Hi@all
thanks for the links. In the meantime i used a spanquery with some
custom syntax highlighting to gather search suggestions. This works
pretty well, except some performance issues.
I use IndexReader.getTermVector to get the terms for a single document
to construct my suggestion spans according the the Spans collected from
SpanQuery.getSpans().
However, this operations seems to be quite slow. Any ideas how to tune
this? I only need a TermsEnum for the terms matching a span, so i do not
want to iterate over the whole document.
Thanks in advance,
Mirko
Am 04.09.2014 01:21, schrieb Jack Krupansky:
Here's the (skimpy) Lucene Javadoc:
http://lucene.apache.org/core/4_9_0/queryparser/org/apache/lucene/queryparser/complexPhrase/ComplexPhraseQueryParser.html
Take a look at the unit tests:
http://svn.apache.org/viewvc/lucene/dev/tags/lucene_solr_4_10_0/lucene/queryparser/src/test/org/apache/lucene/queryparser/complexPhrase/TestComplexPhraseQuery.java?revision=1622067&view=markup
-- Jack Krupansky
-----Original Message----- From: Erick Erickson
Sent: Wednesday, September 3, 2014 7:14 PM
To: java-user
Subject: Re: Question regarding complex queries and long tail suggestions
Take a look at the ComplexPhraseQueryParser here:
https://cwiki.apache.org/confluence/display/solr/Other+Parsers#OtherParsers-ComplexPhraseQueryParser
Best,
Erick
On Wed, Sep 3, 2014 at 12:41 PM, Mirko Sertic <mirko.ser...@web.de>
wrote:
Hi@all
I am using Lucene 4.9 for a search application.
Now i'd like to create complex queries such as:
"hello world * this is interesting"
This should match every dokument with the phrases "hello world" and
"this is
interesting" with any number of terms between them. I also want to go
further and create a custom query parser to generate queries like
"hello *
world this is * funny * sometimes" and so on. I want to combine this
syntax
with wildcards, for instance "hello wor* this * is funny", where wor*
can be
expanded to world and so on.
The key question is: do i have to construct this with nested queries
which
can become very complex, or would it be easier to use the Automaton
API and
construct an AutomatonQuery from it? What would be more performant? I
have
only read about the Automaton API, so code examples would be great!
Later, i want to create long tail suggestions for such kind of queries.
Unfortunately there is not very much documentation available for
this. Are
there code examples available?
Thanks in advance,
Mirko
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org
For additional commands, e-mail: java-user-h...@lucene.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org
For additional commands, e-mail: java-user-h...@lucene.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org
For additional commands, e-mail: java-user-h...@lucene.apache.org
