//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testEmptyTerm() throws Exception
public virtual void testEmptyTerm()
{
Random random = random();
Analyzer a = new AnalyzerAnonymousInnerClassHelper4(this);
checkAnalysisConsistency(random, a, random.nextBoolean(), "");
Analyzer b = new AnalyzerAnonymousInnerClassHelper5(this);
checkAnalysisConsistency(random, b, random.nextBoolean(), "");
}
public virtual void TestEmptyTerm()
{
CharArraySet dict = makeDictionary("a", "e", "i", "o", "u", "y", "bc", "def");
Analyzer a = new AnalyzerAnonymousInnerClassHelper4(this, dict);
CheckOneTerm(a, "", "");
//InputSource @is = new InputSource(this.GetType().getResource("da_UTF8.xml").toExternalForm());
using (var @is = this.GetType().getResourceAsStream("da_UTF8.xml"))
{
HyphenationTree hyphenator = HyphenationCompoundWordTokenFilter.GetHyphenationTree(@is);
Analyzer b = new AnalyzerAnonymousInnerClassHelper5(this, hyphenator);
CheckOneTerm(b, "", "");
}
}
public virtual void TestLotsOfConcatenating2()
{
WordDelimiterFlags flags = WordDelimiterFlags.PRESERVE_ORIGINAL
| WordDelimiterFlags.GENERATE_WORD_PARTS
| WordDelimiterFlags.GENERATE_NUMBER_PARTS
| WordDelimiterFlags.CATENATE_WORDS
| WordDelimiterFlags.CATENATE_NUMBERS
| WordDelimiterFlags.CATENATE_ALL
| WordDelimiterFlags.SPLIT_ON_CASE_CHANGE
| WordDelimiterFlags.SPLIT_ON_NUMERICS
| WordDelimiterFlags.STEM_ENGLISH_POSSESSIVE;
/* analyzer that uses whitespace + wdf */
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this, flags);
AssertAnalyzesTo(a, "abc-def-123-456", new string[] { "abc-def-123-456", "abc", "abcdef", "abcdef123456", "def", "123", "123456", "456" }, new int[] { 0, 0, 0, 0, 4, 8, 8, 12 }, new int[] { 15, 3, 7, 15, 7, 11, 15, 15 }, new int[] { 1, 0, 0, 0, 1, 1, 0, 1 });
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testEmptyTerm() throws Exception
public virtual void testEmptyTerm()
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.analysis.util.CharArraySet dict = makeDictionary("a", "e", "i", "o", "u", "y", "bc", "def");
CharArraySet dict = makeDictionary("a", "e", "i", "o", "u", "y", "bc", "def");
Analyzer a = new AnalyzerAnonymousInnerClassHelper4(this, dict);
checkOneTerm(a, "", "");
InputSource @is = new InputSource(this.GetType().getResource("da_UTF8.xml").toExternalForm());
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.analysis.compound.hyphenation.HyphenationTree hyphenator = HyphenationCompoundWordTokenFilter.getHyphenationTree(is);
HyphenationTree hyphenator = HyphenationCompoundWordTokenFilter.getHyphenationTree(@is);
Analyzer b = new AnalyzerAnonymousInnerClassHelper5(this, hyphenator);
checkOneTerm(b, "", "");
}
public virtual void TestMockGraphTokenFilterBeforeHolesRandom()
{
for (int iter = 0; iter < 10 * RANDOM_MULTIPLIER; iter++)
{
if (VERBOSE)
{
Console.WriteLine("\nTEST: iter=" + iter);
}
// Make new analyzer each time, because MGTF has fixed
// seed:
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this);
Random random = Random;
CheckRandomData(random, a, 5, AtLeast(100));
}
}
public virtual void TestVanishingTerms()
{
string testFile = "aaa => aaaa1 aaaa2 aaaa3\n" + "bbb => bbbb1 bbbb2\n";
SolrSynonymParser parser = new SolrSynonymParser(true, true, new MockAnalyzer(Random()));
parser.Parse(new StringReader(testFile));
SynonymMap map = parser.Build();
Analyzer analyzer = new AnalyzerAnonymousInnerClassHelper5(this, map);
// where did my pot go?!
AssertAnalyzesTo(analyzer, "xyzzy bbb pot of gold", new string[] { "xyzzy", "bbbb1", "pot", "bbbb2", "of", "gold" });
// this one nukes 'pot' and 'of'
// xyzzy aaa pot of gold -> xyzzy aaaa1 aaaa2 aaaa3 gold
AssertAnalyzesTo(analyzer, "xyzzy aaa pot of gold", new string[] { "xyzzy", "aaaa1", "pot", "aaaa2", "of", "aaaa3", "gold" });
}
public virtual void TestEmptyTerm()
{
Random random = Random();
for (int i = 0; i < 512; i++)
{
WordDelimiterFlags flags = (WordDelimiterFlags)i;
CharArraySet protectedWords;
if (random.nextBoolean())
{
protectedWords = new CharArraySet(TEST_VERSION_CURRENT, new string[] { "a", "b", "cd" }, false);
}
else
{
protectedWords = null;
}
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this, flags, protectedWords);
// depending upon options, this thing may or may not preserve the empty term
CheckAnalysisConsistency(random, a, random.nextBoolean(), "");
}
}
public virtual void TestUnigramsAndBigramsHanOnly()
{
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this);
AssertAnalyzesTo(a, "多くの学生が試験に落ちた。", new string[] { "多", "く", "の", "学", "学生", "生", "が", "試", "試験", "験", "に", "落", "ち", "た" }, new int[] { 0, 1, 2, 3, 3, 4, 5, 6, 6, 7, 8, 9, 10, 11 }, new int[] { 1, 2, 3, 4, 5, 5, 6, 7, 8, 8, 9, 10, 11, 12 }, new string[] { "<SINGLE>", "<HIRAGANA>", "<HIRAGANA>", "<SINGLE>", "<DOUBLE>", "<SINGLE>", "<HIRAGANA>", "<SINGLE>", "<DOUBLE>", "<SINGLE>", "<HIRAGANA>", "<SINGLE>", "<HIRAGANA>", "<HIRAGANA>", "<SINGLE>" }, new int[] { 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 1 }, new int[] { 1, 1, 1, 1, 2, 1, 1, 1, 2, 1, 1, 1, 1, 1 });
}
public virtual void TestVersion36()
{
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this);
AssertAnalyzesTo(a, "this is just a t\u08E6st [email protected]", new string[] { "this", "is", "just", "a", "t", "st", "*****@*****.**" }); // new combining mark in 6.1
}
/// <summary>
/// concat numbers + words + all + preserve original </summary>
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testLotsOfConcatenating2() throws Exception
public virtual void testLotsOfConcatenating2()
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int flags = PRESERVE_ORIGINAL | GENERATE_WORD_PARTS | GENERATE_NUMBER_PARTS | CATENATE_WORDS | CATENATE_NUMBERS | CATENATE_ALL | SPLIT_ON_CASE_CHANGE | SPLIT_ON_NUMERICS | STEM_ENGLISH_POSSESSIVE;
int flags = PRESERVE_ORIGINAL | GENERATE_WORD_PARTS | GENERATE_NUMBER_PARTS | CATENATE_WORDS | CATENATE_NUMBERS | CATENATE_ALL | SPLIT_ON_CASE_CHANGE | SPLIT_ON_NUMERICS | STEM_ENGLISH_POSSESSIVE;
/* analyzer that uses whitespace + wdf */
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this, flags);
assertAnalyzesTo(a, "abc-def-123-456", new string[] {"abc-def-123-456", "abc", "abcdef", "abcdef123456", "def", "123", "123456", "456"}, new int[] {0, 0, 0, 0, 4, 8, 8, 12}, new int[] {15, 3, 7, 15, 7, 11, 15, 15}, new int[] {1, 0, 0, 0, 1, 1, 0, 1});
}
public virtual void TestLotsOfConcatenating2()
{
int flags = WordDelimiterFilter.PRESERVE_ORIGINAL | WordDelimiterFilter.GENERATE_WORD_PARTS | WordDelimiterFilter.GENERATE_NUMBER_PARTS | WordDelimiterFilter.CATENATE_WORDS | WordDelimiterFilter.CATENATE_NUMBERS | WordDelimiterFilter.CATENATE_ALL | WordDelimiterFilter.SPLIT_ON_CASE_CHANGE | WordDelimiterFilter.SPLIT_ON_NUMERICS | WordDelimiterFilter.STEM_ENGLISH_POSSESSIVE;
/* analyzer that uses whitespace + wdf */
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this, flags);
AssertAnalyzesTo(a, "abc-def-123-456", new string[] { "abc-def-123-456", "abc", "abcdef", "abcdef123456", "def", "123", "123456", "456" }, new int[] { 0, 0, 0, 0, 4, 8, 8, 12 }, new int[] { 15, 3, 7, 15, 7, 11, 15, 15 }, new int[] { 1, 0, 0, 0, 1, 1, 0, 1 });
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testUnigramsAndBigramsHanOnly() throws Exception
public virtual void testUnigramsAndBigramsHanOnly()
{
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this);
assertAnalyzesTo(a, "多くの学生が試験に落ちた。", new string[] {"多", "く", "の", "学", "学生", "生", "が", "試", "試験", "験", "に", "落", "ち", "た"}, new int[] {0, 1, 2, 3, 3, 4, 5, 6, 6, 7, 8, 9, 10, 11}, new int[] {1, 2, 3, 4, 5, 5, 6, 7, 8, 8, 9, 10, 11, 12}, new string[] {"<SINGLE>", "<HIRAGANA>", "<HIRAGANA>", "<SINGLE>", "<DOUBLE>", "<SINGLE>", "<HIRAGANA>", "<SINGLE>", "<DOUBLE>", "<SINGLE>", "<HIRAGANA>", "<SINGLE>", "<HIRAGANA>", "<HIRAGANA>", "<SINGLE>"}, new int[] {1, 1, 1, 1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 1}, new int[] {1, 1, 1, 1, 2, 1, 1, 1, 2, 1, 1, 1, 1, 1});
}
public virtual void TestMockGraphTokenFilterBeforeHolesRandom()
{
for (int iter = 0; iter < 10 * RANDOM_MULTIPLIER; iter++)
{
if (VERBOSE)
{
Console.WriteLine("\nTEST: iter=" + iter);
}
// Make new analyzer each time, because MGTF has fixed
// seed:
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this);
Random random = Random();
CheckRandomData(random, a, 5, AtLeast(100));
}
}
public virtual void TestEmptyTerm()
{
CharArraySet dict = makeDictionary("a", "e", "i", "o", "u", "y", "bc", "def");
Analyzer a = new AnalyzerAnonymousInnerClassHelper4(this, dict);
CheckOneTerm(a, "", "");
//InputSource @is = new InputSource(this.GetType().getResource("da_UTF8.xml").toExternalForm());
using (var @is = this.GetType().getResourceAsStream("da_UTF8.xml"))
{
HyphenationTree hyphenator = HyphenationCompoundWordTokenFilter.GetHyphenationTree(@is);
Analyzer b = new AnalyzerAnonymousInnerClassHelper5(this, hyphenator);
CheckOneTerm(b, "", "");
}
}
public virtual void TestLetterUnicode()
{
Random random = Random();
Analyzer left = new MockAnalyzer(Random(), jvmLetter, false);
Analyzer right = new AnalyzerAnonymousInnerClassHelper5(this);
for (int i = 0; i < 1000; i++)
{
string s = TestUtil.RandomUnicodeString(random);
assertEquals(s, left.TokenStream("foo", newStringReader(s)), right.TokenStream("foo", newStringReader(s)));
}
}
public virtual void TestVanishingTerms()
{
string testFile = "aaa => aaaa1 aaaa2 aaaa3\n" + "bbb => bbbb1 bbbb2\n";
SolrSynonymParser parser = new SolrSynonymParser(true, true, new MockAnalyzer(Random()));
parser.Parse(new StringReader(testFile));
SynonymMap map = parser.Build();
Analyzer analyzer = new AnalyzerAnonymousInnerClassHelper5(this, map);
// where did my pot go?!
AssertAnalyzesTo(analyzer, "xyzzy bbb pot of gold", new string[] { "xyzzy", "bbbb1", "pot", "bbbb2", "of", "gold" });
// this one nukes 'pot' and 'of'
// xyzzy aaa pot of gold -> xyzzy aaaa1 aaaa2 aaaa3 gold
AssertAnalyzesTo(analyzer, "xyzzy aaa pot of gold", new string[] { "xyzzy", "aaaa1", "pot", "aaaa2", "of", "aaaa3", "gold" });
}
public virtual void TestVersion36()
{
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this);
AssertAnalyzesTo(a, "this is just a t\u08E6st [email protected]", new string[] { "this", "is", "just", "a", "t", "st", "*****@*****.**" }); // new combining mark in 6.1
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testEmptyTerm() throws Exception
public virtual void testEmptyTerm()
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.analysis.util.CharArraySet dict = makeDictionary("a", "e", "i", "o", "u", "y", "bc", "def");
CharArraySet dict = makeDictionary("a", "e", "i", "o", "u", "y", "bc", "def");
Analyzer a = new AnalyzerAnonymousInnerClassHelper4(this, dict);
checkOneTerm(a, "", "");
InputSource @is = new InputSource(this.GetType().getResource("da_UTF8.xml").toExternalForm());
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.analysis.compound.hyphenation.HyphenationTree hyphenator = HyphenationCompoundWordTokenFilter.getHyphenationTree(is);
HyphenationTree hyphenator = HyphenationCompoundWordTokenFilter.getHyphenationTree(@is);
Analyzer b = new AnalyzerAnonymousInnerClassHelper5(this, hyphenator);
checkOneTerm(b, "", "");
}
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testEmptyTerm() throws java.io.IOException
public virtual void testEmptyTerm()
{
Random random = random();
for (int i = 0; i < 512; i++)
{
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final int flags = i;
int flags = i;
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final org.apache.lucene.analysis.util.CharArraySet protectedWords;
CharArraySet protectedWords;
if (random.nextBoolean())
{
protectedWords = new CharArraySet(TEST_VERSION_CURRENT, new HashSet<>("a", "b", "cd"), false);
}
else
{
protectedWords = null;
}
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this, flags, protectedWords);
// depending upon options, this thing may or may not preserve the empty term
checkAnalysisConsistency(random, a, random.nextBoolean(), "");
}
}
// LUCENE-3375
//JAVA TO C# CONVERTER WARNING: Method 'throws' clauses are not available in .NET:
//ORIGINAL LINE: public void testVanishingTerms() throws Exception
public virtual void testVanishingTerms()
{
string testFile = "aaa => aaaa1 aaaa2 aaaa3\n" + "bbb => bbbb1 bbbb2\n";
SolrSynonymParser parser = new SolrSynonymParser(true, true, new MockAnalyzer(random()));
parser.parse(new StringReader(testFile));
//JAVA TO C# CONVERTER WARNING: The original Java variable was marked 'final':
//ORIGINAL LINE: final SynonymMap map = parser.build();
SynonymMap map = parser.build();
Analyzer analyzer = new AnalyzerAnonymousInnerClassHelper5(this, map);
// where did my pot go?!
assertAnalyzesTo(analyzer, "xyzzy bbb pot of gold", new string[] {"xyzzy", "bbbb1", "pot", "bbbb2", "of", "gold"});
// this one nukes 'pot' and 'of'
// xyzzy aaa pot of gold -> xyzzy aaaa1 aaaa2 aaaa3 gold
assertAnalyzesTo(analyzer, "xyzzy aaa pot of gold", new string[] {"xyzzy", "aaaa1", "pot", "aaaa2", "of", "aaaa3", "gold"});
}
public virtual void TestEmptyTerm()
{
Random random = Random();
for (int i = 0; i < 512; i++)
{
int flags = i;
CharArraySet protectedWords;
if (random.nextBoolean())
{
protectedWords = new CharArraySet(TEST_VERSION_CURRENT, new string[] { "a", "b", "cd" }, false);
}
else
{
protectedWords = null;
}
Analyzer a = new AnalyzerAnonymousInnerClassHelper5(this, flags, protectedWords);
// depending upon options, this thing may or may not preserve the empty term
CheckAnalysisConsistency(random, a, random.nextBoolean(), "");
}
}
