Hi -
We are using SOLR to process XML input files using the Data Import
Handler. I didn't see a way to move the xml files out of the way after
processing, so I wrote a small extension to allow this. The "How to
Contribute <http://wiki.apache.org/solr/HowToContribute>" page says to pitch
the request to the developer list in order to decide whether or not to
submit a patch. As such, here goes:
The new code basically extends FileDataSource and wraps the underlying
reader such that when the "close" method on the input stream is called, the
file is moved to a configurable archive directory. It is unclear to me
whether this is the correct place to put it (I pondered changing the
FileListEntityProcessor but this somehow felt safer). I realize that a more
robust implementation would consider the success status of the file being
processed and would also allow for configurable policies rather than a
concrete implementation. Nonetheless, I didn't want the perfect to be the
enemy of the good.
Please peruse the attached source code file and provide feedback as to
the merit of the idea, whether I ought to submit a JIRA ticket/patch and if
my approach is correct.
Thanks!
Josh Harness
ArchivingFileDataSource.java
Description: Binary data
--------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
