public virtual void TestMaxPosition3WithSynomyms()
{
foreach (bool consumeAll in new bool[] { true, false })
{
MockTokenizer tokenizer = new MockTokenizer(new StringReader("one two three four five"), MockTokenizer.WHITESPACE, false);
// if we are consuming all tokens, we can use the checks, otherwise we can't
tokenizer.EnableChecks = consumeAll;
SynonymMap.Builder builder = new SynonymMap.Builder(true);
builder.Add(new CharsRef("one"), new CharsRef("first"), true);
builder.Add(new CharsRef("one"), new CharsRef("alpha"), true);
builder.Add(new CharsRef("one"), new CharsRef("beguine"), true);
CharsRef multiWordCharsRef = new CharsRef();
SynonymMap.Builder.Join(new string[] { "and", "indubitably", "single", "only" }, multiWordCharsRef);
builder.Add(new CharsRef("one"), multiWordCharsRef, true);
SynonymMap.Builder.Join(new string[] { "dopple", "ganger" }, multiWordCharsRef);
builder.Add(new CharsRef("two"), multiWordCharsRef, true);
SynonymMap synonymMap = builder.Build();
TokenStream stream = new SynonymFilter(tokenizer, synonymMap, true);
stream = new LimitTokenPositionFilter(stream, 3, consumeAll);
// "only", the 4th word of multi-word synonym "and indubitably single only" is not emitted, since its position is greater than 3.
AssertTokenStreamContents(stream, new string[] { "one", "first", "alpha", "beguine", "and", "two", "indubitably", "dopple", "three", "single", "ganger" }, new int[] { 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0 });
}
}
/// <summary>
/// blast some random strings through the analyzer </summary>
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testRandomStrings() throws Exception
public virtual void testRandomStrings()
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int numIters = atLeast(10);
int numIters = atLeast(10);
for (int i = 0; i < numIters; i++)
{
SynonymMap.Builder b = new SynonymMap.Builder(random().nextBoolean());
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int numEntries = atLeast(10);
int numEntries = atLeast(10);
for (int j = 0; j < numEntries; j++)
{
add(b, randomNonEmptyString(), randomNonEmptyString(), random().nextBoolean());
}
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.analysis.synonym.SynonymMap map = b.build();
SynonymMap map = b.build();
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final boolean ignoreCase = random().nextBoolean();
bool ignoreCase = random().nextBoolean();
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.analysis.Analyzer analyzer = new org.apache.lucene.analysis.Analyzer()
Analyzer analyzer = new AnalyzerAnonymousInnerClassHelper(this, map, ignoreCase);
checkRandomData(random(), analyzer, 200);
}
}
public virtual void TestRandomStrings()
{
int numIters = AtLeast(10);
for (int i = 0; i < numIters; i++)
{
SynonymMap.Builder b = new SynonymMap.Builder(Random.nextBoolean());
int numEntries = AtLeast(10);
for (int j = 0; j < numEntries; j++)
{
Add(b, RandomNonEmptyString(), RandomNonEmptyString(), Random.nextBoolean());
}
SynonymMap map = b.Build();
bool ignoreCase = Random.nextBoolean();
Analyzer analyzer = Analyzer.NewAnonymous(createComponents: (fieldName, reader) =>
{
Tokenizer tokenizer = new MockTokenizer(reader, MockTokenizer.SIMPLE, true);
TokenStream stream = new SynonymFilter(tokenizer, map, ignoreCase);
return(new TokenStreamComponents(tokenizer, new RemoveDuplicatesTokenFilter(stream)));
});
CheckRandomData(Random, analyzer, 200);
}
}
public object Create(Random random)
{
SynonymMap.Builder b = new SynonymMap.Builder(random.nextBoolean());
int numEntries = AtLeast(10);
for (int j = 0; j < numEntries; j++)
{
AddSyn(b, RandomNonEmptyString(random), RandomNonEmptyString(random), random.nextBoolean());
}
try
{
return(b.Build());
}
catch (Exception ex)
{
throw ex;
return(null); // unreachable code
}
}
public virtual void TestRandomStrings()
{
int numIters = AtLeast(10);
for (int i = 0; i < numIters; i++)
{
SynonymMap.Builder b = new SynonymMap.Builder(Random.nextBoolean());
int numEntries = AtLeast(10);
for (int j = 0; j < numEntries; j++)
{
Add(b, RandomNonEmptyString(), RandomNonEmptyString(), Random.nextBoolean());
}
SynonymMap map = b.Build();
bool ignoreCase = Random.nextBoolean();
Analyzer analyzer = new AnalyzerAnonymousInnerClassHelper(this, map, ignoreCase);
CheckRandomData(Random, analyzer, 200);
}
}
public object Create(Random random)
{
SynonymMap.Builder b = new SynonymMap.Builder(random.nextBoolean());
int numEntries = AtLeast(10);
for (int j = 0; j < numEntries; j++)
{
AddSyn(b, RandomNonEmptyString(random), RandomNonEmptyString(random), random.nextBoolean());
}
try
{
return(b.Build());
}
catch (Exception /*ex*/)
{
throw; // LUCENENET: CA2200: Rethrow to preserve stack details (https://docs.microsoft.com/en-us/visualstudio/code-quality/ca2200-rethrow-to-preserve-stack-details)
#pragma warning disable 162
return(null); // unreachable code
#pragma warning restore 162
}
}
private void Add(SynonymMap.Builder b, string input, string output, bool keepOrig)
{
b.Add(new CharsRef(Regex.Replace(input, " +", "\u0000")), new CharsRef(Regex.Replace(output, " +", "\u0000")), keepOrig);
}
private void add(SynonymMap.Builder b, string input, string output, bool keepOrig)
{
b.add(new CharsRef(input.replaceAll(" +", "\u0000")), new CharsRef(output.replaceAll(" +", "\u0000")), keepOrig);
}