing from ColdSpring 1.0 > > > > Have a great holiday weekend! > > > > thanks, > > > > Dave > > > > > > >
