Thanks man for helping out on this.. Thanks.. Okay Okay.. so Windows is okay.. I do not have much say in what we use here.. so. Which is fine.. I am happy.
I have the following directories, directly under my C:\EclipseWorkspaces\nutchTest\outputDir directory which nutch created.. crawldb index indexes linkdb Segments I know that these are magic names and the config files states that nutch will look for and use the index directory under this magically named tree so.. I am fine with that.... So I think that I am good.. so can I say "C:\EclipseWorkspaces\nutchTest\outputDir" as a searcher.dir value? I am going to try that.. Pointing Luke at that index directory works great. Now.. I am not running on Tomcat.. I am just running from within Eclipse as a regular old java project. This is not going to have a web interface so.. no need for a web server.. I am looking to output that search directory using a get method off of the config.. -----Original Message----- From: Andrzej Bialecki [mailto:[email protected]] Sent: Friday, March 06, 2009 9:26 AM To: [email protected] Subject: Re: Can not get Nutch query to work.. Can you help.. Lukas, Ray wrote: > Okay.. I did try that.. Good idea.. > I moved the index to C:\ and tried > <value>\outputDir</value> and > <value>c:\outputDir</value> > Did not help.. Basically I am running out of guesses you see.. > Maybe I don't know how to do this on evil MS Windows.. Maybe this does > not work on evil MS Windows? Is that the problem? All the examples I > have seen are unix based... Is this my problem.. Windows.. Ray, Nutch works just fine on Windows, even if it's not the best environment for it .. Please make sure that in your outputDir you have subdirs named exactly like this: linkdb indexes (or index, if you merged indexes using bin/nutch merge command) segments which contain the stuff that you crawled & parsed & indexed. These are "magic" names that the searcher tries to use - especially if it can't find either "indexes" or "index" things won't work. Check the Tomcat/Jetty logs - NutchBean should print an info there that says what directory it tries to use for indexes. -- Best regards, Andrzej Bialecki <>< ___. ___ ___ ___ _ _ __________________________________ [__ || __|__/|__||\/| Information Retrieval, Semantic Web ___|||__|| \| || | Embedded Unix, System Integration http://www.sigram.com Contact: info at sigram dot com
