public override bool accept(AttributeSource source)
{
if (termAtt == null)
{
termAtt = source.addAttribute(typeof(CharTermAttribute));
}
try
{
DateTime date = dateFormat.parse(termAtt.ToString()); //We don't care about the date, just that we can parse it as a date
if (date != null)
{
return true;
}
}
catch (ParseException)
{
}
return false;
}
/// <summary>
/// Fills CharTermAttribute with the current token text.
/// </summary>
public void getText(CharTermAttribute t)
{
t.copyBuffer(zzBuffer, zzStartRead, zzMarkedPos - zzStartRead);
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testRandom() throws Exception
public virtual void testRandom()
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int alphabetSize = org.apache.lucene.util.TestUtil.nextInt(random(), 2, 7);
int alphabetSize = TestUtil.Next(random(), 2, 7);
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int docLen = atLeast(3000);
int docLen = atLeast(3000);
//final int docLen = 50;
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final String document = getRandomString('a', alphabetSize, docLen);
string document = getRandomString('a', alphabetSize, docLen);
if (VERBOSE)
{
Console.WriteLine("TEST: doc=" + document);
}
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int numSyn = atLeast(5);
int numSyn = atLeast(5);
//final int numSyn = 2;
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final java.util.Map<String,OneSyn> synMap = new java.util.HashMap<>();
IDictionary<string, OneSyn> synMap = new Dictionary<string, OneSyn>();
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final java.util.List<OneSyn> syns = new java.util.ArrayList<>();
IList<OneSyn> syns = new List<OneSyn>();
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final boolean dedup = random().nextBoolean();
bool dedup = random().nextBoolean();
if (VERBOSE)
{
Console.WriteLine(" dedup=" + dedup);
}
b = new SynonymMap.Builder(dedup);
for (int synIDX = 0;synIDX < numSyn;synIDX++)
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final String synIn = getRandomString('a', alphabetSize, org.apache.lucene.util.TestUtil.nextInt(random(), 1, 5)).trim();
string synIn = getRandomString('a', alphabetSize, TestUtil.Next(random(), 1, 5)).Trim();
OneSyn s = synMap[synIn];
if (s == null)
{
s = new OneSyn();
s.@in = synIn;
syns.Add(s);
s.@out = new List<>();
synMap[synIn] = s;
s.keepOrig = random().nextBoolean();
}
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final String synOut = getRandomString('0', 10, org.apache.lucene.util.TestUtil.nextInt(random(), 1, 5)).trim();
string synOut = getRandomString('0', 10, TestUtil.Next(random(), 1, 5)).Trim();
[email protected](synOut);
add(synIn, synOut, s.keepOrig);
if (VERBOSE)
{
Console.WriteLine(" syns[" + synIDX + "] = " + s.@in + " -> " + s.@out + " keepOrig=" + s.keepOrig);
}
}
tokensIn = new MockTokenizer(new StringReader("a"), MockTokenizer.WHITESPACE, true);
tokensIn.reset();
assertTrue(tokensIn.incrementToken());
assertFalse(tokensIn.incrementToken());
tokensIn.end();
tokensIn.close();
tokensOut = new SynonymFilter(tokensIn, b.build(), true);
termAtt = tokensOut.addAttribute(typeof(CharTermAttribute));
posIncrAtt = tokensOut.addAttribute(typeof(PositionIncrementAttribute));
posLenAtt = tokensOut.addAttribute(typeof(PositionLengthAttribute));
offsetAtt = tokensOut.addAttribute(typeof(OffsetAttribute));
if (dedup)
{
pruneDups(syns);
}
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final String expected = slowSynMatcher(document, syns, 5);
string expected = slowSynMatcher(document, syns, 5);
if (VERBOSE)
{
Console.WriteLine("TEST: expected=" + expected);
}
verify(document, expected);
}
private void assertEquals(CharTermAttribute term, string expected)
{
assertEquals(expected.Length, term.length());
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final char[] buffer = term.buffer();
char[] buffer = term.buffer();
for (int chIDX = 0;chIDX < expected.Length;chIDX++)
{
assertEquals(expected[chIDX], buffer[chIDX]);
}
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testBasic2() throws Exception
public virtual void testBasic2()
{
b = new SynonymMap.Builder(true);
const bool keepOrig = false;
add("aaa", "aaaa1 aaaa2 aaaa3", keepOrig);
add("bbb", "bbbb1 bbbb2", keepOrig);
tokensIn = new MockTokenizer(new StringReader("a"), MockTokenizer.WHITESPACE, true);
tokensIn.reset();
assertTrue(tokensIn.incrementToken());
assertFalse(tokensIn.incrementToken());
tokensIn.end();
tokensIn.close();
tokensOut = new SynonymFilter(tokensIn, b.build(), true);
termAtt = tokensOut.addAttribute(typeof(CharTermAttribute));
posIncrAtt = tokensOut.addAttribute(typeof(PositionIncrementAttribute));
posLenAtt = tokensOut.addAttribute(typeof(PositionLengthAttribute));
offsetAtt = tokensOut.addAttribute(typeof(OffsetAttribute));
if (keepOrig)
{
verify("xyzzy bbb pot of gold", "xyzzy bbb/bbbb1 pot/bbbb2 of gold");
verify("xyzzy aaa pot of gold", "xyzzy aaa/aaaa1 pot/aaaa2 of/aaaa3 gold");
}
else
{
verify("xyzzy bbb pot of gold", "xyzzy bbbb1 pot/bbbb2 of gold");
verify("xyzzy aaa pot of gold", "xyzzy aaaa1 pot/aaaa2 of/aaaa3 gold");
}
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testOutputHangsOffEnd() throws Exception
public virtual void testOutputHangsOffEnd()
{
b = new SynonymMap.Builder(true);
const bool keepOrig = false;
// b hangs off the end (no input token under it):
add("a", "a b", keepOrig);
tokensIn = new MockTokenizer(new StringReader("a"), MockTokenizer.WHITESPACE, true);
tokensIn.reset();
assertTrue(tokensIn.incrementToken());
assertFalse(tokensIn.incrementToken());
tokensIn.end();
tokensIn.close();
tokensOut = new SynonymFilter(tokensIn, b.build(), true);
termAtt = tokensOut.addAttribute(typeof(CharTermAttribute));
posIncrAtt = tokensOut.addAttribute(typeof(PositionIncrementAttribute));
offsetAtt = tokensOut.addAttribute(typeof(OffsetAttribute));
posLenAtt = tokensOut.addAttribute(typeof(PositionLengthAttribute));
// Make sure endOffset inherits from previous input token:
verify("a", "a b:1");
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testBasic() throws Exception
public virtual void testBasic()
{
b = new SynonymMap.Builder(true);
add("a", "foo", true);
add("a b", "bar fee", true);
add("b c", "dog collar", true);
add("c d", "dog harness holder extras", true);
add("m c e", "dog barks loudly", false);
add("i j k", "feep", true);
add("e f", "foo bar", false);
add("e f", "baz bee", false);
add("z", "boo", false);
add("y", "bee", true);
tokensIn = new MockTokenizer(new StringReader("a"), MockTokenizer.WHITESPACE, true);
tokensIn.reset();
assertTrue(tokensIn.incrementToken());
assertFalse(tokensIn.incrementToken());
tokensIn.end();
tokensIn.close();
tokensOut = new SynonymFilter(tokensIn, b.build(), true);
termAtt = tokensOut.addAttribute(typeof(CharTermAttribute));
posIncrAtt = tokensOut.addAttribute(typeof(PositionIncrementAttribute));
posLenAtt = tokensOut.addAttribute(typeof(PositionLengthAttribute));
offsetAtt = tokensOut.addAttribute(typeof(OffsetAttribute));
verify("a b c", "a/bar b/fee c");
// syn output extends beyond input tokens
verify("x a b c d", "x a/bar b/fee c/dog d/harness holder extras");
verify("a b a", "a/bar b/fee a/foo");
// outputs that add to one another:
verify("c d c d", "c/dog d/harness c/holder/dog d/extras/harness holder extras");
// two outputs for same input
verify("e f", "foo/baz bar/bee");
// verify multi-word / single-output offsets:
verify("g i j k g", "g i/feep:7_3 j k g");
// mixed keepOrig true/false:
verify("a m c e x", "a/foo dog barks loudly x");
verify("c d m c e x", "c/dog d/harness holder/dog extras/barks loudly x");
assertTrue(tokensOut.CaptureCount > 0);
// no captureStates when no syns matched
verify("p q r s t", "p q r s t");
assertEquals(0, tokensOut.CaptureCount);
// no captureStates when only single-input syns, w/ no
// lookahead needed, matched
verify("p q z y t", "p q boo y/bee t");
assertEquals(0, tokensOut.CaptureCount);
}
public InputWindowToken(ShingleFilter outerInstance, AttributeSource attSource)
{
this.outerInstance = outerInstance;
this.attSource = attSource;
this.termAtt = attSource.getAttribute(typeof(CharTermAttribute));
this.offsetAtt = attSource.getAttribute(typeof(OffsetAttribute));
}
/*
* Need to worry about multiple scenarios:
* - need to go for the longest match
* a b => foo #shouldn't match if "a b" is followed by "c d"
* a b c d => bar
* - need to backtrack - retry matches for tokens already read
* a b c d => foo
* b c => bar
* If the input stream is "a b c x", one will consume "a b c d"
* trying to match the first rule... all but "a" should be
* pushed back so a match may be made on "b c".
* - don't try and match generated tokens (thus need separate queue)
* matching is not recursive.
* - handle optional generation of original tokens in all these cases,
* merging token streams to preserve token positions.
* - preserve original positionIncrement of first matched token
*/
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: @Override public boolean incrementToken() throws java.io.IOException
public override bool incrementToken()
{
while (true)
{
// if there are any generated tokens, return them... don't try any
// matches against them, as we specifically don't want recursion.
//JAVA TO C# CONVERTER TODO TASK: Java iterators are only converted within the context of 'while' and 'for' loops:
if (replacement != null && replacement.hasNext())
{
//JAVA TO C# CONVERTER TODO TASK: Java iterators are only converted within the context of 'while' and 'for' loops:
copy(this, replacement.next());
return(true);
}
// common case fast-path of first token not matching anything
AttributeSource firstTok = nextTok();
if (firstTok == null)
{
return(false);
}
CharTermAttribute termAtt = firstTok.addAttribute(typeof(CharTermAttribute));
SlowSynonymMap result = map.submap != null?map.submap.get(termAtt.buffer(), 0, termAtt.length()) : null;
if (result == null)
{
copy(this, firstTok);
return(true);
}
// fast-path failed, clone ourselves if needed
if (firstTok == this)
{
firstTok = cloneAttributes();
}
// OK, we matched a token, so find the longest match.
matched = new LinkedList <>();
result = match(result);
if (result == null)
{
// no match, simply return the first token read.
copy(this, firstTok);
return(true);
}
// reuse, or create new one each time?
List <AttributeSource> generated = new List <AttributeSource>(result.synonyms.Length + matched.Count + 1);
//
// there was a match... let's generate the new tokens, merging
// in the matched tokens (position increments need adjusting)
//
AttributeSource lastTok = matched.Count == 0 ? firstTok : matched.Last.Value;
bool includeOrig = result.includeOrig();
AttributeSource origTok = includeOrig ? firstTok : null;
PositionIncrementAttribute firstPosIncAtt = firstTok.addAttribute(typeof(PositionIncrementAttribute));
int origPos = firstPosIncAtt.PositionIncrement; // position of origTok in the original stream
int repPos = 0; // curr position in replacement token stream
int pos = 0; // current position in merged token stream
for (int i = 0; i < result.synonyms.Length; i++)
{
Token repTok = result.synonyms[i];
AttributeSource newTok = firstTok.cloneAttributes();
CharTermAttribute newTermAtt = newTok.addAttribute(typeof(CharTermAttribute));
OffsetAttribute newOffsetAtt = newTok.addAttribute(typeof(OffsetAttribute));
PositionIncrementAttribute newPosIncAtt = newTok.addAttribute(typeof(PositionIncrementAttribute));
OffsetAttribute lastOffsetAtt = lastTok.addAttribute(typeof(OffsetAttribute));
newOffsetAtt.setOffset(newOffsetAtt.startOffset(), lastOffsetAtt.endOffset());
newTermAtt.copyBuffer(repTok.buffer(), 0, repTok.length());
repPos += repTok.PositionIncrement;
if (i == 0) // make position of first token equal to original
{
repPos = origPos;
}
// if necessary, insert original tokens and adjust position increment
while (origTok != null && origPos <= repPos)
{
PositionIncrementAttribute origPosInc = origTok.addAttribute(typeof(PositionIncrementAttribute));
origPosInc.PositionIncrement = origPos - pos;
generated.Add(origTok);
pos += origPosInc.PositionIncrement;
origTok = matched.Count == 0 ? null : matched.RemoveFirst();
if (origTok != null)
{
origPosInc = origTok.addAttribute(typeof(PositionIncrementAttribute));
origPos += origPosInc.PositionIncrement;
}
}
newPosIncAtt.PositionIncrement = repPos - pos;
generated.Add(newTok);
pos += newPosIncAtt.PositionIncrement;
}
// finish up any leftover original tokens
while (origTok != null)
{
PositionIncrementAttribute origPosInc = origTok.addAttribute(typeof(PositionIncrementAttribute));
origPosInc.PositionIncrement = origPos - pos;
generated.Add(origTok);
pos += origPosInc.PositionIncrement;
origTok = matched.Count == 0 ? null : matched.RemoveFirst();
if (origTok != null)
{
origPosInc = origTok.addAttribute(typeof(PositionIncrementAttribute));
origPos += origPosInc.PositionIncrement;
}
}
// what if we replaced a longer sequence with a shorter one?
// a/0 b/5 => foo/0
// should I re-create the gap on the next buffered token?
replacement = generated.GetEnumerator();
// Now return to the top of the loop to read and return the first
// generated token.. The reason this is done is that we may have generated
// nothing at all, and may need to continue with more matching logic.
}
}
public override bool accept(AttributeSource a)
{
CharTermAttribute termAtt = a.getAttribute(typeof(CharTermAttribute));
return(termAtt.ToString().Equals("Dogs", StringComparison.CurrentCultureIgnoreCase));
}
/// <summary>
/// Not an explicit test, just useful to print out some info on performance
/// </summary>
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void performance() throws Exception
public virtual void performance()
{
int[] tokCount = new int[] { 100, 500, 1000, 2000, 5000, 10000 };
int[] modCounts = new int[] { 1, 2, 5, 10, 20, 50, 100, 200, 500 };
for (int k = 0; k < tokCount.Length; k++)
{
StringBuilder buffer = new StringBuilder();
Console.WriteLine("-----Tokens: " + tokCount[k] + "-----");
for (int i = 0; i < tokCount[k]; i++)
{
buffer.Append(English.intToEnglish(i).toUpperCase(Locale.ROOT)).Append(' ');
}
//make sure we produce the same tokens
TeeSinkTokenFilter teeStream = new TeeSinkTokenFilter(new StandardFilter(TEST_VERSION_CURRENT, new StandardTokenizer(TEST_VERSION_CURRENT, new StringReader(buffer.ToString()))));
TokenStream sink = teeStream.newSinkTokenStream(new ModuloSinkFilter(this, 100));
teeStream.consumeAllTokens();
TokenStream stream = new ModuloTokenFilter(this, new StandardFilter(TEST_VERSION_CURRENT, new StandardTokenizer(TEST_VERSION_CURRENT, new StringReader(buffer.ToString()))), 100);
CharTermAttribute tfTok = stream.addAttribute(typeof(CharTermAttribute));
CharTermAttribute sinkTok = sink.addAttribute(typeof(CharTermAttribute));
for (int i = 0; stream.incrementToken(); i++)
{
assertTrue(sink.incrementToken());
assertTrue(tfTok + " is not equal to " + sinkTok + " at token: " + i, tfTok.Equals(sinkTok) == true);
}
//simulate two fields, each being analyzed once, for 20 documents
for (int j = 0; j < modCounts.Length; j++)
{
int tfPos = 0;
long start = DateTimeHelperClass.CurrentUnixTimeMillis();
for (int i = 0; i < 20; i++)
{
stream = new StandardFilter(TEST_VERSION_CURRENT, new StandardTokenizer(TEST_VERSION_CURRENT, new StringReader(buffer.ToString())));
PositionIncrementAttribute posIncrAtt = stream.getAttribute(typeof(PositionIncrementAttribute));
while (stream.incrementToken())
{
tfPos += posIncrAtt.PositionIncrement;
}
stream = new ModuloTokenFilter(this, new StandardFilter(TEST_VERSION_CURRENT, new StandardTokenizer(TEST_VERSION_CURRENT, new StringReader(buffer.ToString()))), modCounts[j]);
posIncrAtt = stream.getAttribute(typeof(PositionIncrementAttribute));
while (stream.incrementToken())
{
tfPos += posIncrAtt.PositionIncrement;
}
}
long finish = DateTimeHelperClass.CurrentUnixTimeMillis();
Console.WriteLine("ModCount: " + modCounts[j] + " Two fields took " + (finish - start) + " ms");
int sinkPos = 0;
//simulate one field with one sink
start = DateTimeHelperClass.CurrentUnixTimeMillis();
for (int i = 0; i < 20; i++)
{
teeStream = new TeeSinkTokenFilter(new StandardFilter(TEST_VERSION_CURRENT, new StandardTokenizer(TEST_VERSION_CURRENT, new StringReader(buffer.ToString()))));
sink = teeStream.newSinkTokenStream(new ModuloSinkFilter(this, modCounts[j]));
PositionIncrementAttribute posIncrAtt = teeStream.getAttribute(typeof(PositionIncrementAttribute));
while (teeStream.incrementToken())
{
sinkPos += posIncrAtt.PositionIncrement;
}
//System.out.println("Modulo--------");
posIncrAtt = sink.getAttribute(typeof(PositionIncrementAttribute));
while (sink.incrementToken())
{
sinkPos += posIncrAtt.PositionIncrement;
}
}
finish = DateTimeHelperClass.CurrentUnixTimeMillis();
Console.WriteLine("ModCount: " + modCounts[j] + " Tee fields took " + (finish - start) + " ms");
assertTrue(sinkPos + " does not equal: " + tfPos, sinkPos == tfPos);
}
Console.WriteLine("- End Tokens: " + tokCount[k] + "-----");
}
}
