Example usage for org.apache.lucene.analysis CharFilter correctOffset

Introduction

In this page you can find the example usage for org.apache.lucene.analysis CharFilter correctOffset.

Prototype

public final int correctOffset(int currentOff) 

Source Link

Document

Chains the corrected offset through the input CharFilter(s).

Usage

From source file:org.elasticsearch.index.analysis.SimpleIcuNormalizerCharFilterTests.java

License:Apache License

@Test
public void testDefaultSetting() throws Exception {

    Settings settings = Settings.settingsBuilder().put("path.home", createTempDir())
            .put("index.analysis.char_filter.myNormalizerChar.type", "icu_normalizer").build();
    AnalysisService analysisService = createAnalysisService(settings);
    CharFilterFactory charFilterFactory = analysisService.charFilter("myNormalizerChar");

    String input = "5?????g?/??chk?";
    Normalizer2 normalizer = Normalizer2.getInstance(null, "nfkc_cf", Normalizer2.Mode.COMPOSE);
    String expectedOutput = normalizer.normalize(input);
    CharFilter inputReader = (CharFilter) charFilterFactory.create(new StringReader(input));
    char[] tempBuff = new char[10];
    StringBuilder output = new StringBuilder();
    while (true) {
        int length = inputReader.read(tempBuff);
        if (length == -1)
            break;
        output.append(tempBuff, 0, length);
        assertEquals(output.toString(),//from   w w w .  j ava 2s.c om
                normalizer.normalize(input.substring(0, inputReader.correctOffset(output.length()))));
    }
    assertEquals(expectedOutput, output.toString());
}

---

From source file:org.elasticsearch.index.analysis.SimpleIcuNormalizerCharFilterTests.java

License:Apache License

@Test
public void testNameAndModeSetting() throws Exception {

    Settings settings = Settings.settingsBuilder().put("path.home", createTempDir())
            .put("index.analysis.char_filter.myNormalizerChar.type", "icu_normalizer")
            .put("index.analysis.char_filter.myNormalizerChar.name", "nfkc")
            .put("index.analysis.char_filter.myNormalizerChar.mode", "decompose").build();
    AnalysisService analysisService = createAnalysisService(settings);
    CharFilterFactory charFilterFactory = analysisService.charFilter("myNormalizerChar");

    String input = "5?????g?/??chk?";
    Normalizer2 normalizer = Normalizer2.getInstance(null, "nfkc", Normalizer2.Mode.DECOMPOSE);
    String expectedOutput = normalizer.normalize(input);
    CharFilter inputReader = (CharFilter) charFilterFactory.create(new StringReader(input));
    char[] tempBuff = new char[10];
    StringBuilder output = new StringBuilder();
    while (true) {
        int length = inputReader.read(tempBuff);
        if (length == -1)
            break;
        output.append(tempBuff, 0, length);
        assertEquals(output.toString(),//from ww  w  .  j ava 2 s .  c om
                normalizer.normalize(input.substring(0, inputReader.correctOffset(output.length()))));
    }
    assertEquals(expectedOutput, output.toString());
}

---