/// <summary>
/// Builds a new TermQuery instance.
/// <para>This is intended for subclasses that wish to customize the generated queries.</para> </summary>
/// <param name="term"> term </param>
/// <param name="context"> the TermContext to be used to create the low level term query. Can be <code>null</code>. </param>
/// <returns> new TermQuery instance </returns>
protected virtual Query NewTermQuery(Term term, TermContext context)
{
return context == null ? new TermQuery(term) : new TermQuery(term, context);
}
public virtual void CollectTermContext(IndexReader reader, IList<AtomicReaderContext> leaves, TermContext[] contextArray, Term[] queryTerms)
{
TermsEnum termsEnum = null;
foreach (AtomicReaderContext context in leaves)
{
Fields fields = context.AtomicReader.Fields;
if (fields == null)
{
// reader has no fields
continue;
}
for (int i = 0; i < queryTerms.Length; i++)
{
Term term = queryTerms[i];
TermContext termContext = contextArray[i];
Terms terms = fields.Terms(term.Field);
if (terms == null)
{
// field does not exist
continue;
}
termsEnum = terms.Iterator(termsEnum);
Debug.Assert(termsEnum != null);
if (termsEnum == TermsEnum.EMPTY)
{
continue;
}
if (termsEnum.SeekExact(term.Bytes))
{
if (termContext == null)
{
contextArray[i] = new TermContext(reader.Context, termsEnum.TermState(), context.Ord, termsEnum.DocFreq(), termsEnum.TotalTermFreq());
}
else
{
termContext.Register(termsEnum.TermState(), context.Ord, termsEnum.DocFreq(), termsEnum.TotalTermFreq());
}
}
}
}
}
protected internal virtual Query BuildQuery(int maxDoc, TermContext[] contextArray, Term[] queryTerms)
{
var lowFreq = new BooleanQuery(disableCoord);
var highFreq = new BooleanQuery(disableCoord) { Boost = highFreqBoost };
lowFreq.Boost = lowFreqBoost;
var query = new BooleanQuery(true);
for (int i = 0; i < queryTerms.Length; i++)
{
TermContext termContext = contextArray[i];
if (termContext == null)
{
lowFreq.Add(NewTermQuery(queryTerms[i], null), lowFreqOccur);
}
else
{
if ((maxTermFrequency >= 1f && termContext.DocFreq > maxTermFrequency) || (termContext.DocFreq > (int)Math.Ceiling(maxTermFrequency * (float)maxDoc)))
{
highFreq.Add(NewTermQuery(queryTerms[i], termContext), highFreqOccur);
}
else
{
lowFreq.Add(NewTermQuery(queryTerms[i], termContext), lowFreqOccur);
}
}
}
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int numLowFreqClauses = lowFreq.clauses().size();
int numLowFreqClauses = lowFreq.Clauses.Length;
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int numHighFreqClauses = highFreq.clauses().size();
int numHighFreqClauses = highFreq.Clauses.Length;
if (lowFreqOccur == BooleanClause.Occur.SHOULD && numLowFreqClauses > 0)
{
int minMustMatch = CalcLowFreqMinimumNumberShouldMatch(numLowFreqClauses);
lowFreq.MinimumNumberShouldMatch = minMustMatch;
}
if (highFreqOccur == BooleanClause.Occur.SHOULD && numHighFreqClauses > 0)
{
int minMustMatch = CalcHighFreqMinimumNumberShouldMatch(numHighFreqClauses);
highFreq.MinimumNumberShouldMatch = minMustMatch;
}
if (lowFreq.Clauses.Length == 0)
{
/*
* if lowFreq is empty we rewrite the high freq terms in a conjunction to
* prevent slow queries.
*/
if (highFreq.MinimumNumberShouldMatch == 0 && highFreqOccur != BooleanClause.Occur.MUST)
{
foreach (BooleanClause booleanClause in highFreq)
{
booleanClause.Occur_ = BooleanClause.Occur.MUST;
}
}
highFreq.Boost = Boost;
return highFreq;
}
else if (highFreq.Clauses.Length == 0)
{
// only do low freq terms - we don't have high freq terms
lowFreq.Boost = Boost;
return lowFreq;
}
else
{
query.Add(highFreq, BooleanClause.Occur.SHOULD);
query.Add(lowFreq, BooleanClause.Occur.MUST);
query.Boost = Boost;
return query;
}
}
public override Query Rewrite(IndexReader reader)
{
if (this.terms.Count == 0)
{
return new BooleanQuery();
}
else if (this.terms.Count == 1)
{
Query tq = NewTermQuery(this.terms[0], null);
tq.Boost = Boost;
return tq;
}
var leaves = reader.Leaves;
int maxDoc = reader.MaxDoc;
var contextArray = new TermContext[terms.Count];
var queryTerms = this.terms.ToArray();
CollectTermContext(reader, leaves, contextArray, queryTerms);
return BuildQuery(maxDoc, contextArray, queryTerms);
}
/// <summary>
/// Creates a <seealso cref="TermContext"/> from a top-level <seealso cref="IndexReaderContext"/> and the
/// given <seealso cref="Term"/>. this method will lookup the given term in all context's leaf readers
/// and register each of the readers containing the term in the returned <seealso cref="TermContext"/>
/// using the leaf reader's ordinal.
/// <p>
/// Note: the given context must be a top-level context.
/// </summary>
public static TermContext Build(IndexReaderContext context, Term term)
{
Debug.Assert(context != null && context.IsTopLevel);
string field = term.Field;
BytesRef bytes = term.Bytes;
TermContext perReaderTermState = new TermContext(context);
//if (DEBUG) System.out.println("prts.build term=" + term);
foreach (AtomicReaderContext ctx in context.Leaves)
{
//if (DEBUG) System.out.println(" r=" + leaves[i].reader);
Fields fields = ctx.AtomicReader.Fields;
if (fields != null)
{
Terms terms = fields.Terms(field);
if (terms != null)
{
TermsEnum termsEnum = terms.Iterator(null);
if (termsEnum.SeekExact(bytes))
{
TermState termState = termsEnum.TermState();
//if (DEBUG) System.out.println(" found");
perReaderTermState.Register(termState, ctx.Ord, termsEnum.DocFreq(), termsEnum.TotalTermFreq());
}
}
}
}
return perReaderTermState;
}
public virtual void CollectTermContext(IndexReader reader, IList<AtomicReaderContext> leaves, TermContext[] contextArray, Term[] queryTerms)
{
TermsEnum termsEnum = null;
foreach (AtomicReaderContext context in leaves)
{
Fields fields = context.AtomicReader.Fields;
if (fields == null)
{
// reader has no fields
continue;
}
for (int i = 0; i < queryTerms.Length; i++)
{
Term term = queryTerms[i];
TermContext termContext = contextArray[i];
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.index.Terms terms = fields.terms(term.field());
Terms terms = fields.Terms(term.Field());
if (terms == null)
{
// field does not exist
continue;
}
termsEnum = terms.Iterator(termsEnum);
Debug.Assert(termsEnum != null);
if (termsEnum == TermsEnum.EMPTY)
{
continue;
}
if (termsEnum.SeekExact(term.Bytes()))
{
if (termContext == null)
{
contextArray[i] = new TermContext(reader.Context, termsEnum.TermState(), context.Ord, termsEnum.DocFreq(), termsEnum.TotalTermFreq());
}
else
{
termContext.Register(termsEnum.TermState(), context.Ord, termsEnum.DocFreq(), termsEnum.TotalTermFreq());
}
}
}
}
}
