public MergeSortQueue(Sort sort, TopDocs[] shardHits)
: base(shardHits.Length)
{
this.shardHits = new ScoreDoc[shardHits.Length][];
for (int shardIDX = 0; shardIDX < shardHits.Length; shardIDX++)
{
ScoreDoc[] shard = shardHits[shardIDX].ScoreDocs;
//System.out.println(" init shardIdx=" + shardIDX + " hits=" + shard);
if (shard != null)
{
this.shardHits[shardIDX] = shard;
// Fail gracefully if API is misused:
for (int hitIDX = 0; hitIDX < shard.Length; hitIDX++)
{
ScoreDoc sd = shard[hitIDX];
if (!(sd is FieldDoc))
{
throw new System.ArgumentException("shard " + shardIDX + " was not sorted by the provided Sort (expected FieldDoc but got ScoreDoc)");
}
FieldDoc fd = (FieldDoc)sd;
if (fd.Fields == null)
{
throw new System.ArgumentException("shard " + shardIDX + " did not set sort field values (FieldDoc.fields is null); you must pass fillFields=true to IndexSearcher.search on each shard");
}
}
}
}
SortField[] sortFields = sort.GetSort();
comparers = new FieldComparer[sortFields.Length];
reverseMul = new int[sortFields.Length];
for (int compIDX = 0; compIDX < sortFields.Length; compIDX++)
{
SortField sortField = sortFields[compIDX];
comparers[compIDX] = sortField.GetComparer(1, compIDX);
reverseMul[compIDX] = sortField.IsReverse ? -1 : 1;
}
}
public TopDocs Call()
{
TopDocs docs = searcher.Search(Arrays.AsList(slice.Leaves), weight, after, nDocs);
ScoreDoc[] scoreDocs = docs.ScoreDocs;
//it would be so nice if we had a thread-safe insert
@lock.Lock();
try
{
for (int j = 0; j < scoreDocs.Length; j++) // merge scoreDocs into hq
{
ScoreDoc scoreDoc = scoreDocs[j];
if (scoreDoc == hq.InsertWithOverflow(scoreDoc))
{
break;
}
}
}
finally
{
@lock.Unlock();
}
return(docs);
}
public SearcherCallableNoSort(ReentrantLock @lock, IndexSearcher searcher, LeafSlice slice, Weight weight, ScoreDoc after, int nDocs, HitQueue hq)
{
this.@lock = @lock;
this.searcher = searcher;
this.weight = weight;
this.after = after;
this.nDocs = nDocs;
this.hq = hq;
this.slice = slice;
}
/// <summary>
/// Expert: Low-level search implementation. Finds the top <code>n</code>
/// hits for <c>query</c>.
///
/// <para/>Applications should usually call <see cref="IndexSearcher.Search(Query,int)"/> or
/// <see cref="IndexSearcher.Search(Query,Filter,int)"/> instead. </summary>
/// <exception cref="BooleanQuery.TooManyClausesException"> If a query would exceed
/// <see cref="BooleanQuery.MaxClauseCount"/> clauses. </exception>
protected virtual TopDocs Search(IList <AtomicReaderContext> leaves, Weight weight, ScoreDoc after, int nDocs)
{
// single thread
int limit = reader.MaxDoc;
if (limit == 0)
{
limit = 1;
}
nDocs = Math.Min(nDocs, limit);
TopScoreDocCollector collector = TopScoreDocCollector.Create(nDocs, after, !weight.ScoresDocsOutOfOrder);
Search(leaves, weight, collector);
return(collector.GetTopDocs());
}
/// <summary>
/// Finds the top <paramref name="n"/>
/// hits for <paramref name="query"/>, applying <paramref name="filter"/> if non-null,
/// where all results are after a previous result (<paramref name="after"/>).
/// <para/>
/// By passing the bottom result from a previous page as <paramref name="after"/>,
/// this method can be used for efficient 'deep-paging' across potentially
/// large result sets.
/// </summary>
/// <exception cref="BooleanQuery.TooManyClausesException"> If a query would exceed
/// <see cref="BooleanQuery.MaxClauseCount"/> clauses. </exception>
public virtual TopDocs SearchAfter(ScoreDoc after, Query query, Filter filter, int n)
{
return(Search(CreateNormalizedWeight(WrapFilter(query, filter)), after, n));
}
/// <summary>
/// Finds the top <paramref name="n"/>
/// hits for top <paramref name="query"/> where all results are after a previous
/// result (top <paramref name="after"/>).
/// <para/>
/// By passing the bottom result from a previous page as <paramref name="after"/>,
/// this method can be used for efficient 'deep-paging' across potentially
/// large result sets.
/// </summary>
/// <exception cref="BooleanQuery.TooManyClausesException"> If a query would exceed
/// <see cref="BooleanQuery.MaxClauseCount"/> clauses. </exception>
public virtual TopDocs SearchAfter(ScoreDoc after, Query query, int n)
{
return(Search(CreateNormalizedWeight(query), after, n));
}
internal InOrderPagingScoreDocCollector(ScoreDoc after, int numHits)
: base(numHits)
{
this.after = after;
}
public override TopDocs Rescore(IndexSearcher searcher, TopDocs firstPassTopDocs, int topN)
{
ScoreDoc[] hits = (ScoreDoc[])firstPassTopDocs.ScoreDocs.Clone();
Array.Sort(hits, Comparer <ScoreDoc> .Create((a, b) => a.Doc - b.Doc));
IList <AtomicReaderContext> leaves = searcher.IndexReader.Leaves;
Weight weight = searcher.CreateNormalizedWeight(query);
// Now merge sort docIDs from hits, with reader's leaves:
int hitUpto = 0;
int readerUpto = -1;
int endDoc = 0;
int docBase = 0;
Scorer scorer = null;
while (hitUpto < hits.Length)
{
ScoreDoc hit = hits[hitUpto];
int docID = hit.Doc;
AtomicReaderContext readerContext = null;
while (docID >= endDoc)
{
readerUpto++;
readerContext = leaves[readerUpto];
endDoc = readerContext.DocBase + readerContext.Reader.MaxDoc;
}
if (readerContext != null)
{
// We advanced to another segment:
docBase = readerContext.DocBase;
scorer = weight.GetScorer(readerContext, null);
}
int targetDoc = docID - docBase;
int actualDoc = scorer.DocID;
if (actualDoc < targetDoc)
{
actualDoc = scorer.Advance(targetDoc);
}
if (actualDoc == targetDoc)
{
// Query did match this doc:
hit.Score = Combine(hit.Score, true, scorer.GetScore());
}
else
{
// Query did not match this doc:
if (Debugging.AssertsEnabled)
{
Debugging.Assert(actualDoc > targetDoc);
}
hit.Score = Combine(hit.Score, false, 0.0f);
}
hitUpto++;
}
// TODO: we should do a partial sort (of only topN)
// instead, but typically the number of hits is
// smallish:
Array.Sort(hits, Comparer <ScoreDoc> .Create((a, b) =>
{
// Sort by score descending, then docID ascending:
if (a.Score > b.Score)
{
return(-1);
}
else if (a.Score < b.Score)
{
return(1);
}
else
{
// this subtraction can't overflow int
// because docIDs are >= 0:
return(a.Doc - b.Doc);
}
}));
if (topN < hits.Length)
{
ScoreDoc[] subset = new ScoreDoc[topN];
Array.Copy(hits, 0, subset, 0, topN);
hits = subset;
}
return(new TopDocs(firstPassTopDocs.TotalHits, hits, hits[0].Score));
}
public static TopDocs Merge(Sort sort, int start, int size, TopDocs[] shardHits)
{
Util.PriorityQueue <ShardRef> queue;
if (sort == null)
{
queue = new ScoreMergeSortQueue(shardHits);
}
else
{
queue = new MergeSortQueue(sort, shardHits);
}
int totalHitCount = 0;
int availHitCount = 0;
float maxScore = float.MinValue;
for (int shardIDX = 0; shardIDX < shardHits.Length; shardIDX++)
{
TopDocs shard = shardHits[shardIDX];
// totalHits can be non-zero even if no hits were
// collected, when searchAfter was used:
totalHitCount += shard.TotalHits;
if (shard.ScoreDocs != null && shard.ScoreDocs.Length > 0)
{
availHitCount += shard.ScoreDocs.Length;
queue.Add(new ShardRef(shardIDX));
maxScore = Math.Max(maxScore, shard.MaxScore);
//System.out.println(" maxScore now " + maxScore + " vs " + shard.getMaxScore());
}
}
if (availHitCount == 0)
{
maxScore = float.NaN;
}
ScoreDoc[] hits;
if (availHitCount <= start)
{
hits = new ScoreDoc[0];
}
else
{
hits = new ScoreDoc[Math.Min(size, availHitCount - start)];
int requestedResultWindow = start + size;
int numIterOnHits = Math.Min(availHitCount, requestedResultWindow);
int hitUpto = 0;
while (hitUpto < numIterOnHits)
{
Debug.Assert(queue.Count > 0);
ShardRef @ref = queue.Pop();
ScoreDoc hit = shardHits[@ref.ShardIndex].ScoreDocs[@ref.HitIndex++];
hit.ShardIndex = @ref.ShardIndex;
if (hitUpto >= start)
{
hits[hitUpto - start] = hit;
}
//System.out.println(" hitUpto=" + hitUpto);
//System.out.println(" doc=" + hits[hitUpto].doc + " score=" + hits[hitUpto].score);
hitUpto++;
if (@ref.HitIndex < shardHits[@ref.ShardIndex].ScoreDocs.Length)
{
// Not done with this these TopDocs yet:
queue.Add(@ref);
}
}
}
if (sort == null)
{
return(new TopDocs(totalHitCount, hits, maxScore));
}
else
{
return(new TopFieldDocs(totalHitCount, hits, sort.GetSort(), maxScore));
}
}
