public void test_getText() {
        final InputStream in = ResourceUtil.getResourceAsStream("extractor/msoffice/test.ppt");
        final String content = msPowerPointExtractor.getText(in, null).getContent();
        CloseableUtil.closeQuietly(in);
        logger.info(content);
        assertTrue(content.contains("テスト"));
    }

    public void test_getText_null() {
        try {
            msPowerPointExtractor.getText(null, null);

     * @return An ExtractData object containing the filename as content.
     */
    @Override
    public ExtractData getText(final InputStream in, final Map<String, String> params) {
        if (in == null) {
            throw new CrawlerSystemException("The inputstream is null.");
        }
        final String content = params.getOrDefault(ExtractData.RESOURCE_NAME_KEY, StringUtil.EMPTY);
        try {

import org.codelibs.fess.suggest.exception.SuggesterException;
import org.opensearch.action.get.GetResponse;
import org.opensearch.common.collect.Tuple;
import org.opensearch.common.xcontent.XContentType;
import org.opensearch.common.xcontent.json.JsonXContent;
import org.opensearch.core.xcontent.XContentBuilder;
import org.opensearch.index.IndexNotFoundException;
import org.opensearch.transport.client.Client;

    /**
     * Parses the charset from the content string.
     *
     * @param content the content to parse
     * @return the parsed charset name, or null if not found
     */
    protected String parseCharset(final String content) {
        final Pattern pattern = Pattern.compile("; *charset *= *([a-zA-Z0-9\\-_]+)", Pattern.CASE_INSENSITIVE);
        final Matcher matcher = pattern.matcher(content);
        if (matcher.find()) {

import java.util.Locale;
import java.util.Map;
import java.util.concurrent.TimeUnit;
import java.util.regex.Pattern;

import javax.net.ssl.SSLContext;

import org.apache.commons.io.output.DeferredFileOutputStream;
import org.apache.commons.lang3.SystemUtils;
import org.apache.http.Header;
import org.apache.http.HttpEntity;
import org.apache.http.HttpHost;
import org.apache.http.HttpResponse;

        final InputStream in = ResourceUtil.getResourceAsStream("extractor/lha/test.lzh");
        final String content = lhaExtractor.getText(in, null).getContent();
        CloseableUtil.closeQuietly(in);
        logger.info(content);
        assertTrue(content.contains("テスト"));
        assertTrue(content.contains("テキスト"));
    }

    public void test_getText_maxSize() throws IOException {

    }

    /**
     * Returns the default maximum content length.
     * @return The default maximum content length in bytes.
     */
    public long getDefaultMaxLength() {
        return defaultMaxLength;
    }

    /**
     * Sets the default maximum content length.
     * @param defaultMaxLength The default maximum content length to set.
     */

    }

    public void test_getText() {
        final InputStream in = ResourceUtil.getResourceAsStream("extractor/test.txt");
        final String content = textExtractor.getText(in, null).getContent();
        CloseableUtil.closeQuietly(in);
        logger.info(content);
        assertTrue(content.contains("テスト"));
    }

    public void test_getText_null() {
        try {
            textExtractor.getText(null, null);

            ExtractData data = emlExtractor.getText(in, null);
            final String content = data.getContent();
            logger.info(content);
            assertTrue(content.contains("Exkursion und Museumsbesuch"));
            assertTrue(content.contains("Fahrt nach Baruth"));
            assertTrue(content.contains("Technische Universität"));
        }
    }

    public void test_getText_null() {

        final String content = data.getContent();
        CloseableUtil.closeQuietly(in);
        logger.info(content);
        assertTrue(content.contains("テスト"));
        assertEquals("タイトル", data.getValues("title")[0]);
    }

    public void test_getHtml_empty() {
        final InputStream in = new ByteArrayInputStream("".getBytes());
        final String content = htmlExtractor.getText(in, null).getContent();