A simple wrapper for MoreLikeThis for use in scenarios where a Query object is required eg in custom QueryParser extensions. At query.rewrite() time the reader is used to construct the actual MoreLikeThis object and obtain the real Query object.
Inheritance: Lucene.Net.Search.Query
Exemplo n.º 1
0
 public void TestMoreLikeThisQuery()
 {
     Query query = new MoreLikeThisQuery("this is a test", new[] { "text" }, new MockAnalyzer(Random()), "text");
     QueryUtils.Check(Random(), query, searcher, Similarity);
 }
Exemplo n.º 2
0
        /// <summary>
        /// (non-Javadoc)
        /// @see org.apache.lucene.xmlparser.QueryObjectBuilder#process(org.w3c.dom.Element)
        /// </summary>
        public virtual Query GetQuery(XmlElement e)
        {
            string fieldsList = e.GetAttribute("fieldNames"); //a comma-delimited list of fields
            string[] fields = defaultFieldNames;
            if ((fieldsList != null) && (fieldsList.Trim().Length > 0))
            {
                fields = fieldsList.Trim().Split(new char[] { ',' }, StringSplitOptions.RemoveEmptyEntries);
                //trim the fieldnames
                for (int i = 0; i < fields.Length; i++)
                {
                    fields[i] = fields[i].Trim();
                }
            }

            //Parse any "stopWords" attribute
            //TODO MoreLikeThis needs to ideally have per-field stopWords lists - until then
            //I use all analyzers/fields to generate multi-field compatible stop list
            string stopWords = e.GetAttribute("stopWords");
            HashSet<string> stopWordsSet = null;
            if ((stopWords != null) && (fields != null))
            {
                stopWordsSet = new HashSet<string>();
                foreach (string field in fields)
                {
                    TokenStream ts = null;
                    try
                    {
                        ts = analyzer.TokenStream(field, stopWords);
                        ICharTermAttribute termAtt = ts.AddAttribute<ICharTermAttribute>();
                        ts.Reset();
                        while (ts.IncrementToken())
                        {
                            stopWordsSet.Add(termAtt.ToString());
                        }
                        ts.End();
                    }
                    catch (IOException ioe)
                    {
                        throw new ParserException("IoException parsing stop words list in "
                            + GetType().Name + ":" + ioe.Message);
                    }
                    finally
                    {
                        IOUtils.CloseWhileHandlingException(ts);
                    }
                }
            }

            MoreLikeThisQuery mlt = new MoreLikeThisQuery(DOMUtils.GetText(e), fields, analyzer, fields[0]);
            mlt.MaxQueryTerms = DOMUtils.GetAttribute(e, "maxQueryTerms", DEFAULT_MAX_QUERY_TERMS);
            mlt.MinTermFrequency = DOMUtils.GetAttribute(e, "minTermFrequency", DEFAULT_MIN_TERM_FREQUENCY);
            mlt.PercentTermsToMatch = DOMUtils.GetAttribute(e, "percentTermsToMatch", DEFAULT_PERCENT_TERMS_TO_MATCH) / 100;
            mlt.StopWords = stopWordsSet;
            int minDocFreq = DOMUtils.GetAttribute(e, "minDocFreq", -1);
            if (minDocFreq >= 0)
            {
                mlt.MinDocFreq = minDocFreq;
            }

            mlt.Boost = DOMUtils.GetAttribute(e, "boost", 1.0f);

            return mlt;
        }