The source code to Lucene in Action (www.lucenebook.com) has an
example of extending the built-in QueryParser to use SpanNearQuery
for in-order phrase searching instead of the default unordered
PhraseQuery behavior. That example will show how to use
SpanNearQuery made up of SpanTermQuery's. The source code to
QueryParser <http://svn.apache.org/repos/asf/lucene/java/trunk/src/
java/org/apache/lucene/queryParser/QueryParser.jj> shows how to use
PhraseQuery directly, if all you want is an example.
Also as Paul mentioned, the test cases in Lucene's source code make
great ways to learn how Lucene's API works.
If what you're trying to do is put an in-order wrapper around all of
the query types you mentioned, you will have a fair bit more work to
do in order to create a parser that can create a SpanNearQuery with
SpanRegexQuery's (new in contrib's trunk), and there currently is not
a SpanQuery version of FuzzyQuery that I'm aware of.
Erik
On Feb 7, 2006, at 4:06 PM, RONALD MANTAY wrote:
Hi, can any one help, I am new to lucene.
Due to requiring a specific Query syntax, I have created a
bespoke query parser to build a boolean query which at any time may
comprising a mixture of :
a) Fuzzy terms with similarity level set according to length of
term.text
b) Wild Card Querys
c) Term Querys
I would like to ability to define if the Query parser terms
require to be in Order or not (but when in order to allow gaps in
the term positions).
I can not see see how I can use PhraseQuery or SpanNearQuery
directly, can anyone point me in the right direction in how to
simulate inOrder to boolean queries.
Any help will be appreciated.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
