return htmlTagPattern;
    }

    /**
     * Gets the pattern used for extracting charset from meta tags.
     *
     * @return the meta charset pattern
     */
    public Pattern getMetaCharsetPattern() {
        return metaCharsetPattern;
    }

    /**
     * Sets the pattern used for extracting charset from meta tags.
     *

            return new String(data, Constants.UTF_8_CHARSET);
        }
    }

    /**
     * Gets the map of parser features.
     *
     * @return the feature map
     */
    public Map<String, String> getFeatureMap() {
        return featureMap;
    }

    /**
     * Sets the map of parser features.
     *
     * @param featureMap the feature map to set
     */

        public void endDocument() {
            // nothing
        }

        /**
         * Gets the parsed sitemap set.
         * @return the sitemap set
         */
        public SitemapSet getSitemapSet() {
            return sitemapSet;
        }

    }

    /**
     * Sets the preload size for checking file format.
     * @param preloadSize the preload size in bytes
     */

    /**
     * The priority of this URL relative to other URLs on your site. Valid
     * values range from 0.0 to 1.0. This value does not affect how your pages
     * are compared to pages on other sites—it only lets the search engines know
     * which pages you deem most important for the crawlers.
     *
     * The default priority of a page is 0.5.
     *
     * Please note that the priority you assign to a page is not likely to

        }
    }

    /**
     * Sets the filter cache expiration time.
     * @param filterCacheExpireAfterWrite The expiration time in seconds.
     */
    public void setFilterCacheExpireAfterWrite(final int filterCacheExpireAfterWrite) {
        this.filterCacheExpireAfterWrite = filterCacheExpireAfterWrite;
    }

    /**
     * Sets the maximum load size for the cache.

        sitemapList.remove(sitemap);
    }

    /**
     * Gets all sitemaps in this set as an array.
     * @return an array of sitemaps
     */
    public Sitemap[] getSitemaps() {
        return sitemapList.toArray(new Sitemap[sitemapList.size()]);
    }

    /**
     * Sets the type of this sitemap set.
     * @param type the type to set (URLSET or INDEX)
     */

        }
        connected = false;
    }

    /**
     * Gets the thread pool from the underlying client.
     *
     * @return The thread pool.
     */
    @Override
    public ThreadPool threadPool() {
        return client.threadPool();
    }

    /**
     * Gets the admin client for cluster and index operations.
     *
     * @return The admin client.
     */

     */
    @Override
    public void update(final List<OpenSearchAccessResult> accessResultList) {
        insertAll(accessResultList, OpType.INDEX);
    }

    /**
     * Gets the count of access results for a specific session.
     *
     * @param sessionId The session ID.
     * @return The number of access results for the session.
     */
    @Override

        idProp.setValue(target, id);
    }

    /**
     * Gets the OpenSearch index name.
     *
     * @return The index name.
     */
    public String getIndex() {
        return index;
    }

    /**
     * Sets the OpenSearch index name.
     *
     * @param index The index name.
     */
    public void setIndex(final String index) {

            extractData.putValue(name, value);
        }
    }

    /**
     * Gets the timeout for PDF extraction in milliseconds.
     * @return the timeout in milliseconds
     */
    public long getTimeout() {
        return timeout;
    }

    /**
     * Sets the timeout for PDF extraction in milliseconds.
     * @param timeout the timeout in milliseconds
     */