com.cybozu.labs.langdetect.Detector Java Examples

private List<DetectedLanguage> detectLanguage(String content) {
	if (content.trim().length() == 0) { // to be consistent with the tika impl?
		log.debug("No input text to detect language from, returning empty list");
		return Collections.emptyList();
	}

	try {
		Detector detector = DetectorFactory.create();
		detector.append(content);
		ArrayList<Language> langlist = detector.getProbabilities();
		ArrayList<DetectedLanguage> solrLangList = new ArrayList<>();
		for (Language l: langlist) 
		{
			if((autoDetectQueryLocales.size() == 0) || (autoDetectQueryLocales.contains(l.lang)))
			{
			    solrLangList.add(new DetectedLanguage(l.lang, l.prob));
			}
		}
		return solrLangList;
	} catch (LangDetectException e) {
		log.debug("Could not determine language, returning empty list: ", e);
		return Collections.emptyList();
	}
}
 

/**
 * Gets a customized detector for a given language.
 *
 * TODO(P1) Adding priority on the language seems to be relatively useless.
 * To be reviewed.
 *
 * @param language
 * @return a {@link Detector} customized for that language
 * @throws LangDetectException
 */
private Detector getDetectorForLanguage(String language) throws LangDetectException {
    Detector detector = DetectorFactory.create();
    HashMap<String, Double> priorityMap = new HashMap();

    for (String supportedLanguage : getSupportedLanguages()) {
        if (supportedLanguage.equals(language)) {
            priorityMap.put(supportedLanguage, 0.8);
        } else if (supportedLanguage.equals("en") && !"en".equals(language)) {
            priorityMap.put(supportedLanguage, 0.5);
        } else {
            priorityMap.put(supportedLanguage, 0.1);
        }
    }

    detector.setPriorMap(priorityMap);

    return detector;
}
 

@Override
public IngestDocument execute(IngestDocument ingestDocument) throws Exception {
    Detector detector = DetectorFactory.create();
    detector.setMaxTextLength(maxLength.bytesAsInt());

    String content;
    try {
        content = ingestDocument.getFieldValue(field, String.class);
    } catch (IllegalArgumentException e) {
        if (ignoreMissing) {
            return ingestDocument;
        }
        throw e;
    }
    if (Strings.isEmpty(content)) {
        return ingestDocument;
    }

    detector.append(content);
    String language = detector.detect();

    ingestDocument.setFieldValue(targetField, language);

    return ingestDocument;
}
 

/**
 * Perform the detection 
 * 
 * @param text to test
 * @return the detected language
 */
public LanguageDetectionResult detectLanguage(String text) {
    try {
        Detector detector = DetectorFactory.create(0.15);
        // issue#47 correction
        detector.append(text.toLowerCase());
        ArrayList<Language> languages = detector.getProbabilities();
        Language detectedLanguage =  
                extractLangWithHighestProbability(languages);
        return new LanguageDetectionResult(detectedLanguage, text, languages.size()>1);
    } catch (LangDetectException ex) {
        LOGGER.warn(ex);
    }
    return null;
}
 

@Override
public DetectionResult detect(String text) {
  Detector detector;
  try {
    detector = DetectorFactory.create();
  } catch (LangDetectException e) {
    // TODO(skreft): log the reason
    return UNKNOWN;
  }

  detector.append(text);
  List<Language> results = detector.getProbabilities();
  if (!results.isEmpty()) {
    Language bestLang = results.get(0);
    return new DetectionResult(bestLang.lang, bestLang.prob);
  }

  return UNKNOWN;
}
 

@Override
public String identifyLanguage(String html)
        throws IOException
{
    // extracting plain html text
    Document doc = Jsoup.parse(html);
    String text = doc.text();

    // we might have removed everything -> no lang
    if (text.isEmpty()) {
        return UNKNOWN_LANGUAGE;
    }

    try {
        Detector detector = DetectorFactory.create();
        detector.append(text);
        String detectedLang = detector.detect();

        ArrayList<Language> detectedProbabilities = detector.getProbabilities();

        if (detectedProbabilities.get(0).prob > PROBABILITY_THRESHOLD) {
            return detectedLang;
        }
        else {
            return UNKNOWN_LANGUAGE;
        }
    }
    catch (LangDetectException e) {
        return UNKNOWN_LANGUAGE;
    }
}
 

/**
 * Try to detect the language of the text in the String.
 * 
 * @param page
 * @return true if the String contains English language, false otherwise
 */
public Boolean isEnglish(String content) {
    try {

        if (content == null || content.isEmpty()) {
            return false;
        }

        Detector detector = DetectorFactory.create();
        detector.append(content);
        ArrayList<Language> langs = detector.getProbabilities();

        if (langs.size() == 0) {
            return false;
        }

        for (Language l : langs) {
            if (l.lang.equals("en")) {
                return true;
            }
        }
        return false;
    } catch (Exception ex) {
        logger.warn("Problem while detecting language in text: " + content, ex);
        return false;
    }
}
 

/**
 * Constructor.
 * Detector instance can be constructed via {@link DetectorFactory#create()}.
 * @param factory {@link DetectorFactory} instance (only DetectorFactory inside)
 */
public Detector(DetectorFactory factory) {
    this.wordLangProbMap = factory.wordLangProbMap;
    this.langlist = factory.langlist;
    this.text = new StringBuilder();

    this.weight = this.alpha / BASE_FREQ;
}
 

/**
 * {@inheritDoc}
 */
public NutchDocument filter(NutchDocument doc, Parse parse, Text url,
		CrawlDatum datum, Inlinks inlinks) throws IndexingException {
	if (conf == null) {
		throw new IndexingException("Not Yet Initialization.");
	}
	if (cause != null) {
		throw new IndexingException("Initialization Failed.", cause);
	}

	String lang = parse.getData().getParseMeta().get(Metadata.LANGUAGE);
	if (lang == null) {
		StringBuilder text = new StringBuilder();
		text.append(parse.getData().getTitle()).append(" ")
				.append(parse.getText());
		try {
			Detector detector = DetectorFactory.create();
			detector.setMaxTextLength(textsize_upper_limit);
			detector.append(text.toString());
			lang = detector.detect();
		} catch (LangDetectException e) {
			throw new IndexingException("Detection failed.", e);
		}
	}
	if (lang == null) lang = "unknown";

	doc.add("lang", lang);
	return doc;
}
 

/**
 * Constructor.
 * Detector instance can be constructed via {@link DetectorFactory#create()}.
 * @param factory {@link DetectorFactory} instance (only DetectorFactory inside)
 */
public Detector(DetectorFactory factory) {
    this.wordLangProbMap = factory.wordLangProbMap;
    this.langlist = factory.langlist;
    this.text = new StringBuffer();
    this.seed  = factory.seed;
}
 

@Test
public static void langDetectSample() {

    long startTime;
    String lang = "none";
    ArrayList<Language> langlist = null;

    try {

        // Initialize
        startTime = System.currentTimeMillis();
        DetectorFactory.create();
        System.out.println("Initialization finished in " + (System.currentTimeMillis() - startTime) + " ms");

        // Detect
        startTime = System.currentTimeMillis();
        Detector detector = DetectorFactory.create();
        detector.append("The quick brown fox jumps over the lazy dog.");
        lang = detector.detect();
        System.out.println("Detection finished in " + (System.currentTimeMillis() - startTime) + " ms");

        // Get probabilities
        langlist = detector.getProbabilities();

    } catch (LangDetectException e) {
        System.err.println("Detection failed");
        e.printStackTrace();
    }

    System.out.println("Detected language: " + lang);
    for (Language s : langlist) {
        System.out.println(s);
    }

}
 

/**
 * {@inheritDoc}
 */
public NutchDocument filter(NutchDocument doc, Parse parse, Text url,
		CrawlDatum datum, Inlinks inlinks) throws IndexingException {
	if (conf == null) {
		throw new IndexingException("Not Yet Initialization.");
	}
	if (cause != null) {
		throw new IndexingException("Initialization Failed.", cause);
	}

	String lang = parse.getData().getParseMeta().get(Metadata.LANGUAGE);
	if (lang == null) {
		StringBuilder text = new StringBuilder();
		text.append(parse.getData().getTitle()).append(" ")
				.append(parse.getText());
		try {
			Detector detector = DetectorFactory.create();
			detector.setMaxTextLength(textsize_upper_limit);
			detector.append(text.toString());
			lang = detector.detect();
		} catch (LangDetectException e) {
			throw new IndexingException("Detection failed.", e);
		}
	}
	if (lang == null) lang = "unknown";

	doc.add("lang", lang);
	return doc;
}
 

public Detector getDetector() {
    return detector;
}
 

public void setDetector(Detector detector) {
    this.detector = detector;
}
 

public static String detect(String text) throws LangDetectException {

        Detector detector = DetectorFactory.create(0.5);
        detector.append(text);

        return detector.detect();
    }