Examples with AnsjTokenizer org.ansj.lucene.util.AnsjTokenizer used on opensource projects

Example 1 with AnsjTokenizer

use of org.ansj.lucene.util.AnsjTokenizer in project ansj_seg by NLPchina.

the class AppTest method parse.

private static void parse(Analysis an, List<StopRecognition> filters) throws IOException {
    Tokenizer tokenizer = new AnsjTokenizer(an, filters, null);
    while (tokenizer.incrementToken()) {
        CharTermAttribute attribute = tokenizer.getAttribute(CharTermAttribute.class);
        System.out.println(attribute);
    }
}

Example 2 with AnsjTokenizer

use of org.ansj.lucene.util.AnsjTokenizer in project ansj_seg by NLPchina.

the class AnsjAnalyzer method getTokenizer.

/**
	 * 获得一个tokenizer
	 * 
	 * @param reader
	 * @param type
	 * @param filter
	 * @return
	 */
public static Tokenizer getTokenizer(Reader reader, Map<String, String> args) {
    if (LOG.isDebugEnabled()) {
        LOG.debug("to create tokenizer " + args);
    }
    Analysis analysis = null;
    String temp = null;
    String type = args.get("type");
    if (type == null) {
        type = AnsjAnalyzer.TYPE.base_ansj.name();
    }
    switch(AnsjAnalyzer.TYPE.valueOf(type)) {
        case base_ansj:
            analysis = new BaseAnalysis();
            break;
        case index_ansj:
            analysis = new IndexAnalysis();
            break;
        case dic_ansj:
            analysis = new DicAnalysis();
            break;
        case query_ansj:
            analysis = new ToAnalysis();
            break;
        case nlp_ansj:
            analysis = new NlpAnalysis();
            if (StringUtil.isNotBlank(temp = args.get(CrfLibrary.DEFAULT))) {
                ((NlpAnalysis) analysis).setCrfModel(CrfLibrary.get(temp));
            }
            break;
        default:
            analysis = new BaseAnalysis();
    }
    if (reader != null) {
        analysis.resetContent(reader);
    }
    if (StringUtil.isNotBlank(temp = args.get(DicLibrary.DEFAULT))) {
        //用户自定义词典
        String[] split = temp.split(",");
        Forest[] forests = new Forest[split.length];
        for (int i = 0; i < forests.length; i++) {
            if (StringUtil.isBlank(split[i])) {
                continue;
            }
            forests[i] = DicLibrary.get(split[i]);
        }
        analysis.setForests(forests);
    }
    List<StopRecognition> filters = null;
    if (StringUtil.isNotBlank(temp = args.get(StopLibrary.DEFAULT))) {
        //用户自定义词典
        String[] split = temp.split(",");
        filters = new ArrayList<StopRecognition>();
        for (String key : split) {
            StopRecognition stop = StopLibrary.get(key.trim());
            if (stop != null)
                filters.add(stop);
        }
    }
    List<SynonymsRecgnition> synonyms = null;
    if (StringUtil.isNotBlank(temp = args.get(SynonymsLibrary.DEFAULT))) {
        //同义词词典
        String[] split = temp.split(",");
        synonyms = new ArrayList<SynonymsRecgnition>();
        for (String key : split) {
            SmartForest<List<String>> sf = SynonymsLibrary.get(key.trim());
            if (sf != null)
                synonyms.add(new SynonymsRecgnition(sf));
        }
    }
    if (StringUtil.isNotBlank(temp = args.get(AmbiguityLibrary.DEFAULT))) {
        //歧义词典
        analysis.setAmbiguityForest(AmbiguityLibrary.get(temp.trim()));
    }
    if (StringUtil.isNotBlank(temp = args.get("isNameRecognition"))) {
        // 是否开启人名识别
        analysis.setIsNameRecognition(Boolean.valueOf(temp));
    }
    if (StringUtil.isNotBlank(temp = args.get("isNumRecognition"))) {
        // 是否开启数字识别
        analysis.setIsNumRecognition(Boolean.valueOf(temp));
    }
    if (StringUtil.isNotBlank(temp = args.get("isQuantifierRecognition"))) {
        //量词识别
        analysis.setIsQuantifierRecognition(Boolean.valueOf(temp));
    }
    if (StringUtil.isNotBlank(temp = args.get("isRealName"))) {
        //是否保留原字符
        analysis.setIsRealName(Boolean.valueOf(temp));
    }
    return new AnsjTokenizer(analysis, filters, synonyms);
}

Example 3 with AnsjTokenizer

use of org.ansj.lucene.util.AnsjTokenizer in project ansj_seg by NLPchina.

the class AnsjAnalyzer method getTokenizer.

/**
	 * 获得一个tokenizer
	 * 
	 * @param reader
	 * @param type
	 * @param filter
	 * @return
	 */
public static Tokenizer getTokenizer(Reader reader, Map<String, String> args) {
    if (LOG.isDebugEnabled()) {
        LOG.debug("to create tokenizer " + args);
    }
    Analysis analysis = null;
    String temp = null;
    String type = args.get("type");
    if (type == null) {
        type = AnsjAnalyzer.TYPE.base_ansj.name();
    }
    switch(AnsjAnalyzer.TYPE.valueOf(type)) {
        case base_ansj:
            analysis = new BaseAnalysis();
            break;
        case index_ansj:
            analysis = new IndexAnalysis();
            break;
        case dic_ansj:
            analysis = new DicAnalysis();
            break;
        case query_ansj:
            analysis = new ToAnalysis();
            break;
        case nlp_ansj:
            analysis = new NlpAnalysis();
            if (StringUtil.isNotBlank(temp = args.get(CrfLibrary.DEFAULT))) {
                ((NlpAnalysis) analysis).setCrfModel(CrfLibrary.get(temp));
            }
            break;
        default:
            analysis = new BaseAnalysis();
    }
    if (reader != null) {
        analysis.resetContent(reader);
    }
    if (StringUtil.isNotBlank(temp = args.get(DicLibrary.DEFAULT))) {
        //用户自定义词典
        String[] split = temp.split(",");
        Forest[] forests = new Forest[split.length];
        for (int i = 0; i < forests.length; i++) {
            if (StringUtil.isBlank(split[i])) {
                continue;
            }
            forests[i] = DicLibrary.get(split[i]);
        }
        analysis.setForests(forests);
    }
    List<StopRecognition> filters = null;
    if (StringUtil.isNotBlank(temp = args.get(StopLibrary.DEFAULT))) {
        //用户自定义词典
        String[] split = temp.split(",");
        filters = new ArrayList<StopRecognition>();
        for (String key : split) {
            StopRecognition stop = StopLibrary.get(key.trim());
            if (stop != null)
                filters.add(stop);
        }
    }
    List<SynonymsRecgnition> synonyms = null;
    if (StringUtil.isNotBlank(temp = args.get(SynonymsLibrary.DEFAULT))) {
        //同义词词典
        String[] split = temp.split(",");
        synonyms = new ArrayList<SynonymsRecgnition>();
        for (String key : split) {
            SmartForest<List<String>> sf = SynonymsLibrary.get(key.trim());
            if (sf != null)
                synonyms.add(new SynonymsRecgnition(sf));
        }
    }
    if (StringUtil.isNotBlank(temp = args.get(AmbiguityLibrary.DEFAULT))) {
        //歧义词典
        analysis.setAmbiguityForest(AmbiguityLibrary.get(temp.trim()));
    }
    if (StringUtil.isNotBlank(temp = args.get("isNameRecognition"))) {
        // 是否开启人名识别
        analysis.setIsNameRecognition(Boolean.valueOf(temp));
    }
    if (StringUtil.isNotBlank(temp = args.get("isNumRecognition"))) {
        // 是否开启数字识别
        analysis.setIsNumRecognition(Boolean.valueOf(temp));
    }
    if (StringUtil.isNotBlank(temp = args.get("isQuantifierRecognition"))) {
        //量词识别
        analysis.setIsQuantifierRecognition(Boolean.valueOf(temp));
    }
    if (StringUtil.isNotBlank(temp = args.get("isRealName"))) {
        //是否保留原字符
        analysis.setIsRealName(Boolean.valueOf(temp));
    }
    return new AnsjTokenizer(analysis, filters, synonyms);
}

Example 4 with AnsjTokenizer

use of org.ansj.lucene.util.AnsjTokenizer in project ansj_seg by NLPchina.

the class AppTest method parse.

private static void parse(Analysis an, List<StopRecognition> filters) throws IOException {
    Tokenizer tokenizer = new AnsjTokenizer(an, filters, null);
    while (tokenizer.incrementToken()) {
        CharTermAttribute attribute = tokenizer.getAttribute(CharTermAttribute.class);
        System.out.println(attribute);
    }
}