Examples with UnicodeSet android.icu.text.UnicodeSet used on opensource projects

Example 66 with UnicodeSet

use of android.icu.text.UnicodeSet in project j2objc by google.

the class LocaleDataTest method TestEnglishExemplarCharacters.

// Simple test case for checking exemplar character type coverage
@Test
public void TestEnglishExemplarCharacters() {
    final char[] testChars = { // standard
    0x61, // auxiliary
    0xE1, // index
    0x41, // filler for deprecated currency exemplar
    0, // punctuation
    0x2D };
    LocaleData ld = LocaleData.getInstance(ULocale.ENGLISH);
    for (int type = 0; type < LocaleData.ES_COUNT; type++) {
        UnicodeSet exSet = ld.getExemplarSet(0, type);
        if (exSet != null) {
            if (testChars[type] > 0 && !exSet.contains(testChars[type])) {
                errln("Character '" + testChars[type] + "' is not included in exemplar type " + type);
            }
        }
    }
    try {
        // out of bounds value
        ld.getExemplarSet(0, LocaleData.ES_COUNT);
        throw new ICUException("Test failure; should throw exception");
    } catch (IllegalArgumentException e) {
        assertEquals("", "java.lang.ArrayIndexOutOfBoundsException", e.getCause().getClass().getName());
    }
}

Example 67 with UnicodeSet

use of android.icu.text.UnicodeSet in project j2objc by google.

the class LocaleDataTest method TestExemplarSetTypes.

// Test case created for checking type coverage of static getExemplarSet method.
// See #9785, #9794 and #9795
@Test
public void TestExemplarSetTypes() {
    final String[] testLocales = { // No auxiliary / index exemplars as of ICU 50
    "am", "en", // #9785
    "th", // Bogus locale
    "foo" };
    final int[] testTypes = { LocaleData.ES_STANDARD, LocaleData.ES_AUXILIARY, LocaleData.ES_INDEX, LocaleData.ES_CURRENCY, LocaleData.ES_PUNCTUATION };
    final String[] testTypeNames = { "ES_STANDARD", "ES_AUXILIARY", "ES_INDEX", "ES_CURRENCY", "ES_PUNCTUATION" };
    for (String locstr : testLocales) {
        ULocale loc = new ULocale(locstr);
        for (int i = 0; i < testTypes.length; i++) {
            try {
                UnicodeSet set = LocaleData.getExemplarSet(loc, 0, testTypes[i]);
                if (set == null) {
                    // Not sure null is really OK (#9795)
                    logln(loc + "(" + testTypeNames[i] + ") returned null");
                } else if (set.isEmpty()) {
                    // This is probably reasonable when data is absent
                    logln(loc + "(" + testTypeNames[i] + ") returned an empty set");
                }
            } catch (Exception e) {
                errln(loc + "(" + testTypeNames[i] + ") Exception:" + e.getMessage());
            }
        }
    }
}

Example 68 with UnicodeSet

use of android.icu.text.UnicodeSet in project j2objc by google.

the class CollationDataBuilder method setDigitTags.

protected void setDigitTags() {
    UnicodeSet digits = new UnicodeSet("[:Nd:]");
    UnicodeSetIterator iter = new UnicodeSetIterator(digits);
    while (iter.next()) {
        assert (iter.codepoint != UnicodeSetIterator.IS_STRING);
        int c = iter.codepoint;
        int ce32 = trie.get(c);
        if (ce32 != Collation.FALLBACK_CE32 && ce32 != Collation.UNASSIGNED_CE32) {
            int index = addCE32(ce32);
            if (index > Collation.MAX_INDEX) {
                throw new IndexOutOfBoundsException("too many mappings");
            // BufferOverflowException is a better fit
            // but cannot be constructed with a message string.
            }
            ce32 = Collation.makeCE32FromTagIndexAndLength(Collation.DIGIT_TAG, index, // u_charDigitValue(c)
            UCharacter.digit(c));
            trie.set(c, ce32);
        }
    }
}

Example 69 with UnicodeSet

use of android.icu.text.UnicodeSet in project j2objc by google.

the class CollationBuilder method addTailComposites.

private void addTailComposites(CharSequence nfdPrefix, CharSequence nfdString) {
    // Look for the last starter in the NFD string.
    int lastStarter;
    int indexAfterLastStarter = nfdString.length();
    for (; ; ) {
        // no starter at all
        if (indexAfterLastStarter == 0) {
            return;
        }
        lastStarter = Character.codePointBefore(nfdString, indexAfterLastStarter);
        if (nfd.getCombiningClass(lastStarter) == 0) {
            break;
        }
        indexAfterLastStarter -= Character.charCount(lastStarter);
    }
    // No closure to Hangul syllables since we decompose them on the fly.
    if (Hangul.isJamoL(lastStarter)) {
        return;
    }
    // Are there any composites whose decomposition starts with the lastStarter?
    // Note: Normalizer2Impl does not currently return start sets for NFC_QC=Maybe characters.
    // We might find some more equivalent mappings here if it did.
    UnicodeSet composites = new UnicodeSet();
    if (!nfcImpl.getCanonStartSet(lastStarter, composites)) {
        return;
    }
    StringBuilder newNFDString = new StringBuilder(), newString = new StringBuilder();
    long[] newCEs = new long[Collation.MAX_EXPANSION_LENGTH];
    UnicodeSetIterator iter = new UnicodeSetIterator(composites);
    while (iter.next()) {
        assert (iter.codepoint != UnicodeSetIterator.IS_STRING);
        int composite = iter.codepoint;
        String decomp = nfd.getDecomposition(composite);
        if (!mergeCompositeIntoString(nfdString, indexAfterLastStarter, composite, decomp, newNFDString, newString)) {
            continue;
        }
        int newCEsLength = dataBuilder.getCEs(nfdPrefix, newNFDString, newCEs, 0);
        if (newCEsLength > Collation.MAX_EXPANSION_LENGTH) {
            // Ignore mappings that we cannot store.
            continue;
        }
        // Note: It is possible that the newCEs do not make use of the mapping
        // for which we are adding the tail composites, in which case we might be adding
        // unnecessary mappings.
        // For example, when we add tail composites for ae^ (^=combining circumflex),
        // UCA discontiguous-contraction matching does not find any matches
        // for ae_^ (_=any combining diacritic below) *unless* there is also
        // a contraction mapping for ae.
        // Thus, if there is no ae contraction, then the ae^ mapping is ignored
        // while fetching the newCEs for ae_^.
        // TODO: Try to detect this effectively.
        // (Alternatively, print a warning when prefix contractions are missing.)
        // We do not need an explicit mapping for the NFD strings.
        // It is fine if the NFD input collates like this via a sequence of mappings.
        // It also saves a little bit of space, and may reduce the set of characters with contractions.
        int ce32 = addIfDifferent(nfdPrefix, newString, newCEs, newCEsLength, Collation.UNASSIGNED_CE32);
        if (ce32 != Collation.UNASSIGNED_CE32) {
            // was different, was added
            addOnlyClosure(nfdPrefix, newNFDString, newCEs, newCEsLength, ce32);
        }
    }
}

Example 70 with UnicodeSet

use of android.icu.text.UnicodeSet in project j2objc by google.

the class TestBoilerplate method getSet.

public static <T> UnicodeSet getSet(Map<Integer, T> m, T value) {
    UnicodeSet result = new UnicodeSet();
    for (Iterator<Integer> it = m.keySet().iterator(); it.hasNext(); ) {
        Integer key = it.next();
        T val = m.get(key);
        if (!val.equals(value))
            continue;
        result.add(key.intValue());
    }
    return result;
}