Community
Participate
Working Groups
Since the migration to Bugzilla 4.0 on Eclipse.org I'm seeing incoming notifications in the task editor for the Resolution, Keywords and Status Whiteboard fields even though these don't appear to have changed (see screenshot).
Created attachment 209015 [details] tooltip
Created attachment 209016 [details] task editor
Steffen, I found the source of this problem. Starting with version 4.0 of Bugzilla we get empty tags for empty fields. In version 3.6 the tags are not in the output. Example: new <bug_status>CONFIRMED</bug_status> <resolution/> <bug_file_loc/> <status_whiteboard/> <keywords/> <priority>Lowest</priority> and old <bug_status>NEW</bug_status> <priority>Lowest</priority> as an result of this we get an list with an empty String and not an empty list. This is only an migration problem. When we change this we must implement an smarter version of TaskAttributDiff.hasChanges(). Thoughts?
Steffen, you can fix this by add the following lines in SaxMultiBugReportContentHandler.endElement(String uri, String localName, String qName) after the line String parsedText = characters.toString(); insert the next three lines if (parsedText.equals("")) { //$NON-NLS-1$ return; }
Created attachment 209178 [details] mylyn/context/zip
Thanks for investigating that! We should add a test case and back port this to 3.6.5.
Steffen, please use c8345e59018b2b81dd2fe990cbd90b5eddac3db7 for backport this.
Created attachment 209399 [details] mylyn/context/zip
*** Bug 368847 has been marked as a duplicate of this bug. ***
This caused a regression with bugs that have empty descriptions: org.eclipse.core.runtime.AssertionFailedException: null argument: at org.eclipse.core.runtime.Assert.isNotNull(Assert.java:85) at org.eclipse.core.runtime.Assert.isNotNull(Assert.java:73) at org.eclipse.mylyn.tasks.core.data.TaskAttribute.setValue(TaskAttribute.java:563) at org.eclipse.mylyn.internal.bugzilla.core.SaxMultiBugReportContentHandler.addDescription(SaxMultiBugReportContentHandler.java:682) at org.eclipse.mylyn.internal.bugzilla.core.SaxMultiBugReportContentHandler.addDescriptionAndComments(SaxMultiBugReportContentHandler.java:637) at org.eclipse.mylyn.internal.bugzilla.core.SaxMultiBugReportContentHandler.endElement(SaxMultiBugReportContentHandler.java:438) at com.sun.org.apache.xerces.internal.parsers.AbstractSAXParser.endElement(AbstractSAXParser.java:601) at com.sun.org.apache.xerces.internal.impl.dtd.XMLNSDTDValidator.endNamespaceScope(XMLNSDTDValidator.java:263) at com.sun.org.apache.xerces.internal.impl.dtd.XMLDTDValidator.handleEndElement(XMLDTDValidator.java:2028) at com.sun.org.apache.xerces.internal.impl.dtd.XMLDTDValidator.endElement(XMLDTDValidator.java:901) at com.sun.org.apache.xerces.internal.impl.XMLDocumentFragmentScannerImpl.scanEndElement(XMLDocumentFragmentScannerImpl.java:1782) at com.sun.org.apache.xerces.internal.impl.XMLDocumentFragmentScannerImpl$FragmentContentDriver.next(XMLDocumentFragmentScannerImpl.java:2938) at com.sun.org.apache.xerces.internal.impl.XMLDocumentScannerImpl.next(XMLDocumentScannerImpl.java:648) at com.sun.org.apache.xerces.internal.impl.XMLNSDocumentScannerImpl.next(XMLNSDocumentScannerImpl.java:140) at com.sun.org.apache.xerces.internal.impl.XMLDocumentFragmentScannerImpl.scanDocument(XMLDocumentFragmentScannerImpl.java:511) at com.sun.org.apache.xerces.internal.parsers.XML11Configuration.parse(XML11Configuration.java:808) at com.sun.org.apache.xerces.internal.parsers.XML11Configuration.parse(XML11Configuration.java:737) at com.sun.org.apache.xerces.internal.parsers.XMLParser.parse(XMLParser.java:119) at com.sun.org.apache.xerces.internal.parsers.AbstractSAXParser.parse(AbstractSAXParser.java:1205) at org.eclipse.mylyn.internal.bugzilla.core.AbstractReportFactory.collectResults(AbstractReportFactory.java:123) at org.eclipse.mylyn.internal.bugzilla.core.MultiBugReportFactory.populateReport(MultiBugReportFactory.java:48) at org.eclipse.mylyn.internal.bugzilla.core.BugzillaClient.getTaskData(BugzillaClient.java:2140) at org.eclipse.mylyn.internal.bugzilla.core.BugzillaTaskDataHandler.getMultiTaskData(BugzillaTaskDataHandler.java:367) at org.eclipse.mylyn.internal.tasks.core.sync.SynchronizeTasksJob.synchronizeTasks(SynchronizeTasksJob.java:305) at org.eclipse.mylyn.internal.tasks.core.sync.SynchronizeTasksJob.runInternal(SynchronizeTasksJob.java:207) at org.eclipse.mylyn.internal.tasks.core.sync.SynchronizeTasksJob.run(SynchronizeTasksJob.java:153) at org.eclipse.mylyn.internal.tasks.core.sync.SynchronizeTasksJob.run(SynchronizeTasksJob.java:129) at org.eclipse.core.internal.jobs.Worker.run(Worker.java:54) I have reverted the commit. Frank, if you have time can you look into covering this in a test case and propose a fix? Thanks!
Steffen, if we want an general solution me mud define if an empty tag is an null String or an empty String. Otherwise we need to enter code that trade some attributes in one case and other different. Thoughts?
Steffen, here is a review which fix the problems. http://review.mylyn.org/216
Created attachment 209887 [details] mylyn/context/zip
Thanks. If I understand the proposed fix correctly, tags with empty values are now always being ignored? What is the purpose of testFirstEmptyTag? Is that only relevant for nested tags?
(In reply to comment #14) > Thanks. If I understand the proposed fix correctly, tags with empty values are > now always being ignored? Yes empty tags for resolution, bug_file_loc, status_whiteboard and keywords need to not set the value to SPACE > What is the purpose of testFirstEmptyTag? Is that only relevant for nested tags? Yes example: ... </thetext> </long_desc> ...
Thanks. My sense is that it's too much risk in terms of regressions to backport this change. Since the impact of the bug is minor it should be sufficient to have this fix in master only.
I think I may have prematurely marked this as resolved. Was the change actually merged into master? I noticed that I see incoming notifications for the status whiteboard even though it has not changed on an internal Bugzilla 3.6 repository when I bootstrap on the change.
I recrate the patch and start an new review http://review.mylyn.org/#change,322
Created attachment 211601 [details] mylyn/context/zip
I have done a little bit of testing and didn't notice any unexpected highlighting. I have pushed the change to master. Thanks for looking into this, Frank!
I'm now getting a bunch of highlighting for tasks that I had already synchronized. We need a different approach to resolving this.
I have proposed a fix here: http://review.mylyn.org/354. Frank, can you do a review?
I create Patchset 2 with my modifications. please review.
Created attachment 212026 [details] mylyn/context/zip
Thanks for the review. I have pushed the latest patch set to master.
