Hi Dror,
thanks for your answer.
> >
> > I'm having problems understanding query parsers handling of AND and OR
> > if there's more than one operator.
> >
> > E.g.
> > a OR b AND c
> > gives the same number of hits as
> > b AND c
> > (only scores are different)
>
> This would make sense if all the document that have a also have both B
> and C in them.
>
Then the query should be equivalent to (a OR b) AND c.
But it isn't. For specific a, b and c I get 766 hits for a OR b AND c
and 1086 for (a OR b) AND c.
> >
> > and
> > a AND b OR c AND d
> > seems to be equivalent to
> > a AND b AND C AND d
> >
>
> That's not what I get.
> http://www.fastbuzz.com/search/results.jsp?query=dean+AND+kerry+AND+clark+AND+gephardt&days=
> returns 479 items
> but
> http://www.fastbuzz.com/search/results.jsp?query=dean+AND+kerry+OR+clark+AND+gephardt&days=
> returns 564 items which indicates that the OR does make a difference.
> As expcted, you end up getting more items with the OR.
>
Hmm. I was sloppy not specifying the lucene version.
My tests were on 1.2.
But I reindex a part of my documents using 1.3rc3 and find the same.
What version does fastbuzz use?
I wrote s small test programm indexing all documents consisting of
one or zero occurences of a, b, c and d (ignoring order, so without
the empty document, that's just 15 docs) and performing some queries
on it.
Programm see below, this is what I get:
a OR b AND c > a +b +c
4 documents found
a b c
a b c d
b c
b c d
(a OR b) AND c > +(a b) +c
6 documents found
a b c
a b c d
a c
b c
a c d
b c d
a OR (b AND c) > a (+b +c)
10 documents found
a b c
a b c d
b c
a
b c d
a b
a c
a d
a b d
a c d
b AND c > +b +c
4 documents found
b c
a b c
b c d
a b c d
a AND b OR c AND d > +a +b +c +d
1 documents found
a b c d
(a AND b) OR (c AND d) > (+a +b) (+c +d)
7 documents found
a b c d
a b
c d
a b c
a b d
a c d
b c d
a AND (b OR c) AND d > +a +(b c) +d
3 documents found
a b c d
a b d
a c d
((a AND b) OR c) AND d > +((+a +b) c) +d
5 documents found
a b c d
a b d
c d
a c d
b c d
a AND (b OR (c AND d)) > +a +(b (+c +d))
5 documents found
a b c d
a c d
a b
a b c
a b d
a AND b AND c AND d > +a +b +c +d
1 documents found
a b c d
Using 1.3rc3, 1.3rc2 or 1.3rc1; I get the same results with a slightly
different order for 1.2.
So I still get the same for
a OR b AND c and b AND c
and
a AND b OR c AND d and a AND b AND c AND d
(note, that the result of the toString method of the query is equal in
both cases)
but different results for any operator grouping, I can think of.
So to me, the question remains, what does AND and OR mean, if they are
combined in one expression?
I can understand all the query results where AND and OR queries are
explicitly grouped by paranthesis, and the results are, what I expect.
But the rules for combined AND and OR aren't what I would expect.
greetings
Morus
PS: the test program:
import org.apache.lucene.document.*;
import org.apache.lucene.analysis.*;
import org.apache.lucene.analysis.standard.StandardAnalyzer;
import org.apache.lucene.index.*;
import org.apache.lucene.store.*;
import org.apache.lucene.search.*;
import org.apache.lucene.queryParser.QueryParser;
class LuceneTest
{
static String[] docs = {
"a", "b", "c", "d",
"a b", "a c", "a d", "b c", "b d", "c d",
"a b c", "a b d", "a c d", "b c d",
"a b c d"
};
static String[] queries = {
"a OR b AND c",
"(a OR b) AND c",
"a OR (b AND c)",
"b AND c",
"a AND b OR c AND d",
"(a AND b) OR (c AND d)",
"a AND (b OR c) AND d",
"((a AND b) OR c) AND d",
"a AND (b OR (c AND d))",
"a AND b AND c AND d"
};
public static void main(String argv[]) throws Exception {
Directory dir = new RAMDirectory();
String[] stop = {};
Analyzer analyzer = new StandardAnalyzer(stop);
IndexWriter writer = new IndexWriter(dir, analyzer, true);
for ( int i=0; i < docs.length; i++ ) {
Document doc = new Document();
doc.add(Field.Text("text", docs[i]));
writer.addDocument(doc);
}
writer.close();
Searcher searcher = new IndexSearcher(dir);
for ( int i=0; i < queries.length; i++ ) {
Query query = QueryParser.parse(queries[i], "text", analyzer);
System.out.println(queries[i] + " > " + query.toString("text"));
Hits hits = searcher.search(query);
System.out.println(" " + hits.length() + " documents found");
for ( int j=0; j < hits.length(); j++ ) {
Document doc = hits.doc(j);
System.out.println("\t"+doc.get("text"));
}
}
}
}

To unsubscribe, email: luceneuserunsubscribe@jakarta.apache.org
For additional commands, email: luceneuserhelp@jakarta.apache.org
