November 2007 - jboss-dev-forums - Jboss List Archives

[Design the new POJO MicroContainer] - Re: New behaviour of AbstractMetaDataContext.isEmpty()

by alesj

http://jira.jboss.com/jira/browse/JBMICROCONT-218 View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4105172#4105172 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4105172

18 years, 4 months

1
0
0 / 0

[QA of JBoss Portal] - Re: Problems with maven

by pgier

"bdaw" wrote : So with cargo most off those dependencies are not present in my plugin classpath. If I don't use test scope for dependencies everything is fine. | Any idea? Before I start to debug maven.... ;) | | So the results of mavenProject.getTestClasspathElements() is different depending on whether you're using the cargo plugin or not? That seems strange. I guess you could look at the cargo source to see if/where it is modifying the test classpath. If there's nothing you can do to change cargo's behavior maybe you could add the dependencies you need for your plugin as plugin dependencies instead of project test dependencies. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4105171#4105171 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4105171

18 years, 4 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: New behaviour of AbstractMetaDataContext.isEmpty()

by adrian＠jboss.org

"adrian(a)jboss.org" wrote : | But on a related issue, I don't like the fact that this logic is in the bridge. | The bridge should be just delegating the request to the retrieval. | i.e. a plain retrieval will just match its scope level while a context | will have to filter its local retrievals. It's more OO. i.e. There should be a | List<MetaDataRetrieval> getScopedRetrievals(ScopeLevel level) | on the retrieval instance and the two branches of the if statement moved into the relevant abstract implementation (context or retrieval). P.S. I'd also guess that the testcase for this functionality isn't very good. ;-) View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4105169#4105169 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4105169

18 years, 4 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: New behaviour of AbstractMetaDataContext.isEmpty()

by kabir.khan＠jboss.com

I think you could perhaps run each aop-mc-int profile as a separate hudson run, and make them dependent on the main MC run, or the main MC run triggers these separate runs. I am not sure exactly how that works, but Ryan Campbell did something like that for the aop tests http://www.jboss.com/index.html?module=bb&op=viewtopic&t=122864 View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4105168#4105168 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4105168

18 years, 4 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: New behaviour of AbstractMetaDataContext.isEmpty()

by pgier

If you want, I can turn them all on by default. Or just turn them all on in hudson. I'll have to check with Prabhat to see if we can configure hudson to display all the test results nicely since they aren't in the normal location. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4105166#4105166 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4105166

18 years, 4 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: New behaviour of AbstractMetaDataContext.isEmpty()

by adrian＠jboss.org

"kabir.khan(a)jboss.com" wrote : | The instanceMetaData.toString() method shows | | | org.jboss.metadata.spi.retrieval.MetaDataRetrievalToMetaDataBridge@1286c00{[JVM=THIS, CLASS=org.jboss.test.microcontainer.matrix.AnnotatedChild, INSTANCE=AnnotatedChild, WORK=1209895]} | | | That's clearly wrong. The invocation of: MetaData instanceMetaData = metaData.getScopeMetaData(CommonLevels.INSTANCE); should be giving you back a MetaData object that only contains the INSTANCE=AnnotatedChild. The bug is fairly obvious. The single scope metadata shouldn't have a parent! :-) | public MetaData getScopeMetaData(ScopeLevel level) | { | if (level == null) | throw new IllegalArgumentException("Null scope level"); | | if (retrieval instanceof MetaDataContext) | { | MetaDataContext context = (MetaDataContext)retrieval; | List<MetaDataRetrieval> matchingRetrievals = new ArrayList<MetaDataRetrieval>(); | List<MetaDataRetrieval> localRetrievals = context.getLocalRetrievals(); | for (MetaDataRetrieval localRetrieval : localRetrievals) | { | ScopeKey scopeKey = localRetrieval.getScope(); | | if (scopeKey.getScopeLevel(level) != null) | matchingRetrievals.add(localRetrieval); | } | if (matchingRetrievals.isEmpty() == false) | - return new MetaDataRetrievalToMetaDataBridge(new AbstractMetaDataContext(context, matchingRetrievals)); | + return new MetaDataRetrievalToMetaDataBridge(new AbstractMetaDataContext(null, matchingRetrievals)); | } | else if (retrieval.getScope().getScopeLevel(level) != null) | return new MetaDataRetrievalToMetaDataBridge(new AbstractMetaDataContext(retrieval)); | return null; | } | But on a related issue, I don't like the fact that this logic is in the bridge. The bridge should be just delegating the request to the retrieval. i.e. a plain retrieval will just match its scope level while a context will have to filter its local retrievals. It's more OO. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4105165#4105165 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4105165

18 years, 4 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: New behaviour of AbstractMetaDataContext.isEmpty()

by kabir.khan＠jboss.com

These tests are not being run automatically, I've asked before why on another thread. I think Paul mentioned that the reason for this is that "they take a long time". Anyway, looking at the code change that caused this they would have been broken for about 2 months now, so I think they should be enabled. I agree that they don't necesarily need to be run when doing a local "mvn install", but they should be enabled when being run on Hudson. The tests are not part of the normal profile, to run the aop-mc-int tests you have to | $cd aop-mc-int | $mvn surefire-report:report -Ptests-no-weave | $mvn surefire-report:report -Ptests-no-weave-secure | $mvn surefire-report:report -Pant-tests-weave | $mvn surefire-report:report -Pant-tests-weave-secure | View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4105158#4105158 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4105158

18 years, 4 months

1
0
0 / 0

[QA of JBoss Portal] - Re: Problems with maven

by pgier

"bdaw" wrote : | How could I check if the property is present and break the maven build with some nice help message? | | The maven enforcer plugin can check for a property and then fail if it's not there. http://maven.apache.org/plugins/maven-enforcer-plugin/rules/requireProper... View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4105157#4105157 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4105157

18 years, 4 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: New behaviour of AbstractMetaDataContext.isEmpty()

by alesj

I'll have a look at that, since I wrote that. ;-( I remember having a todo on tests, but forgot, since no exception occurred. Btw: why are these tests excluded from maven tests? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4105156#4105156 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4105156

18 years, 4 months

1
0
0 / 0

[Design of POJO Server] - Re: Metamodel and collections

by wolfc

Yup. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4105152#4105152 Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4105152

18 years, 4 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 November 2007