[ https://issues.apache.org/jira/browse/LUCENE-1198?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Michael McCandless updated LUCENE-1198: --------------------------------------- Attachment: LUCENE-1198.patch Attached patch. I added a unit test showing the issue. To do this, I added this method to IndexWriter: boolean testPoint(String name) Then in DocumentsWriter.ThreadState.init I added this: assert writer.testPoint("DocumentsWriter.ThreadState.init start"); Then, tests can subclass IndexWriter and do interesting things at each of these test points. We can add further test points over time... Then I fixed the issue and the tests (& all tests) pass. I'll commit in a day or two. > Exception in DocumentsWriter.ThreadState.init leads to corruption > ----------------------------------------------------------------- > > Key: LUCENE-1198 > URL: https://issues.apache.org/jira/browse/LUCENE-1198 > Project: Lucene - Java > Issue Type: Bug > Components: Index > Affects Versions: 2.3 > Reporter: Michael McCandless > Assignee: Michael McCandless > Priority: Minor > Fix For: 2.4 > > Attachments: LUCENE-1198.patch > > > If an exception is hit in the init method, DocumentsWriter incorrectly > increments numDocsInRAM when in fact the document is not added. > Spinoff of this thread: > http://markmail.org/message/e76hgkgldxhakuaa > The root cause that led to the exception in init was actually due to > incorrect use of Lucene's APIs (one thread still modifying the > Document while IndexWriter.addDocument is adding it) but still we > should protect against any exceptions coming out of init. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]