July 2008 - jboss-dev-forums - Jboss List Archives

[Design the new POJO MicroContainer] - Re: JBoss with spaces in the directory name

by alesj

I still think the path should be decoded asap in the whole execution stack. Since there might be a few similar cache points to this one I found in FileHandler: | public VirtualFileHandler createChildHandler(String name) throws IOException | { | File parentFile = getFile(); | File child = new File(parentFile, name); | VirtualFileHandler handler = childCache.get(name); | // if a child has already been created use that | // if the child has been modified on disk then create a new handler | if (handler != null && handler.hasBeenModified()) | { | handler = null; | } | if (handler == null) | { | FileSystemContext context = getVFSContext(); | handler = context.createVirtualFileHandler(this, child); | if (handler != null) | childCache.put(name, handler); | } | return handler; | } | e.g. if the name once comes with %20 and the 2nd time w/o, then we'll have two entries for the same file. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166419#4166419 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4166419

17 years, 8 months

1
0
0 / 0

[Design of AOP on JBoss (Aspects/JBoss)] - Re: AOPConstructorJoinpoint and methodHasSubInstanceMetaData

by kabir.khan＠jboss.com

OK, I've reverted the changes and deleted the issue View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166418#4166418 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4166418

17 years, 8 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: KernelRegistry - unnecessary Exception construction

by adrian＠jboss.org

"adrian(a)jboss.org" wrote : "alesj" wrote : Can you create JIRA's for both items: | | | | https://jira.jboss.org/jira/browse/JBMICROCONT-318 | Ales, if you keep assigning every task to yourself, especially the unscheduled ones or things you're not currently working on (or plan to work on) then no other (potential) contributor is ever going to volunteer to do the work. Scott causes the same problem in the appserver where people think he is working on all sorts of issues that he's never even looked at in years. ;-) View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166417#4166417 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4166417

17 years, 8 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: KernelRegistry - unnecessary Exception construction

by adrian＠jboss.org

"alesj" wrote : Can you create JIRA's for both items: | https://jira.jboss.org/jira/browse/JBMICROCONT-318 https://jira.jboss.org/jira/browse/JBMICROCONT-319 anonymous wrote : | - adding findEntry (we're getting inconsistent with usage of get/find :-), since in VFS it's just the opposite) | They both have the same brain dead protocol, creating and throwing an error for not found instead of the more efficient returning null. :-) View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166415#4166415 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4166415

17 years, 8 months

1
0
0 / 0

[Design of JBoss Profiler] - JPA getReference() does not work with 2.0beta1

by valldeby

We're developing a app with JBoss, MySQL and JPA/Hibernate and decided to profile our project with JBoss Profiler. But no luck! When debugging, something always goes wrong on a this line: Table table = em.getReference(Table.class, tableId); Where em is defined as EntityManager em = emf.createEntityManager(); No exception is thrown. And the debugger ends up in the finally-block of our try/catch. We're using: JBoss 4.2.2GA on Windows for development (will be deployed on Linux). Hibernate 3.2.4 JDK 1.6.0_03 I've also noticed that both our project and the profiler uses javassist.jar (as instructed in README.txt) but unfortunately different versions. /Martin PS No small example at the moment. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166413#4166413 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4166413

17 years, 8 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: VFS Performance

by adrian＠jboss.org

"adrian(a)jboss.org" wrote : "adrian(a)jboss.org" wrote : | | Reading byte code from zip files | | | | | java.util.zip.Inflater.inflateBytes(Inflater.java:Unknown line) | | | java.util.zip.Inflater.inflate(Inflater.java:215) | | | java.util.zip.InflaterInputStream.read(InflaterInputStream.java:128) | | | java.io.FilterInputStream.read(FilterInputStream.java:90) | | | org.jboss.virtual.plugins.context.zip.ZipEntryInputStream.read(ZipEntryInputStream.java:112) | | | org.jboss.classloader.plugins.ClassLoaderUtils.loadByteCode(ClassLoaderUtils.java:150) | | | org.jboss.classloader.spi.base.BaseClassLoader$2.run(BaseClassLoader.java:475) | | | org.jboss.classloader.spi.base.BaseClassLoader$2.run(BaseClassLoader.java:464) | | | java.security.AccessController.doPrivileged(AccessController.java:Unknown line) | | | org.jboss.classloader.spi.base.BaseClassLoader.loadClassLocally(BaseClassLoader.java:462) | | | | | | | This one might improve if we wrap the VFS stream in a BufferedInputStream | in ClassLoaderUtils? | Although ClassLoaderUtils is already using a 1K buffer, so its probably not relevant introducing more buffering. However BufferedInputStream uses an 8K buffer by default, so maybe we can increase the buffer size to see if it helps. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166412#4166412 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4166412

17 years, 8 months

1
0
0 / 0

[Design of AOP on JBoss (Aspects/JBoss)] - Re: AOPConstructorJoinpoint and methodHasSubInstanceMetaData

by adrian＠jboss.org

"adrian(a)jboss.org" wrote : | I've started a thread on this in the MC forum | http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166401#4166401 Hold off on making changes or JIRA issues until we know what we are going to do, otherwise it is likely you'll have to unpick it all anyway. :-) View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166408#4166408 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4166408

17 years, 8 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: Passing MetaData to AOPConstructorJoinPoint

by adrian＠jboss.org

Previously we had a bizarre method on BeanInfo/ClassAdapter which passed an opaque object (which was the MetaData) but this was also very hacky. Maybe we just need an extra method on the JoinpointFactory something like: | /** | * Get a constructor join point | * | * @param constructorInfo the constructor info | * @param metaData an opaque object | * @return the constructor join point | * @throws JoinpointException when no such constructor | */ | ConstructorJoinpoint getConstructorJoinpoint(ConstructorInfo constructorInfo, Object metaData) throws JoinpointException; | which InstantiateAction can use to pass the MetaData instead of the MetaDataStack. This at least hides the ugliness from the main api and doesn't interfere with other uses fo ConstructorJoinpoints that use the original method. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166406#4166406 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4166406

17 years, 8 months

1
0
0 / 0

[Design of AOP on JBoss (Aspects/JBoss)] - Re: AOPConstructorJoinpoint and methodHasSubInstanceMetaData

by adrian＠jboss.org

So the TomcatDeployer is a similar problem, but this time it is JBossXB's usage of BeanInfo.newInstance() which also dispatches a construct joinpoint. That's one is harder to fix since jboss-reflect can't depend on jboss-mdr (the dependency is the reverse) to mask the metadata stack. Maybe we need to think about doing this in a different way so it is less hacky to differentiate constructions that need to pass a MetaData object (and should explicitly do so) versus the ad hoc usage of BeanInfo/ConstructorJoinPoints which don't and shouldn't therefore be creating AOP proxies. I've started a thread on this in the MC forum http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166401#4166401 View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166402#4166402 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4166402

17 years, 8 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Passing MetaData to AOPConstructorJoinPoint

by adrian＠jboss.org

Kabir and I have come across an issue relating to how the MC and AOP negotiate the MetaData to use. http://www.jboss.com/index.html?module=bb&op=viewtopic&t=139625 This solutions are getting pretty hacky, so I think we need to rethink how this api works. Basically being able to differentiate constructor calls that need to create AOP proxies and use Metadata vs other uses of ConstructorJoinPoints. Suggestions? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4166401#4166401 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4166401

17 years, 8 months

1
0
0 / 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

jboss-dev-forums July 2008