| public DocFreq ( Lucene.Net.Index.Term term ) : int | ||
| term | Lucene.Net.Index.Term | |
| 리턴 | int | |
/// <summary> Check whether the word exists in the index.</summary>
/// <param name="word">String
/// </param>
/// <throws> IOException </throws>
/// <returns> true iff the word exists in the index
/// </returns>
public virtual bool Exist(System.String word)
{
// obtainSearcher calls ensureOpen
IndexSearcher indexSearcher = ObtainSearcher();
try
{
return(indexSearcher.DocFreq(F_WORD_TERM.CreateTerm(word)) > 0);
}
finally
{
ReleaseSearcher(indexSearcher);
}
}
public virtual void TestKnownSetOfDocuments()
{
System.String test1 = "eating chocolate in a computer lab"; //6 terms
System.String test2 = "computer in a computer lab"; //5 terms
System.String test3 = "a chocolate lab grows old"; //5 terms
System.String test4 = "eating chocolate with a chocolate lab in an old chocolate colored computer lab"; //13 terms
System.Collections.IDictionary test4Map = new System.Collections.Hashtable();
test4Map["chocolate"] = 3;
test4Map["lab"] = 2;
test4Map["eating"] = 1;
test4Map["computer"] = 1;
test4Map["with"] = 1;
test4Map["a"] = 1;
test4Map["colored"] = 1;
test4Map["in"] = 1;
test4Map["an"] = 1;
test4Map["computer"] = 1;
test4Map["old"] = 1;
Document testDoc1 = new Document();
SetupDoc(testDoc1, test1);
Document testDoc2 = new Document();
SetupDoc(testDoc2, test2);
Document testDoc3 = new Document();
SetupDoc(testDoc3, test3);
Document testDoc4 = new Document();
SetupDoc(testDoc4, test4);
Directory dir = new MockRAMDirectory();
try
{
IndexWriter writer = new IndexWriter(dir, new SimpleAnalyzer(), true, IndexWriter.MaxFieldLength.LIMITED);
Assert.IsTrue(writer != null);
writer.AddDocument(testDoc1);
writer.AddDocument(testDoc2);
writer.AddDocument(testDoc3);
writer.AddDocument(testDoc4);
writer.Close();
IndexSearcher knownSearcher = new IndexSearcher(dir, true);
TermEnum termEnum = knownSearcher.reader_ForNUnit.Terms();
TermDocs termDocs = knownSearcher.reader_ForNUnit.TermDocs();
//System.out.println("Terms: " + termEnum.size() + " Orig Len: " + termArray.length);
Similarity sim = knownSearcher.Similarity;
while (termEnum.Next() == true)
{
Term term = termEnum.Term;
//System.out.println("Term: " + term);
termDocs.Seek(term);
while (termDocs.Next())
{
int docId = termDocs.Doc;
int freq = termDocs.Freq;
//System.out.println("Doc Id: " + docId + " freq " + freq);
ITermFreqVector vector = knownSearcher.reader_ForNUnit.GetTermFreqVector(docId, "field");
float tf = sim.Tf(freq);
float idf = sim.Idf(knownSearcher.DocFreq(term), knownSearcher.MaxDoc);
//float qNorm = sim.queryNorm()
//This is fine since we don't have stop words
float lNorm = sim.LengthNorm("field", vector.GetTerms().Length);
//float coord = sim.coord()
//System.out.println("TF: " + tf + " IDF: " + idf + " LenNorm: " + lNorm);
Assert.IsTrue(vector != null);
System.String[] vTerms = vector.GetTerms();
int[] freqs = vector.GetTermFrequencies();
for (int i = 0; i < vTerms.Length; i++)
{
if (term.Text.Equals(vTerms[i]))
{
Assert.IsTrue(freqs[i] == freq);
}
}
}
//System.out.println("--------");
}
Query query = new TermQuery(new Term("field", "chocolate"));
ScoreDoc[] hits = knownSearcher.Search(query, null, 1000).ScoreDocs;
//doc 3 should be the first hit b/c it is the shortest match
Assert.IsTrue(hits.Length == 3);
float score = hits[0].Score;
/*System.out.println("Hit 0: " + hits.id(0) + " Score: " + hits.score(0) + " String: " + hits.doc(0).toString());
System.out.println("Explain: " + knownSearcher.explain(query, hits.id(0)));
System.out.println("Hit 1: " + hits.id(1) + " Score: " + hits.score(1) + " String: " + hits.doc(1).toString());
System.out.println("Explain: " + knownSearcher.explain(query, hits.id(1)));
System.out.println("Hit 2: " + hits.id(2) + " Score: " + hits.score(2) + " String: " + hits.doc(2).toString());
System.out.println("Explain: " + knownSearcher.explain(query, hits.id(2)));*/
Assert.IsTrue(hits[0].Doc == 2);
Assert.IsTrue(hits[1].Doc == 3);
Assert.IsTrue(hits[2].Doc == 0);
ITermFreqVector vector2 = knownSearcher.reader_ForNUnit.GetTermFreqVector(hits[1].Doc, "field");
Assert.IsTrue(vector2 != null);
//System.out.println("Vector: " + vector);
System.String[] terms = vector2.GetTerms();
int[] freqs2 = vector2.GetTermFrequencies();
Assert.IsTrue(terms != null && terms.Length == 10);
for (int i = 0; i < terms.Length; i++)
{
System.String term = terms[i];
//System.out.println("Term: " + term);
int freq = freqs2[i];
Assert.IsTrue(test4.IndexOf(term) != - 1);
System.Int32 freqInt = -1;
try
{
freqInt = (System.Int32) test4Map[term];
}
catch (Exception)
{
Assert.IsTrue(false);
}
Assert.IsTrue(freqInt == freq);
}
SortedTermVectorMapper mapper = new SortedTermVectorMapper(new TermVectorEntryFreqSortedComparator());
knownSearcher.reader_ForNUnit.GetTermFreqVector(hits[1].Doc, mapper);
var vectorEntrySet = mapper.TermVectorEntrySet;
Assert.IsTrue(vectorEntrySet.Count == 10, "mapper.getTermVectorEntrySet() Size: " + vectorEntrySet.Count + " is not: " + 10);
TermVectorEntry last = null;
foreach(TermVectorEntry tve in vectorEntrySet)
{
if (tve != null && last != null)
{
Assert.IsTrue(last.Frequency >= tve.Frequency, "terms are not properly sorted");
System.Int32 expectedFreq = (System.Int32) test4Map[tve.Term];
//we expect double the expectedFreq, since there are two fields with the exact same text and we are collapsing all fields
Assert.IsTrue(tve.Frequency == 2 * expectedFreq, "Frequency is not correct:");
}
last = tve;
}
FieldSortedTermVectorMapper fieldMapper = new FieldSortedTermVectorMapper(new TermVectorEntryFreqSortedComparator());
knownSearcher.reader_ForNUnit.GetTermFreqVector(hits[1].Doc, fieldMapper);
var map = fieldMapper.FieldToTerms;
Assert.IsTrue(map.Count == 2, "map Size: " + map.Count + " is not: " + 2);
vectorEntrySet = map["field"];
Assert.IsTrue(vectorEntrySet != null, "vectorEntrySet is null and it shouldn't be");
Assert.IsTrue(vectorEntrySet.Count == 10, "vectorEntrySet Size: " + vectorEntrySet.Count + " is not: " + 10);
knownSearcher.Close();
}
catch (System.IO.IOException e)
{
System.Console.Error.WriteLine(e.StackTrace);
Assert.IsTrue(false);
}
}
public virtual void TestKnownSetOfDocuments()
{
System.String test1 = "eating chocolate in a computer lab"; //6 terms
System.String test2 = "computer in a computer lab"; //5 terms
System.String test3 = "a chocolate lab grows old"; //5 terms
System.String test4 = "eating chocolate with a chocolate lab in an old chocolate colored computer lab"; //13 terms
System.Collections.IDictionary test4Map = new System.Collections.Hashtable();
test4Map["chocolate"] = 3;
test4Map["lab"] = 2;
test4Map["eating"] = 1;
test4Map["computer"] = 1;
test4Map["with"] = 1;
test4Map["a"] = 1;
test4Map["colored"] = 1;
test4Map["in"] = 1;
test4Map["an"] = 1;
test4Map["computer"] = 1;
test4Map["old"] = 1;
Document testDoc1 = new Document();
SetupDoc(testDoc1, test1);
Document testDoc2 = new Document();
SetupDoc(testDoc2, test2);
Document testDoc3 = new Document();
SetupDoc(testDoc3, test3);
Document testDoc4 = new Document();
SetupDoc(testDoc4, test4);
Directory dir = new MockRAMDirectory();
try
{
IndexWriter writer = new IndexWriter(dir, new SimpleAnalyzer(), true, IndexWriter.MaxFieldLength.LIMITED);
Assert.IsTrue(writer != null);
writer.AddDocument(testDoc1);
writer.AddDocument(testDoc2);
writer.AddDocument(testDoc3);
writer.AddDocument(testDoc4);
writer.Close();
IndexSearcher knownSearcher = new IndexSearcher(dir, true);
TermEnum termEnum = knownSearcher.reader_ForNUnit.Terms();
TermDocs termDocs = knownSearcher.reader_ForNUnit.TermDocs();
//System.out.println("Terms: " + termEnum.size() + " Orig Len: " + termArray.length);
Similarity sim = knownSearcher.Similarity;
while (termEnum.Next() == true)
{
Term term = termEnum.Term;
//System.out.println("Term: " + term);
termDocs.Seek(term);
while (termDocs.Next())
{
int docId = termDocs.Doc;
int freq = termDocs.Freq;
//System.out.println("Doc Id: " + docId + " freq " + freq);
ITermFreqVector vector = knownSearcher.reader_ForNUnit.GetTermFreqVector(docId, "field");
float tf = sim.Tf(freq);
float idf = sim.Idf(knownSearcher.DocFreq(term), knownSearcher.MaxDoc);
//float qNorm = sim.queryNorm()
//This is fine since we don't have stop words
float lNorm = sim.LengthNorm("field", vector.GetTerms().Length);
//float coord = sim.coord()
//System.out.println("TF: " + tf + " IDF: " + idf + " LenNorm: " + lNorm);
Assert.IsTrue(vector != null);
System.String[] vTerms = vector.GetTerms();
int[] freqs = vector.GetTermFrequencies();
for (int i = 0; i < vTerms.Length; i++)
{
if (term.Text.Equals(vTerms[i]))
{
Assert.IsTrue(freqs[i] == freq);
}
}
}
//System.out.println("--------");
}
Query query = new TermQuery(new Term("field", "chocolate"));
ScoreDoc[] hits = knownSearcher.Search(query, null, 1000).ScoreDocs;
//doc 3 should be the first hit b/c it is the shortest match
Assert.IsTrue(hits.Length == 3);
float score = hits[0].Score;
/*System.out.println("Hit 0: " + hits.id(0) + " Score: " + hits.score(0) + " String: " + hits.doc(0).toString());
* System.out.println("Explain: " + knownSearcher.explain(query, hits.id(0)));
* System.out.println("Hit 1: " + hits.id(1) + " Score: " + hits.score(1) + " String: " + hits.doc(1).toString());
* System.out.println("Explain: " + knownSearcher.explain(query, hits.id(1)));
* System.out.println("Hit 2: " + hits.id(2) + " Score: " + hits.score(2) + " String: " + hits.doc(2).toString());
* System.out.println("Explain: " + knownSearcher.explain(query, hits.id(2)));*/
Assert.IsTrue(hits[0].Doc == 2);
Assert.IsTrue(hits[1].Doc == 3);
Assert.IsTrue(hits[2].Doc == 0);
ITermFreqVector vector2 = knownSearcher.reader_ForNUnit.GetTermFreqVector(hits[1].Doc, "field");
Assert.IsTrue(vector2 != null);
//System.out.println("Vector: " + vector);
System.String[] terms = vector2.GetTerms();
int[] freqs2 = vector2.GetTermFrequencies();
Assert.IsTrue(terms != null && terms.Length == 10);
for (int i = 0; i < terms.Length; i++)
{
System.String term = terms[i];
//System.out.println("Term: " + term);
int freq = freqs2[i];
Assert.IsTrue(test4.IndexOf(term) != -1);
System.Int32 freqInt = -1;
try
{
freqInt = (System.Int32)test4Map[term];
}
catch (Exception)
{
Assert.IsTrue(false);
}
Assert.IsTrue(freqInt == freq);
}
SortedTermVectorMapper mapper = new SortedTermVectorMapper(new TermVectorEntryFreqSortedComparator());
knownSearcher.reader_ForNUnit.GetTermFreqVector(hits[1].Doc, mapper);
var vectorEntrySet = mapper.TermVectorEntrySet;
Assert.IsTrue(vectorEntrySet.Count == 10, "mapper.getTermVectorEntrySet() Size: " + vectorEntrySet.Count + " is not: " + 10);
TermVectorEntry last = null;
foreach (TermVectorEntry tve in vectorEntrySet)
{
if (tve != null && last != null)
{
Assert.IsTrue(last.Frequency >= tve.Frequency, "terms are not properly sorted");
System.Int32 expectedFreq = (System.Int32)test4Map[tve.Term];
//we expect double the expectedFreq, since there are two fields with the exact same text and we are collapsing all fields
Assert.IsTrue(tve.Frequency == 2 * expectedFreq, "Frequency is not correct:");
}
last = tve;
}
FieldSortedTermVectorMapper fieldMapper = new FieldSortedTermVectorMapper(new TermVectorEntryFreqSortedComparator());
knownSearcher.reader_ForNUnit.GetTermFreqVector(hits[1].Doc, fieldMapper);
var map = fieldMapper.FieldToTerms;
Assert.IsTrue(map.Count == 2, "map Size: " + map.Count + " is not: " + 2);
vectorEntrySet = map["field"];
Assert.IsTrue(vectorEntrySet != null, "vectorEntrySet is null and it shouldn't be");
Assert.IsTrue(vectorEntrySet.Count == 10, "vectorEntrySet Size: " + vectorEntrySet.Count + " is not: " + 10);
knownSearcher.Close();
}
catch (System.IO.IOException e)
{
System.Console.Error.WriteLine(e.StackTrace);
Assert.IsTrue(false);
}
}
