Redirecting to lucene-user, which is more appropriate. I'm not sure what exactly the question is here, but:
Parse your XML document and for each <p> element you encounter create a new Document instance, and then populate its fields with some data, like the URI data you mentioned. If you parse with DOM - just walk the node tree and make new Document whenever you encounter an element you want as a separate Document. If you are using the SAX API you'll probably want some logic in start/endElement and characters methods. When you reach the end of the element you are done with your Document instance, so add it to the IndexWriter instance that you opened once, before the parser. When you are done with the whole XML document close the IndexWriter. Otis --- Murray Altheim <[EMAIL PROTECTED]> wrote: > Hi, > > I'm trying to develop a class to handle an XML document, where > the contents aren't so much indexed on a per-document basis, > rather on an element basis. Each element has a unique ID, so > I'm looking to create a class/method similar to Lucene's > Document.Document(). By way of example, I'll use some XHTML > markup to illustrate what I'm trying to do: > > <html> > <base href="http://purl.org/ceryle/blat.xml"/> > [...] > <body> > <p id="p1"> > some text to index... > </p> > <p id="p2"> > some more text to index... > </p> > <p id="p3"> > even more text to index... > </p> > </body> > </html> > > I'd very much appreciate any help in explaining how I'd go about > creating a method to return a Lucene Document to index this via > ID. Would I want a separate Document per <p>? (There are many > thousands of such elements.) Everything in my system, both at the > document and the individual element level is done via URL, so > the method should create URLs for each <p> element like > > http://purl.org/ceryle/blat.xml#p1 > http://purl.org/ceryle/blat.xml#p2 > http://purl.org/ceryle/blat.xml#p3 > etc. > > I don't need anyone to go to the trouble of coding this, just point > me to how it might be done, or to any existing examples that do this > kind of thing. > > Thanks very much! > > Murray > > ...................................................................... > Murray Altheim > http://kmi.open.ac.uk/people/murray/ > Knowledge Media Institute > The Open University, Milton Keynes, Bucks, MK7 6AA, UK > . > > "If we can just get the people that can reconcile themselves > to the new dispensation out of the way and then kill the few > thousand people who can't reconcile themselves, then we can > let the remaining 98 percent come back and live out their > lives," Pike said. "If we bomb the place to the ground, those > peace-loving people won't have a home to live in. [...] If we > simply pulverize the city, it would look bad on TV." -- John Pike > > U.S., Iraqi troops mass for assault on Fallujah > STRATEGY: U.S. to employ snipers, robots to cut down casualties > Matthew B. Stannard, San Francisco Chronicle > > http://www.sfgate.com/cgi-bin/article.cgi?file=/c/a/2004/11/06/MNGHL9NBU11.DTL > > "We have a growing, maturing insurgency group. We see larger > and more coordinated military attacks. They are getting better > and they can self-regenerate. The idea there are x number of > insurgents, and that when they're all dead we can get out is > wrong. The insurgency has shown an ability to regenerate itself > because there are people willing to fill the ranks of those who > are killed. The political culture is more hostile to the US > presence. The longer we stay, the more they are confirmed in > that view." -- W Andrew Terrill > > Far Graver Than Vietnam, Sidney Blumenthal, The Guardian > http://www.guardian.co.uk/comment/story/0,,1305360,00.html > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
