public virtual void TestRandomLongs()
{
Directory dir = NewDirectory();
var w = new RandomIndexWriter(Random(), dir);
int numDocs = AtLeast(1000);
if (VERBOSE)
{
Console.WriteLine("TEST: numDocs=" + numDocs);
}
long[] values = new long[numDocs];
long minValue = long.MaxValue;
long maxValue = long.MinValue;
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
long v = Random().NextLong();
values[i] = v;
doc.Add(new NumericDocValuesField("field", v));
doc.Add(new LongField("field", v, Field.Store.NO));
w.AddDocument(doc);
minValue = Math.Min(minValue, v);
maxValue = Math.Max(maxValue, v);
}
IndexReader r = w.Reader;
IndexSearcher s = NewSearcher(r);
FacetsConfig config = new FacetsConfig();
int numIters = AtLeast(10);
for (int iter = 0; iter < numIters; iter++)
{
if (VERBOSE)
{
Console.WriteLine("TEST: iter=" + iter);
}
int numRange = TestUtil.NextInt(Random(), 1, 100);
LongRange[] ranges = new LongRange[numRange];
int[] expectedCounts = new int[numRange];
long minAcceptedValue = long.MaxValue;
long maxAcceptedValue = long.MinValue;
for (int rangeID = 0; rangeID < numRange; rangeID++)
{
long min;
if (rangeID > 0 && Random().Next(10) == 7)
{
// Use an existing boundary:
LongRange prevRange = ranges[Random().Next(rangeID)];
if (Random().NextBoolean())
{
min = prevRange.min;
}
else
{
min = prevRange.max;
}
}
else
{
min = Random().NextLong();
}
long max;
if (rangeID > 0 && Random().Next(10) == 7)
{
// Use an existing boundary:
LongRange prevRange = ranges[Random().Next(rangeID)];
if (Random().NextBoolean())
{
max = prevRange.min;
}
else
{
max = prevRange.max;
}
}
else
{
max = Random().NextLong();
}
if (min > max)
{
long x = min;
min = max;
max = x;
}
bool minIncl;
bool maxIncl;
if (min == max)
{
minIncl = true;
maxIncl = true;
}
else
{
minIncl = Random().NextBoolean();
maxIncl = Random().NextBoolean();
}
ranges[rangeID] = new LongRange("r" + rangeID, min, minIncl, max, maxIncl);
if (VERBOSE)
{
Console.WriteLine(" range " + rangeID + ": " + ranges[rangeID]);
}
// Do "slow but hopefully correct" computation of
// expected count:
for (int i = 0; i < numDocs; i++)
{
bool accept = true;
if (minIncl)
{
accept &= values[i] >= min;
}
else
{
accept &= values[i] > min;
}
if (maxIncl)
{
accept &= values[i] <= max;
}
else
{
accept &= values[i] < max;
}
if (accept)
{
expectedCounts[rangeID]++;
minAcceptedValue = Math.Min(minAcceptedValue, values[i]);
maxAcceptedValue = Math.Max(maxAcceptedValue, values[i]);
}
}
}
FacetsCollector sfc = new FacetsCollector();
s.Search(new MatchAllDocsQuery(), sfc);
Filter fastMatchFilter;
if (Random().NextBoolean())
{
if (Random().NextBoolean())
{
fastMatchFilter = NumericRangeFilter.NewLongRange("field", minValue, maxValue, true, true);
}
else
{
fastMatchFilter = NumericRangeFilter.NewLongRange("field", minAcceptedValue, maxAcceptedValue, true, true);
}
}
else
{
fastMatchFilter = null;
}
ValueSource vs = new LongFieldSource("field");
Facets facets = new LongRangeFacetCounts("field", vs, sfc, fastMatchFilter, ranges);
FacetResult result = facets.GetTopChildren(10, "field");
Assert.AreEqual(numRange, result.LabelValues.Length);
for (int rangeID = 0; rangeID < numRange; rangeID++)
{
if (VERBOSE)
{
Console.WriteLine(" range " + rangeID + " expectedCount=" + expectedCounts[rangeID]);
}
LabelAndValue subNode = result.LabelValues[rangeID];
Assert.AreEqual("r" + rangeID, subNode.label);
Assert.AreEqual(expectedCounts[rangeID], (int)subNode.value);
LongRange range = ranges[rangeID];
// Test drill-down:
DrillDownQuery ddq = new DrillDownQuery(config);
if (Random().NextBoolean())
{
if (Random().NextBoolean())
{
ddq.Add("field", NumericRangeFilter.NewLongRange("field", range.min, range.max, range.minInclusive, range.maxInclusive));
}
else
{
ddq.Add("field", NumericRangeQuery.NewLongRange("field", range.min, range.max, range.minInclusive, range.maxInclusive));
}
}
else
{
ddq.Add("field", range.GetFilter(fastMatchFilter, vs));
}
Assert.AreEqual(expectedCounts[rangeID], s.Search(ddq, 10).TotalHits);
}
}
IOUtils.Close(w, r, dir);
}
public virtual void TestBasic()
{
Store.Directory dir = NewDirectory();
Store.Directory taxoDir = NewDirectory();
// Writes facet ords to a separate directory from the
// main index:
var taxoWriter = new DirectoryTaxonomyWriter(taxoDir, IndexWriterConfig.OpenMode_e.CREATE);
FacetsConfig config = new FacetsConfig();
config.SetHierarchical("Publish Date", true);
RandomIndexWriter writer = new RandomIndexWriter(Random(), dir);
Document doc = new Document();
doc.Add(new FacetField("Author", "Bob"));
doc.Add(new FacetField("Publish Date", "2010", "10", "15"));
writer.AddDocument(config.Build(taxoWriter, doc));
doc = new Document();
doc.Add(new FacetField("Author", "Lisa"));
doc.Add(new FacetField("Publish Date", "2010", "10", "20"));
writer.AddDocument(config.Build(taxoWriter, doc));
doc = new Document();
doc.Add(new FacetField("Author", "Lisa"));
doc.Add(new FacetField("Publish Date", "2012", "1", "1"));
writer.AddDocument(config.Build(taxoWriter, doc));
doc = new Document();
doc.Add(new FacetField("Author", "Susan"));
doc.Add(new FacetField("Publish Date", "2012", "1", "7"));
writer.AddDocument(config.Build(taxoWriter, doc));
doc = new Document();
doc.Add(new FacetField("Author", "Frank"));
doc.Add(new FacetField("Publish Date", "1999", "5", "5"));
writer.AddDocument(config.Build(taxoWriter, doc));
// NRT open
IndexSearcher searcher = NewSearcher(writer.Reader);
// NRT open
var taxoReader = new DirectoryTaxonomyReader(taxoWriter);
// Aggregate the facet counts:
FacetsCollector c = new FacetsCollector();
// MatchAllDocsQuery is for "browsing" (counts facets
// for all non-deleted docs in the index); normally
// you'd use a "normal" query, and use MultiCollector to
// wrap collecting the "normal" hits and also facets:
searcher.Search(new MatchAllDocsQuery(), c);
Facets facets = new FastTaxonomyFacetCounts(taxoReader, config, c);
// Retrieve & verify results:
Assert.AreEqual("dim=Publish Date path=[] value=5 childCount=3\n 2010 (2)\n 2012 (2)\n 1999 (1)\n", facets.GetTopChildren(10, "Publish Date").ToString());
Assert.AreEqual("dim=Author path=[] value=5 childCount=4\n Lisa (2)\n Bob (1)\n Susan (1)\n Frank (1)\n", facets.GetTopChildren(10, "Author").ToString());
// Now user drills down on Publish Date/2010:
DrillDownQuery q2 = new DrillDownQuery(config);
q2.Add("Publish Date", "2010");
c = new FacetsCollector();
searcher.Search(q2, c);
facets = new FastTaxonomyFacetCounts(taxoReader, config, c);
Assert.AreEqual("dim=Author path=[] value=2 childCount=2\n Bob (1)\n Lisa (1)\n", facets.GetTopChildren(10, "Author").ToString());
Assert.AreEqual(1, facets.GetSpecificValue("Author", "Lisa"));
Assert.Null(facets.GetTopChildren(10, "Non exitent dim"));
// Smoke test PrintTaxonomyStats:
ByteArrayOutputStream bos = new ByteArrayOutputStream();
PrintTaxonomyStats.PrintStats(taxoReader, Console.Out, true);
string result = bos.ToString();
Assert.True(result.IndexOf("/Author: 4 immediate children; 5 total categories", StringComparison.Ordinal) != -1);
Assert.True(result.IndexOf("/Publish Date: 3 immediate children; 12 total categories", StringComparison.Ordinal) != -1);
// Make sure at least a few nodes of the tree came out:
Assert.True(result.IndexOf(" /1999", StringComparison.Ordinal) != -1);
Assert.True(result.IndexOf(" /2012", StringComparison.Ordinal) != -1);
Assert.True(result.IndexOf(" /20", StringComparison.Ordinal) != -1);
IOUtils.Close(writer, taxoWriter, searcher.IndexReader, taxoReader, taxoDir, dir);
}
public virtual void TestMixedRangeAndNonRangeTaxonomy()
{
Directory d = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(Random(), d);
Directory td = NewDirectory();
DirectoryTaxonomyWriter tw = new DirectoryTaxonomyWriter(td, IndexWriterConfig.OpenMode_e.CREATE);
FacetsConfig config = new FacetsConfig();
for (long l = 0; l < 100; l++)
{
Document doc = new Document();
// For computing range facet counts:
doc.Add(new NumericDocValuesField("field", l));
// For drill down by numeric range:
doc.Add(new LongField("field", l, Field.Store.NO));
if ((l & 3) == 0)
{
doc.Add(new FacetField("dim", "a"));
}
else
{
doc.Add(new FacetField("dim", "b"));
}
w.AddDocument(config.Build(tw, doc));
}
IndexReader r = w.Reader;
var tr = new DirectoryTaxonomyReader(tw);
IndexSearcher s = NewSearcher(r);
if (VERBOSE)
{
Console.WriteLine("TEST: searcher=" + s);
}
DrillSideways ds = new DrillSidewaysAnonymousInnerClassHelper(this, s, config, tr);
// First search, no drill downs:
DrillDownQuery ddq = new DrillDownQuery(config);
DrillSidewaysResult dsr = ds.Search(null, ddq, 10);
Assert.AreEqual(100, dsr.Hits.TotalHits);
Assert.AreEqual("dim=dim path=[] value=100 childCount=2\n b (75)\n a (25)\n", dsr.Facets.GetTopChildren(10, "dim").ToString());
Assert.AreEqual("dim=field path=[] value=21 childCount=5\n less than 10 (10)\n less than or equal to 10 (11)\n over 90 (9)\n 90 or above (10)\n over 1000 (0)\n", dsr.Facets.GetTopChildren(10, "field").ToString());
// Second search, drill down on dim=b:
ddq = new DrillDownQuery(config);
ddq.Add("dim", "b");
dsr = ds.Search(null, ddq, 10);
Assert.AreEqual(75, dsr.Hits.TotalHits);
Assert.AreEqual("dim=dim path=[] value=100 childCount=2\n b (75)\n a (25)\n", dsr.Facets.GetTopChildren(10, "dim").ToString());
Assert.AreEqual("dim=field path=[] value=16 childCount=5\n less than 10 (7)\n less than or equal to 10 (8)\n over 90 (7)\n 90 or above (8)\n over 1000 (0)\n", dsr.Facets.GetTopChildren(10, "field").ToString());
// Third search, drill down on "less than or equal to 10":
ddq = new DrillDownQuery(config);
ddq.Add("field", NumericRangeQuery.NewLongRange("field", 0L, 10L, true, true));
dsr = ds.Search(null, ddq, 10);
Assert.AreEqual(11, dsr.Hits.TotalHits);
Assert.AreEqual("dim=dim path=[] value=11 childCount=2\n b (8)\n a (3)\n", dsr.Facets.GetTopChildren(10, "dim").ToString());
Assert.AreEqual("dim=field path=[] value=21 childCount=5\n less than 10 (10)\n less than or equal to 10 (11)\n over 90 (9)\n 90 or above (10)\n over 1000 (0)\n", dsr.Facets.GetTopChildren(10, "field").ToString());
IOUtils.Close(tw, tr, td, w, r, d);
}
public virtual void TestRandomFloats()
{
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(Random(), dir);
int numDocs = AtLeast(1000);
float[] values = new float[numDocs];
float minValue = float.PositiveInfinity;
float maxValue = float.NegativeInfinity;
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
float v = Random().NextFloat();
values[i] = v;
doc.Add(new FloatDocValuesField("field", v));
doc.Add(new FloatField("field", v, Field.Store.NO));
w.AddDocument(doc);
minValue = Math.Min(minValue, v);
maxValue = Math.Max(maxValue, v);
}
IndexReader r = w.Reader;
IndexSearcher s = NewSearcher(r);
FacetsConfig config = new FacetsConfig();
int numIters = AtLeast(10);
for (int iter = 0; iter < numIters; iter++)
{
if (VERBOSE)
{
Console.WriteLine("TEST: iter=" + iter);
}
int numRange = TestUtil.NextInt(Random(), 1, 5);
DoubleRange[] ranges = new DoubleRange[numRange];
int[] expectedCounts = new int[numRange];
float minAcceptedValue = float.PositiveInfinity;
float maxAcceptedValue = float.NegativeInfinity;
if (VERBOSE)
{
Console.WriteLine("TEST: " + numRange + " ranges");
}
for (int rangeID = 0; rangeID < numRange; rangeID++)
{
double min;
if (rangeID > 0 && Random().Next(10) == 7)
{
// Use an existing boundary:
DoubleRange prevRange = ranges[Random().Next(rangeID)];
if (Random().NextBoolean())
{
min = prevRange.Min;
}
else
{
min = prevRange.Max;
}
}
else
{
min = Random().NextDouble();
}
double max;
if (rangeID > 0 && Random().Next(10) == 7)
{
// Use an existing boundary:
DoubleRange prevRange = ranges[Random().Next(rangeID)];
if (Random().NextBoolean())
{
max = prevRange.Min;
}
else
{
max = prevRange.Max;
}
}
else
{
max = Random().NextDouble();
}
if (min > max)
{
double x = min;
min = max;
max = x;
}
// Must truncate to float precision so that the
// drill-down counts (which use NRQ.newFloatRange)
// are correct:
min = (float)min;
max = (float)max;
bool minIncl;
bool maxIncl;
if (min == max)
{
minIncl = true;
maxIncl = true;
}
else
{
minIncl = Random().NextBoolean();
maxIncl = Random().NextBoolean();
}
ranges[rangeID] = new DoubleRange("r" + rangeID, min, minIncl, max, maxIncl);
if (VERBOSE)
{
Console.WriteLine("TEST: range " + rangeID + ": " + ranges[rangeID]);
}
// Do "slow but hopefully correct" computation of
// expected count:
for (int i = 0; i < numDocs; i++)
{
bool accept = true;
if (minIncl)
{
accept &= values[i] >= min;
}
else
{
accept &= values[i] > min;
}
if (maxIncl)
{
accept &= values[i] <= max;
}
else
{
accept &= values[i] < max;
}
if (VERBOSE)
{
Console.WriteLine("TEST: check doc=" + i + " val=" + values[i] + " accept=" + accept);
}
if (accept)
{
expectedCounts[rangeID]++;
minAcceptedValue = Math.Min(minAcceptedValue, values[i]);
maxAcceptedValue = Math.Max(maxAcceptedValue, values[i]);
}
}
}
FacetsCollector sfc = new FacetsCollector();
s.Search(new MatchAllDocsQuery(), sfc);
Filter fastMatchFilter;
if (Random().NextBoolean())
{
if (Random().NextBoolean())
{
fastMatchFilter = NumericRangeFilter.NewFloatRange("field", minValue, maxValue, true, true);
}
else
{
fastMatchFilter = NumericRangeFilter.NewFloatRange("field", minAcceptedValue, maxAcceptedValue, true, true);
}
}
else
{
fastMatchFilter = null;
}
ValueSource vs = new FloatFieldSource("field");
Facets facets = new DoubleRangeFacetCounts("field", vs, sfc, fastMatchFilter, ranges);
FacetResult result = facets.GetTopChildren(10, "field");
Assert.AreEqual(numRange, result.LabelValues.Length);
for (int rangeID = 0; rangeID < numRange; rangeID++)
{
if (VERBOSE)
{
Console.WriteLine("TEST: verify range " + rangeID + " expectedCount=" + expectedCounts[rangeID]);
}
LabelAndValue subNode = result.LabelValues[rangeID];
Assert.AreEqual("r" + rangeID, subNode.label);
Assert.AreEqual(expectedCounts[rangeID], (int)subNode.value);
DoubleRange range = ranges[rangeID];
// Test drill-down:
DrillDownQuery ddq = new DrillDownQuery(config);
if (Random().NextBoolean())
{
if (Random().NextBoolean())
{
ddq.Add("field", NumericRangeFilter.NewFloatRange("field", (float)range.Min, (float)range.Max, range.MinInclusive, range.MaxInclusive));
}
else
{
ddq.Add("field", NumericRangeQuery.NewFloatRange("field", (float)range.Min, (float)range.Max, range.MinInclusive, range.MaxInclusive));
}
}
else
{
ddq.Add("field", range.GetFilter(fastMatchFilter, vs));
}
Assert.AreEqual(expectedCounts[rangeID], s.Search(ddq, 10).TotalHits);
}
}
IOUtils.Close(w, r, dir);
}
public virtual void TestRandomFloats()
{
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
Random, dir);
int numDocs = AtLeast(1000);
float[] values = new float[numDocs];
float minValue = float.PositiveInfinity;
float maxValue = float.NegativeInfinity;
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
float v = Random.NextSingle();
values[i] = v;
doc.Add(new SingleDocValuesField("field", v));
doc.Add(new SingleField("field", v, Field.Store.NO));
w.AddDocument(doc);
minValue = Math.Min(minValue, v);
maxValue = Math.Max(maxValue, v);
}
IndexReader r = w.GetReader();
IndexSearcher s = NewSearcher(r);
FacetsConfig config = new FacetsConfig();
int numIters = AtLeast(10);
for (int iter = 0; iter < numIters; iter++)
{
if (VERBOSE)
{
Console.WriteLine("TEST: iter=" + iter);
}
int numRange = TestUtil.NextInt32(Random, 1, 5);
DoubleRange[] ranges = new DoubleRange[numRange];
int[] expectedCounts = new int[numRange];
float minAcceptedValue = float.PositiveInfinity;
float maxAcceptedValue = float.NegativeInfinity;
if (VERBOSE)
{
Console.WriteLine("TEST: " + numRange + " ranges");
}
for (int rangeID = 0; rangeID < numRange; rangeID++)
{
double min;
if (rangeID > 0 && Random.Next(10) == 7)
{
// Use an existing boundary:
DoubleRange prevRange = ranges[Random.Next(rangeID)];
if (Random.NextBoolean())
{
min = prevRange.Min;
}
else
{
min = prevRange.Max;
}
}
else
{
min = Random.NextDouble();
}
double max;
if (rangeID > 0 && Random.Next(10) == 7)
{
// Use an existing boundary:
DoubleRange prevRange = ranges[Random.Next(rangeID)];
if (Random.NextBoolean())
{
max = prevRange.Min;
}
else
{
max = prevRange.Max;
}
}
else
{
max = Random.NextDouble();
}
if (min > max)
{
double x = min;
min = max;
max = x;
}
// Must truncate to float precision so that the
// drill-down counts (which use NRQ.newFloatRange)
// are correct:
min = (float)min;
max = (float)max;
bool minIncl;
bool maxIncl;
if (min == max)
{
minIncl = true;
maxIncl = true;
}
else
{
minIncl = Random.NextBoolean();
maxIncl = Random.NextBoolean();
}
ranges[rangeID] = new DoubleRange("r" + rangeID, min, minIncl, max, maxIncl);
if (VERBOSE)
{
Console.WriteLine("TEST: range " + rangeID + ": " + ranges[rangeID]);
}
// Do "slow but hopefully correct" computation of
// expected count:
for (int i = 0; i < numDocs; i++)
{
bool accept = true;
if (minIncl)
{
accept &= values[i] >= min;
}
else
{
accept &= values[i] > min;
}
if (maxIncl)
{
accept &= values[i] <= max;
}
else
{
accept &= values[i] < max;
}
if (VERBOSE)
{
Console.WriteLine("TEST: check doc=" + i + " val=" + values[i] + " accept=" + accept);
}
if (accept)
{
expectedCounts[rangeID]++;
minAcceptedValue = Math.Min(minAcceptedValue, values[i]);
maxAcceptedValue = Math.Max(maxAcceptedValue, values[i]);
}
}
}
FacetsCollector sfc = new FacetsCollector();
s.Search(new MatchAllDocsQuery(), sfc);
Filter fastMatchFilter;
if (Random.NextBoolean())
{
if (Random.NextBoolean())
{
fastMatchFilter = NumericRangeFilter.NewSingleRange("field", minValue, maxValue, true, true);
}
else
{
fastMatchFilter = NumericRangeFilter.NewSingleRange("field", minAcceptedValue, maxAcceptedValue, true, true);
}
}
else
{
fastMatchFilter = null;
}
ValueSource vs = new SingleFieldSource("field");
Facets facets = new DoubleRangeFacetCounts("field", vs, sfc, fastMatchFilter, ranges);
FacetResult result = facets.GetTopChildren(10, "field");
Assert.AreEqual(numRange, result.LabelValues.Length);
for (int rangeID = 0; rangeID < numRange; rangeID++)
{
if (VERBOSE)
{
Console.WriteLine("TEST: verify range " + rangeID + " expectedCount=" + expectedCounts[rangeID]);
}
LabelAndValue subNode = result.LabelValues[rangeID];
Assert.AreEqual("r" + rangeID, subNode.Label);
Assert.AreEqual(expectedCounts[rangeID], (int)subNode.Value);
DoubleRange range = ranges[rangeID];
// Test drill-down:
DrillDownQuery ddq = new DrillDownQuery(config);
if (Random.NextBoolean())
{
if (Random.NextBoolean())
{
ddq.Add("field", NumericRangeFilter.NewSingleRange("field", (float)range.Min, (float)range.Max, range.MinInclusive, range.MaxInclusive));
}
else
{
ddq.Add("field", NumericRangeQuery.NewSingleRange("field", (float)range.Min, (float)range.Max, range.MinInclusive, range.MaxInclusive));
}
}
else
{
ddq.Add("field", range.GetFilter(fastMatchFilter, vs));
}
Assert.AreEqual(expectedCounts[rangeID], s.Search(ddq, 10).TotalHits);
}
}
IOUtils.Dispose(w, r, dir);
}
public virtual void TestCustomDoublesValueSource()
{
Directory dir = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(Random(), dir);
Document doc = new Document();
writer.AddDocument(doc);
writer.AddDocument(doc);
writer.AddDocument(doc);
// Test wants 3 docs in one segment:
writer.ForceMerge(1);
var vs = new ValueSourceAnonymousInnerClassHelper(this, doc);
FacetsConfig config = new FacetsConfig();
FacetsCollector fc = new FacetsCollector();
IndexReader r = writer.Reader;
IndexSearcher s = NewSearcher(r);
s.Search(new MatchAllDocsQuery(), fc);
DoubleRange[] ranges = new DoubleRange[] { new DoubleRange("< 1", 0.0, true, 1.0, false), new DoubleRange("< 2", 0.0, true, 2.0, false), new DoubleRange("< 5", 0.0, true, 5.0, false), new DoubleRange("< 10", 0.0, true, 10.0, false), new DoubleRange("< 20", 0.0, true, 20.0, false), new DoubleRange("< 50", 0.0, true, 50.0, false) };
Filter fastMatchFilter;
AtomicBoolean filterWasUsed = new AtomicBoolean();
if (Random().NextBoolean())
{
// Sort of silly:
fastMatchFilter = new CachingWrapperFilterAnonymousInnerClassHelper(this, new QueryWrapperFilter(new MatchAllDocsQuery()), filterWasUsed);
}
else
{
fastMatchFilter = null;
}
if (VERBOSE)
{
Console.WriteLine("TEST: fastMatchFilter=" + fastMatchFilter);
}
Facets facets = new DoubleRangeFacetCounts("field", vs, fc, fastMatchFilter, ranges);
Assert.AreEqual("dim=field path=[] value=3 childCount=6\n < 1 (0)\n < 2 (1)\n < 5 (3)\n < 10 (3)\n < 20 (3)\n < 50 (3)\n", facets.GetTopChildren(10, "field").ToString());
Assert.True(fastMatchFilter == null || filterWasUsed.Get());
DrillDownQuery ddq = new DrillDownQuery(config);
ddq.Add("field", ranges[1].GetFilter(fastMatchFilter, vs));
// Test simple drill-down:
Assert.AreEqual(1, s.Search(ddq, 10).TotalHits);
// Test drill-sideways after drill-down
DrillSideways ds = new DrillSidewaysAnonymousInnerClassHelper2(this, s, config, (TaxonomyReader)null, vs, ranges, fastMatchFilter);
DrillSidewaysResult dsr = ds.Search(ddq, 10);
Assert.AreEqual(1, dsr.Hits.TotalHits);
Assert.AreEqual("dim=field path=[] value=3 childCount=6\n < 1 (0)\n < 2 (1)\n < 5 (3)\n < 10 (3)\n < 20 (3)\n < 50 (3)\n", dsr.Facets.GetTopChildren(10, "field").ToString());
IOUtils.Close(r, writer, dir);
}
public virtual void TestCustomDoublesValueSource()
{
Directory dir = NewDirectory();
RandomIndexWriter writer = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
Random, dir);
Document doc = new Document();
writer.AddDocument(doc);
writer.AddDocument(doc);
writer.AddDocument(doc);
// Test wants 3 docs in one segment:
writer.ForceMerge(1);
var vs = new ValueSourceAnonymousInnerClassHelper(this, doc);
FacetsConfig config = new FacetsConfig();
FacetsCollector fc = new FacetsCollector();
IndexReader r = writer.GetReader();
IndexSearcher s = NewSearcher(r);
s.Search(new MatchAllDocsQuery(), fc);
DoubleRange[] ranges = new DoubleRange[] { new DoubleRange("< 1", 0.0, true, 1.0, false), new DoubleRange("< 2", 0.0, true, 2.0, false), new DoubleRange("< 5", 0.0, true, 5.0, false), new DoubleRange("< 10", 0.0, true, 10.0, false), new DoubleRange("< 20", 0.0, true, 20.0, false), new DoubleRange("< 50", 0.0, true, 50.0, false) };
Filter fastMatchFilter;
AtomicBoolean filterWasUsed = new AtomicBoolean();
if (Random.NextBoolean())
{
// Sort of silly:
fastMatchFilter = new CachingWrapperFilterAnonymousInnerClassHelper(this, new QueryWrapperFilter(new MatchAllDocsQuery()), filterWasUsed);
}
else
{
fastMatchFilter = null;
}
if (VERBOSE)
{
Console.WriteLine("TEST: fastMatchFilter=" + fastMatchFilter);
}
Facets facets = new DoubleRangeFacetCounts("field", vs, fc, fastMatchFilter, ranges);
Assert.AreEqual("dim=field path=[] value=3 childCount=6\n < 1 (0)\n < 2 (1)\n < 5 (3)\n < 10 (3)\n < 20 (3)\n < 50 (3)\n", facets.GetTopChildren(10, "field").ToString());
Assert.True(fastMatchFilter == null || filterWasUsed.Get());
DrillDownQuery ddq = new DrillDownQuery(config);
ddq.Add("field", ranges[1].GetFilter(fastMatchFilter, vs));
// Test simple drill-down:
Assert.AreEqual(1, s.Search(ddq, 10).TotalHits);
// Test drill-sideways after drill-down
DrillSideways ds = new DrillSidewaysAnonymousInnerClassHelper2(this, s, config, (TaxonomyReader)null, vs, ranges, fastMatchFilter);
DrillSidewaysResult dsr = ds.Search(ddq, 10);
Assert.AreEqual(1, dsr.Hits.TotalHits);
Assert.AreEqual("dim=field path=[] value=3 childCount=6\n < 1 (0)\n < 2 (1)\n < 5 (3)\n < 10 (3)\n < 20 (3)\n < 50 (3)\n", dsr.Facets.GetTopChildren(10, "field").ToString());
IOUtils.Dispose(r, writer, dir);
}
public virtual void TestIntSumAssociationDrillDown()
{
FacetsCollector fc = new FacetsCollector();
IndexSearcher searcher = NewSearcher(reader);
DrillDownQuery q = new DrillDownQuery(config);
q.Add("int", "b");
searcher.Search(q, fc);
Facets facets = new TaxonomyFacetSumIntAssociations("$facets.int", taxoReader, config, fc);
Assert.AreEqual("dim=int path=[] value=-1 childCount=2\n b (150)\n a (100)\n", facets.GetTopChildren(10, "int").ToString());
Assert.AreEqual(100, (int)facets.GetSpecificValue("int", "a"), "Wrong count for category 'a'!");
Assert.AreEqual(150, (int)facets.GetSpecificValue("int", "b"), "Wrong count for category 'b'!");
}
public virtual void TestBasic()
{
Store.Directory dir = NewDirectory();
Store.Directory taxoDir = NewDirectory();
// Writes facet ords to a separate directory from the
// main index:
var taxoWriter = new DirectoryTaxonomyWriter(taxoDir, OpenMode.CREATE);
FacetsConfig config = new FacetsConfig();
config.SetHierarchical("Publish Date", true);
RandomIndexWriter writer = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
Random, dir);
Document doc = new Document();
doc.Add(new FacetField("Author", "Bob"));
doc.Add(new FacetField("Publish Date", "2010", "10", "15"));
writer.AddDocument(config.Build(taxoWriter, doc));
doc = new Document();
doc.Add(new FacetField("Author", "Lisa"));
doc.Add(new FacetField("Publish Date", "2010", "10", "20"));
writer.AddDocument(config.Build(taxoWriter, doc));
doc = new Document();
doc.Add(new FacetField("Author", "Lisa"));
doc.Add(new FacetField("Publish Date", "2012", "1", "1"));
writer.AddDocument(config.Build(taxoWriter, doc));
doc = new Document();
doc.Add(new FacetField("Author", "Susan"));
doc.Add(new FacetField("Publish Date", "2012", "1", "7"));
writer.AddDocument(config.Build(taxoWriter, doc));
doc = new Document();
doc.Add(new FacetField("Author", "Frank"));
doc.Add(new FacetField("Publish Date", "1999", "5", "5"));
writer.AddDocument(config.Build(taxoWriter, doc));
// NRT open
IndexSearcher searcher = NewSearcher(writer.GetReader());
// NRT open
var taxoReader = new DirectoryTaxonomyReader(taxoWriter);
// Aggregate the facet counts:
FacetsCollector c = new FacetsCollector();
// MatchAllDocsQuery is for "browsing" (counts facets
// for all non-deleted docs in the index); normally
// you'd use a "normal" query, and use MultiCollector to
// wrap collecting the "normal" hits and also facets:
searcher.Search(new MatchAllDocsQuery(), c);
Facets facets = new FastTaxonomyFacetCounts(taxoReader, config, c);
// Retrieve & verify results:
Assert.AreEqual("dim=Publish Date path=[] value=5 childCount=3\n 2010 (2)\n 2012 (2)\n 1999 (1)\n", facets.GetTopChildren(10, "Publish Date").ToString());
Assert.AreEqual("dim=Author path=[] value=5 childCount=4\n Lisa (2)\n Bob (1)\n Susan (1)\n Frank (1)\n", facets.GetTopChildren(10, "Author").ToString());
// Now user drills down on Publish Date/2010:
DrillDownQuery q2 = new DrillDownQuery(config);
q2.Add("Publish Date", "2010");
c = new FacetsCollector();
searcher.Search(q2, c);
facets = new FastTaxonomyFacetCounts(taxoReader, config, c);
Assert.AreEqual("dim=Author path=[] value=2 childCount=2\n Bob (1)\n Lisa (1)\n", facets.GetTopChildren(10, "Author").ToString());
Assert.AreEqual(1, facets.GetSpecificValue("Author", "Lisa"));
Assert.Null(facets.GetTopChildren(10, "Non exitent dim"));
// Smoke test PrintTaxonomyStats:
string result;
using (ByteArrayOutputStream bos = new ByteArrayOutputStream())
{
using (StreamWriter w = new StreamWriter(bos, Encoding.UTF8, 2048, true)
{
AutoFlush = true
})
{
PrintTaxonomyStats.PrintStats(taxoReader, w, true);
}
result = bos.ToString();
}
Assert.True(result.IndexOf("/Author: 4 immediate children; 5 total categories", StringComparison.Ordinal) != -1);
Assert.True(result.IndexOf("/Publish Date: 3 immediate children; 12 total categories", StringComparison.Ordinal) != -1);
// Make sure at least a few nodes of the tree came out:
Assert.True(result.IndexOf(" /1999", StringComparison.Ordinal) != -1);
Assert.True(result.IndexOf(" /2012", StringComparison.Ordinal) != -1);
Assert.True(result.IndexOf(" /20", StringComparison.Ordinal) != -1);
IOUtils.Dispose(writer, taxoWriter, searcher.IndexReader, taxoReader, taxoDir, dir);
}
public virtual void TestRandomDoubles()
{
Directory dir = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(Random(), dir, Similarity, TimeZone);
int numDocs = AtLeast(1000);
double[] values = new double[numDocs];
double minValue = double.PositiveInfinity;
double maxValue = double.NegativeInfinity;
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
double v = Random().NextDouble();
values[i] = v;
doc.Add(new DoubleDocValuesField("field", v));
doc.Add(new DoubleField("field", v, Field.Store.NO));
w.AddDocument(doc);
minValue = Math.Min(minValue, v);
maxValue = Math.Max(maxValue, v);
}
IndexReader r = w.Reader;
IndexSearcher s = NewSearcher(r);
FacetsConfig config = new FacetsConfig();
int numIters = AtLeast(10);
for (int iter = 0; iter < numIters; iter++)
{
if (VERBOSE)
{
Console.WriteLine("TEST: iter=" + iter);
}
int numRange = TestUtil.NextInt(Random(), 1, 5);
DoubleRange[] ranges = new DoubleRange[numRange];
int[] expectedCounts = new int[numRange];
double minAcceptedValue = double.PositiveInfinity;
double maxAcceptedValue = double.NegativeInfinity;
for (int rangeID = 0; rangeID < numRange; rangeID++)
{
double min;
if (rangeID > 0 && Random().Next(10) == 7)
{
// Use an existing boundary:
DoubleRange prevRange = ranges[Random().Next(rangeID)];
if (Random().NextBoolean())
{
min = prevRange.Min;
}
else
{
min = prevRange.Max;
}
}
else
{
min = Random().NextDouble();
}
double max;
if (rangeID > 0 && Random().Next(10) == 7)
{
// Use an existing boundary:
DoubleRange prevRange = ranges[Random().Next(rangeID)];
if (Random().NextBoolean())
{
max = prevRange.Min;
}
else
{
max = prevRange.Max;
}
}
else
{
max = Random().NextDouble();
}
if (min > max)
{
double x = min;
min = max;
max = x;
}
bool minIncl;
bool maxIncl;
if (min == max)
{
minIncl = true;
maxIncl = true;
}
else
{
minIncl = Random().NextBoolean();
maxIncl = Random().NextBoolean();
}
ranges[rangeID] = new DoubleRange("r" + rangeID, min, minIncl, max, maxIncl);
// Do "slow but hopefully correct" computation of
// expected count:
for (int i = 0; i < numDocs; i++)
{
bool accept = true;
if (minIncl)
{
accept &= values[i] >= min;
}
else
{
accept &= values[i] > min;
}
if (maxIncl)
{
accept &= values[i] <= max;
}
else
{
accept &= values[i] < max;
}
if (accept)
{
expectedCounts[rangeID]++;
minAcceptedValue = Math.Min(minAcceptedValue, values[i]);
maxAcceptedValue = Math.Max(maxAcceptedValue, values[i]);
}
}
}
FacetsCollector sfc = new FacetsCollector();
s.Search(new MatchAllDocsQuery(), sfc);
Filter fastMatchFilter;
if (Random().NextBoolean())
{
if (Random().NextBoolean())
{
fastMatchFilter = NumericRangeFilter.NewDoubleRange("field", minValue, maxValue, true, true);
}
else
{
fastMatchFilter = NumericRangeFilter.NewDoubleRange("field", minAcceptedValue, maxAcceptedValue, true, true);
}
}
else
{
fastMatchFilter = null;
}
ValueSource vs = new DoubleFieldSource("field");
Facets facets = new DoubleRangeFacetCounts("field", vs, sfc, fastMatchFilter, ranges);
FacetResult result = facets.GetTopChildren(10, "field");
Assert.AreEqual(numRange, result.LabelValues.Length);
for (int rangeID = 0; rangeID < numRange; rangeID++)
{
if (VERBOSE)
{
Console.WriteLine(" range " + rangeID + " expectedCount=" + expectedCounts[rangeID]);
}
LabelAndValue subNode = result.LabelValues[rangeID];
Assert.AreEqual("r" + rangeID, subNode.label);
Assert.AreEqual(expectedCounts[rangeID], (int)subNode.value);
DoubleRange range = ranges[rangeID];
// Test drill-down:
DrillDownQuery ddq = new DrillDownQuery(config);
if (Random().NextBoolean())
{
if (Random().NextBoolean())
{
ddq.Add("field", NumericRangeFilter.NewDoubleRange("field", range.Min, range.Max, range.MinInclusive, range.MaxInclusive));
}
else
{
ddq.Add("field", NumericRangeQuery.NewDoubleRange("field", range.Min, range.Max, range.MinInclusive, range.MaxInclusive));
}
}
else
{
ddq.Add("field", range.GetFilter(fastMatchFilter, vs));
}
Assert.AreEqual(expectedCounts[rangeID], s.Search(ddq, 10).TotalHits);
}
}
IOUtils.Close(w, r, dir);
}
private DrillDownQuery ToQuery(SearchOptions searchOptions, CategoryOptions categoryOptions)
{
var query = new DrillDownQuery(Facets.FacetsConfig, ToQuery(searchOptions.Query));
if (categoryOptions != null && categoryOptions.DrillDown.Any())
{
foreach (var kvp in categoryOptions.DrillDown)
query.add(kvp.Key, kvp.Value.ToArray());
}
return query;
}
public virtual void TestBasic()
{
AssumeTrue("Test requires SortedSetDV support", DefaultCodecSupportsSortedSet());
Directory dir = NewDirectory();
FacetsConfig config = new FacetsConfig();
config.SetMultiValued("a", true);
RandomIndexWriter writer = new RandomIndexWriter(Random(), dir, Similarity, TimeZone);
Document doc = new Document();
doc.Add(new SortedSetDocValuesFacetField("a", "foo"));
doc.Add(new SortedSetDocValuesFacetField("a", "bar"));
doc.Add(new SortedSetDocValuesFacetField("a", "zoo"));
doc.Add(new SortedSetDocValuesFacetField("b", "baz"));
writer.AddDocument(config.Build(doc));
if (Random().NextBoolean())
{
writer.Commit();
}
doc = new Document();
doc.Add(new SortedSetDocValuesFacetField("a", "foo"));
writer.AddDocument(config.Build(doc));
// NRT open
IndexSearcher searcher = NewSearcher(writer.Reader);
// Per-top-reader state:
SortedSetDocValuesReaderState state = new DefaultSortedSetDocValuesReaderState(searcher.IndexReader);
FacetsCollector c = new FacetsCollector();
searcher.Search(new MatchAllDocsQuery(), c);
SortedSetDocValuesFacetCounts facets = new SortedSetDocValuesFacetCounts(state, c);
Assert.AreEqual("dim=a path=[] value=4 childCount=3\n foo (2)\n bar (1)\n zoo (1)\n", facets.GetTopChildren(10, "a").ToString());
Assert.AreEqual("dim=b path=[] value=1 childCount=1\n baz (1)\n", facets.GetTopChildren(10, "b").ToString());
// DrillDown:
DrillDownQuery q = new DrillDownQuery(config);
q.Add("a", "foo");
q.Add("b", "baz");
TopDocs hits = searcher.Search(q, 1);
Assert.AreEqual(1, hits.TotalHits);
IOUtils.Close(writer, searcher.IndexReader, dir);
}
public virtual void TestRandomLongs()
{
Directory dir = NewDirectory();
var w = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
Random, dir);
int numDocs = AtLeast(1000);
if (Verbose)
{
Console.WriteLine("TEST: numDocs=" + numDocs);
}
long[] values = new long[numDocs];
long minValue = long.MaxValue;
long maxValue = long.MinValue;
for (int i = 0; i < numDocs; i++)
{
Document doc = new Document();
long v = Random.NextInt64();
values[i] = v;
doc.Add(new NumericDocValuesField("field", v));
doc.Add(new Int64Field("field", v, Field.Store.NO));
w.AddDocument(doc);
minValue = Math.Min(minValue, v);
maxValue = Math.Max(maxValue, v);
}
IndexReader r = w.GetReader();
IndexSearcher s = NewSearcher(r);
FacetsConfig config = new FacetsConfig();
int numIters = AtLeast(10);
for (int iter = 0; iter < numIters; iter++)
{
if (Verbose)
{
Console.WriteLine("TEST: iter=" + iter);
}
int numRange = TestUtil.NextInt32(Random, 1, 100);
Int64Range[] ranges = new Int64Range[numRange];
int[] expectedCounts = new int[numRange];
long minAcceptedValue = long.MaxValue;
long maxAcceptedValue = long.MinValue;
for (int rangeID = 0; rangeID < numRange; rangeID++)
{
long min;
if (rangeID > 0 && Random.Next(10) == 7)
{
// Use an existing boundary:
Int64Range prevRange = ranges[Random.Next(rangeID)];
if (Random.NextBoolean())
{
min = prevRange.Min;
}
else
{
min = prevRange.Max;
}
}
else
{
min = Random.NextInt64();
}
long max;
if (rangeID > 0 && Random.Next(10) == 7)
{
// Use an existing boundary:
Int64Range prevRange = ranges[Random.Next(rangeID)];
if (Random.NextBoolean())
{
max = prevRange.Min;
}
else
{
max = prevRange.Max;
}
}
else
{
max = Random.NextInt64();
}
if (min > max)
{
long x = min;
min = max;
max = x;
}
bool minIncl;
bool maxIncl;
if (min == max)
{
minIncl = true;
maxIncl = true;
}
else
{
minIncl = Random.NextBoolean();
maxIncl = Random.NextBoolean();
}
ranges[rangeID] = new Int64Range("r" + rangeID, min, minIncl, max, maxIncl);
if (Verbose)
{
Console.WriteLine(" range " + rangeID + ": " + ranges[rangeID]);
}
// Do "slow but hopefully correct" computation of
// expected count:
for (int i = 0; i < numDocs; i++)
{
bool accept = true;
if (minIncl)
{
accept &= values[i] >= min;
}
else
{
accept &= values[i] > min;
}
if (maxIncl)
{
accept &= values[i] <= max;
}
else
{
accept &= values[i] < max;
}
if (accept)
{
expectedCounts[rangeID]++;
minAcceptedValue = Math.Min(minAcceptedValue, values[i]);
maxAcceptedValue = Math.Max(maxAcceptedValue, values[i]);
}
}
}
FacetsCollector sfc = new FacetsCollector();
s.Search(new MatchAllDocsQuery(), sfc);
Filter fastMatchFilter;
if (Random.NextBoolean())
{
if (Random.NextBoolean())
{
fastMatchFilter = NumericRangeFilter.NewInt64Range("field", minValue, maxValue, true, true);
}
else
{
fastMatchFilter = NumericRangeFilter.NewInt64Range("field", minAcceptedValue, maxAcceptedValue, true, true);
}
}
else
{
fastMatchFilter = null;
}
ValueSource vs = new Int64FieldSource("field");
Facets facets = new Int64RangeFacetCounts("field", vs, sfc, fastMatchFilter, ranges);
FacetResult result = facets.GetTopChildren(10, "field");
Assert.AreEqual(numRange, result.LabelValues.Length);
for (int rangeID = 0; rangeID < numRange; rangeID++)
{
if (Verbose)
{
Console.WriteLine(" range " + rangeID + " expectedCount=" + expectedCounts[rangeID]);
}
LabelAndValue subNode = result.LabelValues[rangeID];
Assert.AreEqual("r" + rangeID, subNode.Label);
Assert.AreEqual(expectedCounts[rangeID], (int)subNode.Value);
Int64Range range = ranges[rangeID];
// Test drill-down:
DrillDownQuery ddq = new DrillDownQuery(config);
if (Random.NextBoolean())
{
if (Random.NextBoolean())
{
ddq.Add("field", NumericRangeFilter.NewInt64Range("field", range.Min, range.Max, range.MinInclusive, range.MaxInclusive));
}
else
{
ddq.Add("field", NumericRangeQuery.NewInt64Range("field", range.Min, range.Max, range.MinInclusive, range.MaxInclusive));
}
}
else
{
ddq.Add("field", range.GetFilter(fastMatchFilter, vs));
}
Assert.AreEqual(expectedCounts[rangeID], s.Search(ddq, 10).TotalHits);
}
}
IOUtils.Dispose(w, r, dir);
}
public virtual void TestHugeLabel()
{
Directory indexDir = NewDirectory(), taxoDir = NewDirectory();
IndexWriter indexWriter = new IndexWriter(indexDir, NewIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random())));
DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(taxoDir, OpenMode.CREATE, new Cl2oTaxonomyWriterCache(2, 1f, 1));
FacetsConfig config = new FacetsConfig();
// Add one huge label:
string bigs = null;
int ordinal = -1;
int len = FacetLabel.MAX_CATEGORY_PATH_LENGTH - 4; // for the dimension and separator
bigs = TestUtil.RandomSimpleString(Random(), len, len);
FacetField ff = new FacetField("dim", bigs);
FacetLabel cp = new FacetLabel("dim", bigs);
ordinal = taxoWriter.AddCategory(cp);
Document doc = new Document();
doc.Add(ff);
indexWriter.AddDocument(config.Build(taxoWriter, doc));
// Add tiny ones to cause a re-hash
for (int i = 0; i < 3; i++)
{
string s = TestUtil.RandomSimpleString(Random(), 1, 10);
taxoWriter.AddCategory(new FacetLabel("dim", s));
doc = new Document();
doc.Add(new FacetField("dim", s));
indexWriter.AddDocument(config.Build(taxoWriter, doc));
}
// when too large components were allowed to be added, this resulted in a new added category
Assert.AreEqual(ordinal, taxoWriter.AddCategory(cp));
IOUtils.Close(indexWriter, taxoWriter);
DirectoryReader indexReader = DirectoryReader.Open(indexDir);
var taxoReader = new DirectoryTaxonomyReader(taxoDir);
IndexSearcher searcher = new IndexSearcher(indexReader);
DrillDownQuery ddq = new DrillDownQuery(new FacetsConfig());
ddq.Add("dim", bigs);
Assert.AreEqual(1, searcher.Search(ddq, 10).TotalHits);
IOUtils.Close(indexReader, taxoReader, indexDir, taxoDir);
}
public virtual void TestMixedRangeAndNonRangeTaxonomy()
{
Directory d = NewDirectory();
RandomIndexWriter w = new RandomIndexWriter(
#if FEATURE_INSTANCE_TESTDATA_INITIALIZATION
this,
#endif
Random, d);
Directory td = NewDirectory();
DirectoryTaxonomyWriter tw = new DirectoryTaxonomyWriter(td, OpenMode.CREATE);
FacetsConfig config = new FacetsConfig();
for (long l = 0; l < 100; l++)
{
Document doc = new Document();
// For computing range facet counts:
doc.Add(new NumericDocValuesField("field", l));
// For drill down by numeric range:
doc.Add(new Int64Field("field", l, Field.Store.NO));
if ((l & 3) == 0)
{
doc.Add(new FacetField("dim", "a"));
}
else
{
doc.Add(new FacetField("dim", "b"));
}
w.AddDocument(config.Build(tw, doc));
}
IndexReader r = w.GetReader();
var tr = new DirectoryTaxonomyReader(tw);
IndexSearcher s = NewSearcher(r);
if (VERBOSE)
{
Console.WriteLine("TEST: searcher=" + s);
}
DrillSideways ds = new DrillSidewaysAnonymousInnerClassHelper(this, s, config, tr);
// First search, no drill downs:
DrillDownQuery ddq = new DrillDownQuery(config);
DrillSidewaysResult dsr = ds.Search(null, ddq, 10);
Assert.AreEqual(100, dsr.Hits.TotalHits);
Assert.AreEqual("dim=dim path=[] value=100 childCount=2\n b (75)\n a (25)\n", dsr.Facets.GetTopChildren(10, "dim").ToString());
Assert.AreEqual("dim=field path=[] value=21 childCount=5\n less than 10 (10)\n less than or equal to 10 (11)\n over 90 (9)\n 90 or above (10)\n over 1000 (0)\n", dsr.Facets.GetTopChildren(10, "field").ToString());
// Second search, drill down on dim=b:
ddq = new DrillDownQuery(config);
ddq.Add("dim", "b");
dsr = ds.Search(null, ddq, 10);
Assert.AreEqual(75, dsr.Hits.TotalHits);
Assert.AreEqual("dim=dim path=[] value=100 childCount=2\n b (75)\n a (25)\n", dsr.Facets.GetTopChildren(10, "dim").ToString());
Assert.AreEqual("dim=field path=[] value=16 childCount=5\n less than 10 (7)\n less than or equal to 10 (8)\n over 90 (7)\n 90 or above (8)\n over 1000 (0)\n", dsr.Facets.GetTopChildren(10, "field").ToString());
// Third search, drill down on "less than or equal to 10":
ddq = new DrillDownQuery(config);
ddq.Add("field", NumericRangeQuery.NewInt64Range("field", 0L, 10L, true, true));
dsr = ds.Search(null, ddq, 10);
Assert.AreEqual(11, dsr.Hits.TotalHits);
Assert.AreEqual("dim=dim path=[] value=11 childCount=2\n b (8)\n a (3)\n", dsr.Facets.GetTopChildren(10, "dim").ToString());
Assert.AreEqual("dim=field path=[] value=21 childCount=5\n less than 10 (10)\n less than or equal to 10 (11)\n over 90 (9)\n 90 or above (10)\n over 1000 (0)\n", dsr.Facets.GetTopChildren(10, "field").ToString());
IOUtils.Dispose(tw, tr, td, w, r, d);
}
/// <summary>
/// Searches the Lucene index for Documents that match the specified search criteria.
/// </summary>
/// <param name="criteria">The search criteria.</param>
/// <returns></returns>
/// <exception cref="ArgumentNullException"></exception>
/// <exception cref="System.ArgumentNullException"></exception>
public SearchResult<Guid> Search(SearchCriteria criteria)
{
if (criteria == null)
throw new ArgumentNullException(nameof(criteria));
criteria.Query = String.IsNullOrWhiteSpace(criteria.Query) ? ALL_DOCS_QUERY : criteria.Query;
criteria.TopN = criteria.TopN > 0 ? criteria.TopN : SearchCriteria.DEFAULT_TOP_N;
criteria.ItemsPerPage = criteria.ItemsPerPage > 0 ? criteria.ItemsPerPage : SearchCriteria.DEFAULT_ITEMS_PER_PAGE;
criteria.PageNumber = criteria.PageNumber > 0 ? criteria.PageNumber : 1;
criteria.Validate();
var result = new SearchResult<Guid>(criteria);
var queryParser = new LuceneQueryParser(Schema.StandardField.FULL_TEXT, _compositeAnalyzer, Schema);
var query = queryParser.Parse(criteria.Query);
var instance = _searcherTaxonomyManager.Acquire() as SearcherTaxonomyManagerSearcherAndTaxonomy;
if (instance != null)
{
var searcher = instance.Searcher;
var taxonomyReader = instance.TaxonomyReader;
try
{
var sort = GetSortCriteria(criteria.SortByField);
var selectedFacets = criteria.SelectCategories.ToFacetFields();
var topDocs = (TopDocs)null;
var categories = (IEnumerable<Category>)null;
if (selectedFacets.Count() == 0)
{
// We are not going to do a drill-down on specific facets.
// Instead we will just take the top N facets from the matching Documents.
var facetsCollector = new FacetsCollector();
// Get the matching Documents
topDocs = FacetsCollector.Search(searcher, query, criteria.TopN, sort, facetsCollector);
// Get the Facet counts from the matching Documents
var facetCounts = new FastTaxonomyFacetCounts(taxonomyReader, _facetBuilder.FacetsConfig, facetsCollector);
categories = facetCounts.GetCategories(criteria.TopNCategories);
}
else
{
// Perform a drill-sideways query
var drillDownQuery = new DrillDownQuery(_facetBuilder.FacetsConfig, query);
foreach (var facetField in selectedFacets)
drillDownQuery.Add(facetField.Dim, facetField.Path);
var drillSideways = new DrillSideways(searcher, _facetBuilder.FacetsConfig, taxonomyReader);
var drillSidewaysResult = drillSideways.Search(drillDownQuery, null, null, criteria.TopN, sort, false, false);
// Get the matching documents
topDocs = drillSidewaysResult.Hits;
// Get the Facet counts from the matching Documents
categories = drillSidewaysResult.Facets.GetCategories(criteria.TopNCategories, selectedFacets);
}
// TODO: Don't pass TopDocs; pass an IEnumerable<Guid>
result.PopulateWith(topDocs, categories, id => searcher.Doc(id));
}
finally
{
_searcherTaxonomyManager.Release(instance);
searcher = null;
taxonomyReader = null;
}
}
return result;
}
/// <summary>
/// Search facet content in the existing index.
/// </summary>
/// <param name="text">The text to search for.</param>
/// <param name="indexFields">The array of index fields to search in.</param>
/// <param name="facetPaths">The array of facet paths to perform a drill down search on.</param>
/// <param name="numberToReturn">The maximum number of documents to return.</param>
/// <returns>The facet document.</returns>
/// <remarks>Use wildcard chars ('*', '?', '\'), logical ('AND', 'OR'), Quoted exact phrase ("search this").</remarks>
public Nequeo.Search.Engine.FacetDocument SearchFacetDocument(string text, FacetData.IndexField[] indexFields, FacetPath[] facetPaths, int numberToReturn = Int32.MaxValue)
{
Nequeo.Search.Engine.FacetDocument documents = new FacetDocument();
documents.TotalHits = 0;
try
{
// If text exists.
if (!String.IsNullOrEmpty(text))
{
// Load the searcher.
Lucene.Net.Search.IndexSearcher searcher = new Lucene.Net.Search.IndexSearcher(_reader);
string searchFieldName = "facetcontent";
Query query = null;
DrillDownQuery queryFacet = null;
TopDocs results = null;
// Build the facet configuration information.
FacetsConfig config = new FacetsConfig();
// Add the config.
foreach (FacetData.IndexField item in indexFields)
{
config.SetIndexFieldName(item.DimensionName, item.IndexFieldName);
}
// Get bytes
char[] textArray = text.ToCharArray();
// Search logical.
if (text.Contains("AND") || text.Contains("OR"))
{
// Create the query.
query = CreateLogicalQuery(text, searchFieldName);
}
else if (textArray[0].Equals('"') && textArray[textArray.Length - 1].Equals('"'))
{
// Create the query.
query = CreateQuotedQuery(new string(textArray, 1, textArray.Length - 2), searchFieldName);
}
else
{
// Create the query.
query = CreateBoolenQuery(text, BooleanClause.Occur.SHOULD, searchFieldName);
}
// Create the facet query.
queryFacet = new DrillDownQuery(config, query);
foreach (FacetPath facetPath in facetPaths)
{
// Add the path.
queryFacet.Add(facetPath.DimensionName, facetPath.Path);
}
// The collector.
FacetsCollector collector = new FacetsCollector();
// Search.
if (queryFacet != null)
{
results = FacetsCollector.Search(searcher, queryFacet, numberToReturn, collector);
}
// Get the total number of results that was asked for.
int totalResult = ((results.ScoreDocs != null && results.ScoreDocs.Length > 0) ? results.ScoreDocs.Length : 0);
// If result found.
if (results != null && results.TotalHits > 0)
{
List <TextDataResult> textDataResults = new List <TextDataResult>();
List <FileDocumentResult> fileDocResults = new List <FileDocumentResult>();
List <FacetPathResult> facetPathResults = new List <FacetPathResult>();
IDictionary <string, Facets> facetsMap = new Dictionary <string, Facets>();
// Add the facet count.
foreach (FacetData.IndexField item in indexFields)
{
// Add the facet for each index field.
facetsMap[item.DimensionName] = GetTaxonomyFacetCounts(_facetReader, config, collector, item.IndexFieldName);
}
// Create the multi facet list.
foreach (FacetPath facetPath in facetPaths)
{
try
{
// Add the facets.
Facets facets = facetsMap.First(u => u.Key.ToLower().Contains(facetPath.DimensionName.ToLower())).Value;
float number = facets.GetSpecificValue(facetPath.DimensionName, facetPath.Path);
// Add the path.
facetPathResults.Add(new FacetPathResult(facetPath.DimensionName, number, facetPath.Path));
}
catch { }
}
// For each document found.
for (int i = 0; i < totalResult; i++)
{
FileDocumentResult fileDocument = null;
TextDataResult textData = null;
int docID = results.ScoreDocs[i].Doc;
Lucene.Net.Documents.Document doc = searcher.Doc(docID);
try
{
// Get the data for each field.
IndexableField[] textNameFields = doc.GetFields("textname");
// If this field exists then text data.
if (textNameFields.Length > 0)
{
// Assign the data to the text document.
textData = new TextDataResult();
textData.Name = textNameFields.Length > 0 ? textNameFields[0].StringValue : null;
textData.Score = results.ScoreDocs[i].Score;
textData.Doc = docID;
// Do not know if the text was stored.
IndexableField[] textValueFields = doc.GetFields("textcomplete");
textData.Text = textValueFields.Length > 0 ? textValueFields[0].StringValue : null;
}
}
catch { }
// If text data exists then add.
if (textData != null)
{
textDataResults.Add(textData);
}
try
{
// Get the data for each field.
IndexableField[] pathNameFields = doc.GetFields("path");
IndexableField[] modifiedNameFields = doc.GetFields("modified");
// If this field exists then file document.
if (pathNameFields.Length > 0)
{
// Assign the data to the path document.
fileDocument = new FileDocumentResult();
fileDocument.Path = pathNameFields.Length > 0 ? pathNameFields[0].StringValue : null;
fileDocument.Modified = modifiedNameFields.Length > 0 ? modifiedNameFields[0].StringValue : null;
fileDocument.Score = results.ScoreDocs[i].Score;
fileDocument.Doc = docID;
}
}
catch { }
// If file data exists then add.
if (fileDocument != null)
{
fileDocResults.Add(fileDocument);
}
}
// Assign the facet document values.
documents.MaxScore = results.MaxScore;
documents.TotalHits = results.TotalHits;
documents.FacetPathResults = facetPathResults.ToArray();
documents.TextDataResults = textDataResults.ToArray();
documents.FileDocumentResults = fileDocResults.ToArray();
}
}
// Return the documents.
return(documents);
}
catch (Exception)
{
throw;
}
}
