When creating an index of a large directory structure, DocFetcher sometimes fails to clean up some of the temporary files it creates in /tmp. It looks like there are several different causes for that behavior and I have managed to track down and fix one of them: There can be an exception while trying to extract files from certain broken rar archives. The exception happens in RarTree.java while calling "archive.extractFile(fh, out);". All exceptions from this call are catched but the exception handlers do not clean up the temporary file (which is already created and possibly partly written at that time).
I've attached a simple patch to mitigate the issue and delete the temporary file in the exception handlers. It is based on the latest version in the git (commit 0d28c499df4a25d83d27af21c12259f4a3b9d174).
Anonymous
Hi,
thanks a lot for reporting and investigating this problem. Unfortunately, I don't have time to look into the matter (and the patch) in the near future, but I have noted the bug as high-priority.
Best regards
q:-) <= Quang