Hey Mark, thank you! From what you're saying, this would jibe with
some errors I have seen/heard of. I'll make this change tonight. And
yeah, you're probably right about having a getter there for
isFactoryBean. I'll commit tonight. So hey, with the changes you made
on your version, are your apps running well? Thanks, Chris
Chris Scott
[EMAIL PROTECTED]
http://cdscott.blogspot.com/
http://www.coldspringframework.org/
On Apr 27, 2006, at 3:40 AM, Mark Mandel wrote:
I actually spent some time going through the code, and I can see
the error
Looking at: DefaultXMLBeanFactory,
Line 479 - we loop through all the functions, and call the init() and
the setBeanFactory() functions on a constructBean call, regardless of
whether or not the Bean has come from a factory or not.
Seems like that block of code should be inside the above <cfif> block
(~477) - unless of course you want to also do the check for
setBeanFactory() on factory created beans (which you may wish to). In
which case, there should be a check for if the Bean has a Factory,
before init is called.
So line 482 should read:
<cfif md.functions[functionIndex].name eq "init" AND
beanDef.getFactoryBean() eq "">
(which is how it now reads in my local file)
Side note - why not have a 'isFactoryBean()' method on your
BeanDefintion? - it would save you a whole lot of checking for
'getBeanFactory() eq ""'... or Len(getBeanFactory()) etc...
Hope that helps,
Mark
On 4/27/06, Mark Mandel <[EMAIL PROTECTED]> wrote:
Hey -
The latest version through me for a spin.
It looks like if you have a
<bean id="beanFOne" factory-bean="factory" factory-
method="getBeanOne"/>
when CS goes to get beanFOne, it calls init on the bean.
This tends to make things fall over when the init() takes a param
that
is required, or you don't want this to happen.
I can only assume this is a bug?
I've attached another test bed (in a zip this time!), so hopefully
you
should be able to see what I'm talking about.
Cheers,
Mark
--
E: [EMAIL PROTECTED]
W: www.compoundtheory.com
ICQ: 3094740
--
E: [EMAIL PROTECTED]
W: www.compoundtheory.com
ICQ: 3094740
