I am writing a plugin for eclipse. When calling TextSearchQueryProvider sqProvider = TextSearchQueryProvider.getPreferred();, I get a NoClassDefFoundError.
The funny thing is, I only get this on the exported jar-plugin, not while debugging the plugin. I figured, it might be related to exporting the org.eclipse.search-plugin, but that plugin is so basic, that eclipse doesn't run without it anyway. So I guess the plugin should be there.
I am running eclipse photon (4.8.0).
Some more clarifications:
I have specified org.eclipse.search as dependency in MANIFEST.MF:
Require-Bundle: javax.inject,
org.eclipse.search;bundle-version="3.0.0";visibility:=reexport,
....
I have imported org.eclipse.search.ui.text.TextSearchQueryProvider
Do I need to do anything else, that I am not aware of?
Addition:
The plug-in related views did not show any obvious problems.
Specifically, the 'org.eclipse.search'-dependency is being satisfied by the version '3.11.200.v20180503-1856', which to me implies, that the plugin has been successfully linked?
The problem vanished, after I exported the plug-in with another version postfix.
I had originally called the postfix "beta". After giving it a new postfix with date,
the dependency tree seems to work correctly. It might be that the original package was bad and misconfigured in the MANIFEST, and only after reexporting with a different name, the pacakge dependencies were reevaluated corrrectly.
Related
Introduction
I have spend a lot of time to fix this bug
In our application we have a lot of generated code by cxf and jaxb which produces tons of warnings. We use the maven-build-helper plugin to add this code to our projects automatically.
By adding this enhancement, eclipse JDT enabled the possibility to set the javac -nowarn flag for specific source folders. Unluckily, by updating the maven project the flag gets lost. There are a few threads on SO where users got bugged by bug.
What I have tried so far
So i came up with a clever solution, javac has a nowarn flag. I set it to my maven-compiler-plugin and specified the directory. My maven build was fine, but my eclipse build wasn't. My research told me, that eclipse jdt does only use the maven-compiler-plugin source and target version. So my next step was to try to configure the EclipseCompiler, but this is not possible, because there is no possibility to add custom compiler Arguments in eclipse JDT.
Next Step. Inside of the .classpath file, eclipse JDT adds an ignore_optional_problems attribute for each ignored path. By updating the maven project inside of eclipse, this entry gets lost. So i started to write a maven "ignore-source-folder" plugin which should add the missing attribute. To run the plugin each time eclipse starts a build, i also created a m2e connector to refresh the .classpath file and everything should be fine.
By testing my plugin with my connector i realized, it works, but only 70% of the repetitions.
What happened?
Every time eclipse m2e/Jdt starts a new build, all classpath entries will be removed and populated again. When my maven plugin gets triggered by eclipse, a race condition starts.
So I started to analyze the code of jdt and m2e jdt. The ignore_optional_problems flag gets only once set manually inside of the patch which was provided to JDT and isn't stored somewhere else. By triggering a new build via m2e-jdt this information gets lost.
How to fix this problem
To fix this problem, some element has to be added or extended in jdt which contains all ignored folders. If a new build gets triggered and the classpath file gets newly generated by m2e-jdt this element should be checked for ignored paths. ClasspathEntryDescriptor seems to be a good place for it.
My problem
I checked out jdt and m2e jdt, but i have tons of errors inside of my IDE and i have no idea how to start. And even if I fix the code, I have no idea how to build and test it. I think my effort will be to high and we talk about 20 or 30 simple lines of code.
I am afraid if I add my results to the filed bug at eclipse, no one will care about it.
So, is there any developer able and willing to help me for this tiny job?
You need to "fix" the M2E's JDT project configurator. Probably somewhere in the m2e-core project. So, your change should obtain some kind of configuration flag from the project pom and create corresponding classpath entries for JDT.
I have a plugin which at present doesn't have any extension points (neither does it extend any other plugin). To resolve a certain issue I have to create an extension-point for this plugin and an appropriate extension somewhere else.
When doing this eclipse forced me to set singleton:=true in the Manifest file. I know the reason why this parameter has to be set, but I wanted to know if there are any implications on the functioning of the plugin within the product by introducing this parameter.
Thank you.
"singleton:=true" means that bundle only have one version could exist in the OSGi runtime.
Eclipse automatically adds the flag due to you provide an extension point in your bundle. If there are two version of your bundles providing the same extension point(have the same identity, might different attributes or elements) would make things mess.
From the Book Eclipse plugin Development by Dr Alex Blewitt:
The clause
singleton:=true
is an OSGi directive, which means that only one version of this plug-in can be installed in Eclipse at one time.
For plug-ins which add dependencies to the UI, there is a restriction that they must be singletons. (This constraint is one of the main reasons why installing a new plug-requires the IDE to restart.)
I am working on an RCP project based on eclipse. It has been working fine but recently I thought I'd upgrade it to use a new eclipse version (3.2 -> 3.5).
After a bit of trouble, it was running on the new platform. Then I did something. Don't know what. The end result is that I'm now getting a classpath error when one of my plugins (A) tries to access a class in one of the dependent plugins (B) (also one of mine).
As far as I can see, Plugin A has Plugin B in its' dependency list and the compiler shows no errors. To test, I created a new Plugin C with one class and accessed the class from Plugin A. That works fine.
Does anyone have any hints for troubleshooting such issues? A checklist of settings to check? I've been struggling with this for hours and getting nowhere! Particularly frustrating as it was working until I changed something!
Thanks
Update
I should also say that the project is quite old and uses the plugin.xml and not the manifest. Could this be a problem? Is there a way to upgrade?
Some ideas:
Clean-and-build every project (maybe after a restart of Eclipse)
Check, whether the required packages are also exported
Check the Error log, maybe there is some unloadable plug-in, etc.
Check the Run configuration, whether every required plug-in is present there
If the problem does not happen in a Runtime workbench, but in an Exported RCP app, then check whether you compile the project with the same settings during export than in Eclipse
I hope, something helps about these.
You can try the following in your run configuration:
In the Main tab check Clear (workspace) under Workspace Data
In the Configuration tab check clear the configuration area before launching under Configuration Area
This helps me every time I encounter some strange ClassNotFound errors.
If I understand you correctly you have no compiler errors but when you run you get class not found exceptions? You could try PDE Tools > Update Classpath from the project's context menu to see if you declared any dependencies outside of the MANIFEST.MF file, which would result in compiling code that fails at runtime.
I developed for RCP quite a long ago, but if I'm not wrong, since 3.4 or 3.5 you have to declare in the plugin configration which packages/classes are exported for dependent plugins and which are not.
This is unlike the old convention of using 'internal' in the package to mark non-exported packages.
Since you have plugin.xml from very old version of eclipse, it might be the problem, as exported resources from one plugin to another were not enforced in Eclipse RCP 3.2.
One more complement, hope it helps.
Open plugin.xml, check "build" tab, see how "binary build" are configured. It affects which files will compiled and exported.
See binary build here : http://help.eclipse.org/indigo/index.jsp?topic=%2Forg.eclipse.pde.doc.user%2Fguide%2Ftools%2Feditors%2Fmanifest_editor%2Fbuild.htm
example:
For logging, my code uses log4j. but other jars my code is dependent upon, uses slf4j instead. So both jars must be in the build path. Unfortunately, its possible for my code to directly use (depend on) slf4j now, either by context-assist, or some other developers changes. I would like any use of slf4j to show up as an error, but my application (and tests) will still need it in the classpath when running.
explanation:
I'd like to find out if this is possible in eclipse. This scenario happens often for me. I'll have a large project, that uses alot of 3rd party libraries. And of course those 3rd party jars have their own dependencies as well. So I have to include all dependencies in the classpath ("build path" in eclipse) for the application and its tests to compile and run (from within eclipse).
But I don't want my code to use all of those jars, just the few direct dependencies I've decided upon myself. So if my code accidentally uses a dependency of a dependency, I want it to show up as a compilation error. Ideally, as class not found, but any error would do.
I know I can manually configure the classpath when running outside of eclipse, and even within eclipse I can modify the classpath for a specific class I'm running (in the run configurations), but thats not manageable if you run alot of individual test cases, or have alot of main() classes.
It sounds like your project has enough dependency relationships that you might consider structuring it with OSGi bundles (plug-ins). Each bundle gets its own classloader and gets to specify what bundles (and optionally what version ranges, etc.) it depends on, what packages it exports, whether it re-exports stuff from its dependencies, etc.
Eclipse itself is structured out of Eclipse plug-ins and fragments, which are just OSGi bundles with an optional tiny bit of additional Eclipse wiring (plugin.xml, which is used to declare Eclipse "extension points" and "extensions") attached. Eclipse thus has fairly good tooling for creating and managing bundles built-in (via the Plug-in Development Environment). Much of what you find out there may lead you to conflate "OSGi bundle" with "plug-in that extends the Eclipse IDE", but the two concepts are quite separable.
The Eclipse tooling does distinguish rather clearly (and sometimes annoyingly, but in the "helpful medicine" way) between the bundles in your build environment vs. the bundles that a particular run configuration includes.
After a few years of living in OSGi land, the default Java "flat classpath" feels weird and even kind of broken to me, largely because (as you've experienced) it throws all JARs into one giant arena and hopes they can sort of work things out. The OSGi environment gives me a lot more control over dependency relationships, and as a "side effect" also naturally demands clarification of those relationships. Between these clear declarations and the tooling's enforcement of them, the project's structure is more obvious to everyone on the team.
if my code accidentally uses a dependency of a dependency, I want it to show up as a compilation error. Ideally, as class not found, but any error would do.
Put your code in one plug-in, your direct dependencies in other plug-ins, their dependencies in other plug-ins, etc. and declare each plug-in's dependencies. Eclipse will immediately do exactly what you want. You won't be offered dependencies' dependencies' contents in autocompletes; you'll get red squiggles and build errors; etc.
Why not use access rules to keep your code clean?
It looks like it would better be managed with maven, integrated in eclipse with m2eclipse.
That way, you can only execute part of the maven build lifecycle, and you can manage separate set of dependencies per build steps.
In my experience it helps to be more resrictive, I made the team filling out (paper) forms why this jar is needed and what license...
and they did rather type in a few lines of code instead of drag along 20 jars to open a file using only one line of code, or another fancy 'feature'.
Using maven could help for a while, but when you first spot jars having names like nightly-build or snapshot, you will know you're in jar-hell.
conclusion: Choose dependencies well
Would using the slf4j-over-log4j jar be useful? That allows using slf4j with actual logging going to log4j.
I have a serious probleam with my Eclipse Plugin..
My plugin depends on another two plugins. All of theese plugins (including my plugin) use Mozilla Rhino engine - two of them use js.jar (version 1.4). But my plugin uses new version, which is not released yet and is packed in MyRhino.jar.
While developing, everything goes fine - code completion offers me classes and methods from my special version of rhino, no errors etc.
But when I run my plugin, a different version of Rhino is used and I get runtime exceptions like ClassNotFoundException and so on.
How am I supposed to tell eclipse to use the same jar in runtime as in development time?
When I was developing plugins on NetBeans, the solution was simple - wrapping the jar in a separate module, set which packages to export and declare dependency on this new module..
I found simmilar solution in Eclipse, however it did not work for me - although I declared a dependency on the wrapper plugin, the packages and classes were not available even in development time.
I've spent hours with this probleam and not solved it, so and help is appreciated. Thanks everyone.
Make sure your dependencies are correct. If your code depends on new version of library, then you need to indicate that in your MANIFEST.MF. E.g. if you require bundle org.mozilla.rhino, specify minimum version you need:
Require-Bundle: org.mozilla.rhino;bundle-version="1.5.0"
Your MyRhino.jar will need to specify correct version (e.g. 1.5.0), even though it is not released yet:
Bundle-Version: 1.5.0.qualifier
(If MyRhino.jar isn't converted to plugin yet, you need to do that manually: you can simply craete new plugin project using existing JAR in Eclipse)
Alternativaly, you can wrap MyRhino.jar into your plugin, and remove dependencies on other plugins. To do this, use Bundle-ClassPath (see http://www.aqute.biz/Blog/2007-02-19 for details). If you put your MyRhino.jar directly into your plugin, then Bundle-ClassPath should be:
Bundle-ClassPath: .,MyRhino.jar
Finally I've solved it..
It's simple to create the wrapper plugin form existing jar by the new project wizard in eclipse.. But the catch is that the wizard does not include the jar in bundle classpath (as I expected). Prety confusing (at least for me) that the wizard does only half of the job for you :/
After inserting the jar in the wrapper plugin bundle classpath, everytning works.
Thanks Peter for your answear anyway :)