Bugzilla – Full Text Bug Listing
| Summary: | Exception raised when activating/deactivating editors: "No file system is defined for scheme: pathmap" | ||
|---|---|---|---|
| Product: | [Modeling] Papyrus | Reporter: | Alain Le Guennec <alain.leguennec> |
| Component: | Core | Assignee: | Cedric Dumoulin <cedric.dumoulin> |
| Status: | RESOLVED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | CC: | eclipse-bugzilla, hess, mathieu.velten |
| Version: | 0.8.0 | ||
| Target Milestone: | --- | ||
| Hardware: | PC | ||
| OS: | Windows XP | ||
| Whiteboard: | |||
|
Description
Alain Le Guennec
There is two related bugs in Eclipse: https://bugs.eclipse.org/bugs/show_bug.cgi?id=310446 https://bugs.eclipse.org/bugs/show_bug.cgi?id=324788 In Papyrus, the exception is raised because when a resource is modified, the method org.eclipse.papyrus.core.adaptor.gmf.GmfMultiDiagramDocumentProvider.computeModificationStamp is called. This in turn calls the org.eclipse.core.internal.filesystem.InternalFileSystemCore.getFileSystem() method, which fails to find PATHMAP. This bug also happen in the following condition: -Create a model -Create 2 class diagrams -Import a profile (Ecore for example) The exception is raised, because a pathmap variable is added to the model, and the model is modified. A possible solution is to change the computeModificationStamp() method in order to catch faulty resources (with faulty URIs) before the call to getFileSystem(). The solution remove the exceptions, but I am not sure that it does'nt introduce other bugs. We should understand why we try to get a URI with a scheme 'pathmap' or 'platform', and why this fail. Do we forgot to register something ? I am pretty sure that the problem does not come from papyrus since pathmap and platform URI are valid EMF URIs. I am a little worried that this change breaks something. Why not put WorkspaceSynchronizer.getFile in a try-catch, since file variable is checked against null right after, it should fix the problem with less regression possibilities. more infos here : https://bugs.eclipse.org/bugs/show_bug.cgi?id=333690 looks like we should just ignore the error log entry until it's fixed upstream. I have choose to prevent the exception rather than catching it for performance reasons: it looks like the call to getFile() is time consuming (to be confirmed) when the file is not found. I was reluctant to apply this patch, because it could hide others problems in the future. We could change the code in order to let it report the first exception, and skip the remaining ones. I think this bug is a duplicate of bug 351813. I tried the patched version from Mariot Chauvin which solved the platform/pathmap error for me. This fix is in the indigo repository for SR1 (Source : https://bugs.eclipse.org/bugs/show_bug.cgi?id=351813#c6) Cannot reproduce anymore.
> This fix is in the indigo repository for SR1
I close this task
|