/** * Create a new Normalizer2CharFilter that combines NFKC normalization, Case * Folding, and removes Default Ignorables (NFKC_Casefold) */ public ICUNormalizer2CharFilter(Reader in) { this(in, Normalizer2.getInstance(null, "nfkc_cf", Normalizer2.Mode.COMPOSE)); }
/** * Create a new Normalizer2Filter that combines NFKC normalization, Case * Folding, and removes Default Ignorables (NFKC_Casefold) */ public ICUNormalizer2Filter(TokenStream input) { this(input, Normalizer2.getInstance(null, "nfkc_cf", Normalizer2.Mode.COMPOSE)); }
/** * Create a new Normalizer2Filter that combines NFKC normalization, Case * Folding, and removes Default Ignorables (NFKC_Casefold) */ public ICUNormalizer2Filter(TokenStream input) { this(input, Normalizer2.getInstance(null, "nfkc_cf", Normalizer2.Mode.COMPOSE)); }
public static Normalizer2 getNFCInstance() { return Normalizer2.getInstance(null, "nfc", Normalizer2.Mode.COMPOSE); } }
/** * Constructor that takes {@link Reader} and {@link Normalizer2}. * * @param reader * @param name * @param mode */ public GosenNormalizerCharFilter(Reader reader, String name, Normalizer2.Mode mode) { super(reader); this.normalizer = Normalizer2.getInstance(null, name, mode); buffer.reset(input); }
@Override public TokenStream create(TokenStream tokenStream) { return new org.apache.lucene.analysis.icu.ICUNormalizer2Filter(tokenStream, Normalizer2.getInstance(null, name, Normalizer2.Mode.COMPOSE)); } }
@Override public TokenStream create(TokenStream tokenStream) { return new org.apache.lucene.analysis.icu.ICUNormalizer2Filter(tokenStream, Normalizer2.getInstance(null, "nfkc_cf", Normalizer2.Mode.COMPOSE)); } }));
@Override public TokenStream create(TokenStream tokenStream) { return new org.apache.lucene.analysis.icu.ICUNormalizer2Filter(tokenStream, Normalizer2.getInstance(null, name, Normalizer2.Mode.COMPOSE)); } }
/** * Constructor that takes {@link Reader} and {@link Normalizer2}. * * @param reader * @param name * @param mode */ public GosenNormalizerCharFilter(Reader reader, String name, Normalizer2.Mode mode) { super(reader); this.normalizer = Normalizer2.getInstance(null, name, mode); buffer.reset(input); }
@Override public TokenStream create(TokenStream tokenStream) { return new org.apache.lucene.analysis.icu.ICUNormalizer2Filter(tokenStream, Normalizer2.getInstance(null, "nfkc_cf", Normalizer2.Mode.COMPOSE)); } }));
@Inject public IcuNormalizerCharFilterFactory(Index index, IndexSettingsService indexSettingsService, @Assisted String name, @Assisted Settings settings) { super(index, indexSettingsService.getSettings(), name); this.name = settings.get("name", "nfkc_cf"); String mode = settings.get("mode"); if (!"compose".equals(mode) && !"decompose".equals(mode)) { mode = "compose"; } this.normalizer = Normalizer2.getInstance( null, this.name, "compose".equals(mode) ? Normalizer2.Mode.COMPOSE : Normalizer2.Mode.DECOMPOSE); }
@Inject public IcuNormalizerCharFilterFactory(Index index, @IndexSettings Settings indexSettings, @Assisted String name, @Assisted Settings settings) { super(index, indexSettings, name); this.name = settings.get("name", "nfkc_cf"); String mode = settings.get("mode"); if (!"compose".equals(mode) && !"decompose".equals(mode)) { mode = "compose"; } this.normalizer = Normalizer2.getInstance( null, this.name, "compose".equals(mode) ? Normalizer2.Mode.COMPOSE : Normalizer2.Mode.DECOMPOSE); }
/** Creates a new ICUNormalizer2CharFilterFactory */ public ICUNormalizer2CharFilterFactory(Map<String,String> args) { super(args); String name = get(args, "name", "nfkc_cf"); String mode = get(args, "mode", Arrays.asList("compose", "decompose"), "compose"); Normalizer2 normalizer = Normalizer2.getInstance (null, name, "compose".equals(mode) ? Normalizer2.Mode.COMPOSE : Normalizer2.Mode.DECOMPOSE); String filter = get(args, "filter"); if (filter != null) { UnicodeSet set = new UnicodeSet(filter); if (!set.isEmpty()) { set.freeze(); normalizer = new FilteredNormalizer2(normalizer, set); } } if (!args.isEmpty()) { throw new IllegalArgumentException("Unknown parameters: " + args); } this.normalizer = normalizer; }
/** Creates a new ICUNormalizer2FilterFactory */ public ICUNormalizer2FilterFactory(Map<String,String> args) { super(args); String name = get(args, "name", "nfkc_cf"); String mode = get(args, "mode", Arrays.asList("compose", "decompose"), "compose"); Normalizer2 normalizer = Normalizer2.getInstance (null, name, "compose".equals(mode) ? Normalizer2.Mode.COMPOSE : Normalizer2.Mode.DECOMPOSE); String filter = get(args, "filter"); if (filter != null) { UnicodeSet set = new UnicodeSet(filter); if (!set.isEmpty()) { set.freeze(); normalizer = new FilteredNormalizer2(normalizer, set); } } if (!args.isEmpty()) { throw new IllegalArgumentException("Unknown parameters: " + args); } this.normalizer = normalizer; }
@Override public TokenStream create(TokenStream tokenStream) { // The ICUFoldingFilter is in fact implemented as a ICUNormalizer2Filter. // ICUFoldingFilter lacks a constructor for adding filtering so we implemement it here if (unicodeSetFilter != null) { Normalizer2 base = Normalizer2.getInstance( ICUFoldingFilter.class.getResourceAsStream("utr30.nrm"), "utr30", Normalizer2.Mode.COMPOSE); UnicodeSet unicodeSet = new UnicodeSet(unicodeSetFilter); unicodeSet.freeze(); Normalizer2 filtered = new FilteredNormalizer2(base, unicodeSet); return new org.apache.lucene.analysis.icu.ICUNormalizer2Filter(tokenStream, filtered); } else { return new ICUFoldingFilter(tokenStream); } } }
@Override public TokenStream create(TokenStream tokenStream) { // The ICUFoldingFilter is in fact implemented as a ICUNormalizer2Filter. // ICUFoldingFilter lacks a constructor for adding filtering so we implemement it here if (unicodeSetFilter != null) { Normalizer2 base = Normalizer2.getInstance( ICUFoldingFilter.class.getResourceAsStream("utr30.nrm"), "utr30", Normalizer2.Mode.COMPOSE); UnicodeSet unicodeSet = new UnicodeSet(unicodeSetFilter); unicodeSet.freeze(); Normalizer2 filtered = new FilteredNormalizer2(base, unicodeSet); return new org.apache.lucene.analysis.icu.ICUNormalizer2Filter(tokenStream, filtered); } else { return new ICUFoldingFilter(tokenStream); } } }