public virtual void TestReset()
{
NGramTokenizer tokenizer = new NGramTokenizer(TEST_VERSION_CURRENT, input, 1, 1);
AssertTokenStreamContents(tokenizer, new string[] { "a", "b", "c", "d", "e" }, new int[] { 0, 1, 2, 3, 4 }, new int[] { 1, 2, 3, 4, 5 }, 5); // abcde
tokenizer.SetReader(new StringReader("abcde"));
AssertTokenStreamContents(tokenizer, new string[] { "a", "b", "c", "d", "e" }, new int[] { 0, 1, 2, 3, 4 }, new int[] { 1, 2, 3, 4, 5 }, 5); // abcde
}
public virtual void TestRandomStrings()
{
for (int i = 0; i < 10; i++)
{
int min = TestUtil.NextInt32(Random, 2, 10);
int max = TestUtil.NextInt32(Random, min, 20);
Analyzer a = Analyzer.NewAnonymous(createComponents: (fieldName, reader) =>
{
Tokenizer tokenizer = new NGramTokenizer(TEST_VERSION_CURRENT, reader, min, max);
return(new TokenStreamComponents(tokenizer, tokenizer));
});
CheckRandomData(Random, a, 200 * RandomMultiplier, 20);
CheckRandomData(Random, a, 10 * RandomMultiplier, 1027);
}
}
public void TestOversizedNgrams()
{
NGramTokenizer tokenizer = new NGramTokenizer(input, 6, 7);
AssertTokenStreamContents(tokenizer, new String[0], new int[0], new int[0], 5 /* abcde */);
}
public void TestNgrams()
{
NGramTokenizer tokenizer = new NGramTokenizer(input, 1, 3);
AssertTokenStreamContents(tokenizer,
new String[] { "a", "b", "c", "d", "e", "ab", "bc", "cd", "de", "abc", "bcd", "cde" },
new int[] { 0, 1, 2, 3, 4, 0, 1, 2, 3, 0, 1, 2 },
new int[] { 1, 2, 3, 4, 5, 2, 3, 4, 5, 3, 4, 5 },
5 /* abcde */
);
}
public void TestBigrams()
{
NGramTokenizer tokenizer = new NGramTokenizer(input, 2, 2);
AssertTokenStreamContents(tokenizer, new String[] { "ab", "bc", "cd", "de" }, new int[] { 0, 1, 2, 3 }, new int[] { 2, 3, 4, 5 }, 5 /* abcde */);
}
public void TestUnigrams()
{
NGramTokenizer tokenizer = new NGramTokenizer(input, 1, 1);
AssertTokenStreamContents(tokenizer, new String[] { "a", "b", "c", "d", "e" }, new int[] { 0, 1, 2, 3, 4 }, new int[] { 1, 2, 3, 4, 5 }, 5 /* abcde */);
}
public void TestReset()
{
NGramTokenizer tokenizer = new NGramTokenizer(input, 1, 1);
AssertTokenStreamContents(tokenizer, new String[] { "a", "b", "c", "d", "e" }, new int[] { 0, 1, 2, 3, 4 }, new int[] { 1, 2, 3, 4, 5 }, 5 /* abcde */);
tokenizer.Reset(new StringReader("abcde"));
AssertTokenStreamContents(tokenizer, new String[] { "a", "b", "c", "d", "e" }, new int[] { 0, 1, 2, 3, 4 }, new int[] { 1, 2, 3, 4, 5 }, 5 /* abcde */);
}
public virtual void TestOversizedNgrams()
{
NGramTokenizer tokenizer = new NGramTokenizer(TEST_VERSION_CURRENT, input, 6, 7);
AssertTokenStreamContents(tokenizer, new string[0], new int[0], new int[0], 5); // abcde
}
public virtual void TestNgrams()
{
NGramTokenizer tokenizer = new NGramTokenizer(TEST_VERSION_CURRENT, input, 1, 3);
AssertTokenStreamContents(tokenizer, new string[] { "a", "ab", "abc", "b", "bc", "bcd", "c", "cd", "cde", "d", "de", "e" }, new int[] { 0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 4 }, new int[] { 1, 2, 3, 2, 3, 4, 3, 4, 5, 4, 5, 5 }, null, null, null, 5, false); // abcde
}
public virtual void TestBigrams()
{
NGramTokenizer tokenizer = new NGramTokenizer(TEST_VERSION_CURRENT, input, 2, 2);
AssertTokenStreamContents(tokenizer, new string[] { "ab", "bc", "cd", "de" }, new int[] { 0, 1, 2, 3 }, new int[] { 2, 3, 4, 5 }, 5); // abcde
}
protected override TokenStreamComponents CreateComponents(string fieldName, TextReader reader)
{
Tokenizer tokenizer = new NGramTokenizer(TEST_VERSION_CURRENT, reader, min, max);
return(new TokenStreamComponents(tokenizer, tokenizer));
}