org.elasticsearch.index.analysis.AnalyzerProvider Java Examples

@Override
public Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {
    Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> extra = new LinkedHashMap<>();
    if (settings.getAsBoolean("plugins.xbib.icu.enabled", true)) {
        extra.put("icu_collation", IcuCollationKeyAnalyzerProvider::new);
    }
    if (settings.getAsBoolean("plugins.xbib.hyphen.enabled", true)) {
        extra.put("hyphen", HyphenAnalyzerProvider::new);
    }
    if (settings.getAsBoolean("plugins.xbib.naturalsort.enabled", true)) {
        extra.put("naturalsort", NaturalSortKeyAnalyzerProvider::new);
    }
    if (settings.getAsBoolean("plugins.xbib.sortform.enabled", true)) {
        extra.put("sortform", SortformAnalyzerProvider::new);
    }
    if (settings.getAsBoolean("plugins.xbib.standardnumber.enabled", true)) {
        extra.put("standardnumber", (indexSettings, environment, name, factorySettings) ->
                new StandardnumberAnalyzerProvider(indexSettings, environment, name, factorySettings, standardNumberTypeParser));
    }
    return extra;
}
 

public AnalysisModule(Environment environment, List<AnalysisPlugin> plugins) throws IOException {
    NamedRegistry<AnalysisProvider<CharFilterFactory>> charFilters = setupCharFilters(plugins);
    NamedRegistry<org.apache.lucene.analysis.hunspell.Dictionary> hunspellDictionaries = setupHunspellDictionaries(plugins);
    HunspellService hunspellService = new HunspellService(environment.settings(), environment, hunspellDictionaries.getRegistry());
    NamedRegistry<AnalysisProvider<TokenFilterFactory>> tokenFilters = setupTokenFilters(plugins, hunspellService);
    NamedRegistry<AnalysisProvider<TokenizerFactory>> tokenizers = setupTokenizers(plugins);
    NamedRegistry<AnalysisProvider<AnalyzerProvider<?>>> analyzers = setupAnalyzers(plugins);
    NamedRegistry<AnalysisProvider<AnalyzerProvider<?>>> normalizers = setupNormalizers();

    Map<String, PreConfiguredCharFilter> preConfiguredCharFilters = setupPreConfiguredCharFilters(plugins);
    Map<String, PreConfiguredTokenFilter> preConfiguredTokenFilters = setupPreConfiguredTokenFilters(plugins);
    Map<String, PreConfiguredTokenizer> preConfiguredTokenizers = setupPreConfiguredTokenizers(plugins);
    Map<String, PreBuiltAnalyzerProviderFactory> preConfiguredAnalyzers = setupPreBuiltAnalyzerProviderFactories(plugins);

    analysisRegistry = new AnalysisRegistry(environment,
            charFilters.getRegistry(), tokenFilters.getRegistry(), tokenizers.getRegistry(),
            analyzers.getRegistry(), normalizers.getRegistry(),
            preConfiguredCharFilters, preConfiguredTokenFilters, preConfiguredTokenizers, preConfiguredAnalyzers);
}
 

@Override
public Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {
    Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> extra = new HashMap<>();

    extra.put("hlseg_search", HLSegAnalyzerProvider::getHLSegSearchAnalyzerProvider);
    
    return extra;
}
 

@Override
public Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {
  Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> extra = new HashMap<>();

  extra.put("jieba_search", JiebaAnalyzerProvider::getJiebaSearchAnalyzerProvider);
  extra.put("jieba_index", JiebaAnalyzerProvider::getJiebaIndexAnalyzerProvider);

  return extra;
}
 

@Override
public Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {
    Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> extra = new HashMap<>();

    extra.put("ik_smart", IkAnalyzerProvider::getIkSmartAnalyzerProvider);
    extra.put("ik_max_word", IkAnalyzerProvider::getIkAnalyzerProvider);

    return extra;
}
 

@Override
public Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {
    Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> extra = new HashMap<>();

    extra.put("hanlp", HanLPAnalyzerProvider::getHanLPAnalyzerProvider);
    extra.put("hanlp_standard", HanLPAnalyzerProvider::getHanLPStandardAnalyzerProvider);
    extra.put("hanlp_index", HanLPAnalyzerProvider::getHanLPIndexAnalyzerProvider);
    extra.put("hanlp_nlp", HanLPAnalyzerProvider::getHanLPNLPAnalyzerProvider);
    extra.put("hanlp_n_short", HanLPAnalyzerProvider::getHanLPNShortAnalyzerProvider);
    extra.put("hanlp_dijkstra", HanLPAnalyzerProvider::getHanLPDijkstraAnalyzerProvider);
    extra.put("hanlp_crf", HanLPAnalyzerProvider::getHanLPCRFAnalyzerProvider);
    extra.put("hanlp_speed", HanLPAnalyzerProvider::getHanLPSpeedAnalyzerProvider);

    return extra;
}
 

@Override
public Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {
    Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> analyzers = new HashMap<>();
    analyzers.put("hanlp", HanLPAnalyzerProvider::new);
    analyzers.put("hanlp-index", HanLPIndexAnalyzerProvider::new);
    return analyzers;
}
 

@Override
public Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {

    //开启异步线程，先执行一个core的分词，试图去下载依赖的资源

    Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> map = new HashMap<>();

    map.put("mynlp", MynlpAnalyzerProvider::new);
    map.put("mynlp-core", MynlpAnalyzerProvider::new);
    if (enableCws) {
        map.put("mynlp-cws", MynlpAnalyzerProvider::new);
    }

    return map;
}
 

@Override
public Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {
    Map<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> providerMap
            = new HashMap<String, AnalysisModule.AnalysisProvider<AnalyzerProvider<? extends Analyzer>>>();

    providerMap.put("lc_index", LcPinyinAnalyzerProvider::getIndexAnalyzerProvider);
    providerMap.put("lc_search", LcPinyinAnalyzerProvider::getSmartPinyinAnalyzerProvider);

    return providerMap;
}
 

private NamedRegistry<AnalysisProvider<AnalyzerProvider<?>>> setupAnalyzers(List<AnalysisPlugin> plugins) {
    NamedRegistry<AnalysisProvider<AnalyzerProvider<?>>> analyzers = new NamedRegistry<>("analyzer");
    analyzers.register("default", StandardAnalyzerProvider::new);
    analyzers.register("standard", StandardAnalyzerProvider::new);
    analyzers.register("simple", SimpleAnalyzerProvider::new);
    analyzers.register("stop", StopAnalyzerProvider::new);
    analyzers.register("whitespace", WhitespaceAnalyzerProvider::new);
    analyzers.register("keyword", KeywordAnalyzerProvider::new);
    analyzers.extractAndRegister(plugins, AnalysisPlugin::getAnalyzers);
    return analyzers;
}
 

@Override
public Map<String, AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {
    return singletonMap(ANALYZER_NAME, (indexSettings, environment, name, settings) -> new MorfologikAnalyzerProvider(indexSettings, name, settings));
}
 

@Override
public Map<String, AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {
    return singletonMap("word", ChineseWordAnalyzerProvider::new);
}
 

private NamedRegistry<AnalysisProvider<AnalyzerProvider<?>>> setupNormalizers() {
    return new NamedRegistry<>("normalizer");
}
 

@Override
public Map<String, AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {
    Map<String, AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> analyzers = new TreeMap<>();
    analyzers.put("fingerprint", FingerprintAnalyzerProvider::new);
    analyzers.put("standard_html_strip", StandardHtmlStripAnalyzerProvider::new);
    analyzers.put("pattern", PatternAnalyzerProvider::new);
    analyzers.put("snowball", SnowballAnalyzerProvider::new);

    // Language analyzers:
    analyzers.put("arabic", ArabicAnalyzerProvider::new);
    analyzers.put("armenian", ArmenianAnalyzerProvider::new);
    analyzers.put("basque", BasqueAnalyzerProvider::new);
    analyzers.put("bengali", BengaliAnalyzerProvider::new);
    analyzers.put("brazilian", BrazilianAnalyzerProvider::new);
    analyzers.put("bulgarian", BulgarianAnalyzerProvider::new);
    analyzers.put("catalan", CatalanAnalyzerProvider::new);
    analyzers.put("chinese", ChineseAnalyzerProvider::new);
    analyzers.put("cjk", CjkAnalyzerProvider::new);
    analyzers.put("czech", CzechAnalyzerProvider::new);
    analyzers.put("danish", DanishAnalyzerProvider::new);
    analyzers.put("dutch", DutchAnalyzerProvider::new);
    analyzers.put("english", EnglishAnalyzerProvider::new);
    analyzers.put("finnish", FinnishAnalyzerProvider::new);
    analyzers.put("french", FrenchAnalyzerProvider::new);
    analyzers.put("galician", GalicianAnalyzerProvider::new);
    analyzers.put("german", GermanAnalyzerProvider::new);
    analyzers.put("greek", GreekAnalyzerProvider::new);
    analyzers.put("hindi", HindiAnalyzerProvider::new);
    analyzers.put("hungarian", HungarianAnalyzerProvider::new);
    analyzers.put("indonesian", IndonesianAnalyzerProvider::new);
    analyzers.put("irish", IrishAnalyzerProvider::new);
    analyzers.put("italian", ItalianAnalyzerProvider::new);
    analyzers.put("latvian", LatvianAnalyzerProvider::new);
    analyzers.put("lithuanian", LithuanianAnalyzerProvider::new);
    analyzers.put("norwegian", NorwegianAnalyzerProvider::new);
    analyzers.put("persian", PersianAnalyzerProvider::new);
    analyzers.put("portuguese", PortugueseAnalyzerProvider::new);
    analyzers.put("romanian", RomanianAnalyzerProvider::new);
    analyzers.put("russian", RussianAnalyzerProvider::new);
    analyzers.put("sorani", SoraniAnalyzerProvider::new);
    analyzers.put("spanish", SpanishAnalyzerProvider::new);
    analyzers.put("swedish", SwedishAnalyzerProvider::new);
    analyzers.put("turkish", TurkishAnalyzerProvider::new);
    analyzers.put("thai", ThaiAnalyzerProvider::new);
    return analyzers;
}
 

/**
 * Override to add additional {@link Analyzer}s. See {@link #requiresAnalysisSettings(AnalysisProvider)}
 * how to on get the configuration from the index.
 */
default Map<String, AnalysisProvider<AnalyzerProvider<? extends Analyzer>>> getAnalyzers() {
    return emptyMap();
}