CloseableUtil.closeQuietly(bis);
        assertEquals("UTF-8", encoding);
    }

    public void test_getEncoding_utf16lebom() {
        final InputStream in = ResourceUtil.getResourceAsStream("extractor/xml/test_utf16lebom.xml");
        final BufferedInputStream bis = new BufferedInputStream(in);
        final String encoding = xmlExtractor.getEncoding(bis);

        }
        try {
            return new ExtractData(new String(InputStreamUtil.getBytes(in), getEncoding()));
        } catch (final Exception e) {
            throw new ExtractException(e);
        }
    }

    public String getEncoding() {
        return encoding;
    }

    public void setEncoding(final String encoding) {
        this.encoding = encoding;
    }

        }
        matcher.appendTail(sb);
        return sb.toString().replaceAll("\\s+", " ").trim();
    }

    public String getEncoding() {
        return encoding;
    }

    public void setEncoding(final String encoding) {
        this.encoding = encoding;
    }

    /**
     * @return Returns the preloadSizeForCharset.
     */

                    if (pageEnc != null) {
                        enc = pageEnc;
                    } else if (urlQueue.getEncoding() != null) {
                        enc = urlQueue.getEncoding();
                    }
                }
            }
        } else {
            enc = fessConfig.getCrawlerDocumentFileNameEncoding();
        }

        return listFiles;
    }

    public void setListFiles(boolean listFiles) {
        this.listFiles = listFiles;
    }

    public String getEncoding() {
        return encoding;
    }

    public void setEncoding(String encoding) {
        this.encoding = encoding;
    }

    public boolean isFork() {
        return fork;
    }

    public void setFork(boolean fork) {

        final String encoding = htmlExtractor.getEncoding(bis);
        CloseableUtil.closeQuietly(bis);
        assertEquals("UTF-8", encoding);
    }

    public void test_getEncoding_sjis() {
        final InputStream in = ResourceUtil.getResourceAsStream("extractor/test_sjis.html");
        final BufferedInputStream bis = new BufferedInputStream(in);
        final String encoding = htmlExtractor.getEncoding(bis);

        // create input source
        final InputSource is = new InputSource(new ByteArrayInputStream(accessResultData.getData()));
        if (StringUtil.isNotBlank(accessResultData.getEncoding())) {
            is.setEncoding(accessResultData.getEncoding());
        }

        // create handler
        final DocHandler handler = new DocHandler();

        // create a sax instance

        return optimize;
    }

    public void setOptimize(boolean optimize) {
        this.optimize = optimize;
    }

    @Nullable
    public String getEncoding() {
        return encoding;
    }

    public void setEncoding(@Nullable String encoding) {
        this.encoding = encoding;
    }

    public boolean isForce() {
        return force;
    }

    String getTransformerName();

    void setTransformerName(String transformerName);

    byte[] getData();

    String getDataAsString();

    void setData(byte[] data);

    String getEncoding();

    void setEncoding(String encoding);

        final InputStream in = ResourceUtil.getResourceAsStream("extractor/test_utf8.html");
        final BufferedInputStream bis = new BufferedInputStream(in);
        final String encoding = htmlXpathExtractor.getEncoding(bis);
        CloseableUtil.closeQuietly(bis);
        assertEquals("UTF-8", encoding);
    }

    public void test_getEncoding_sjis() {