import org.w3c.dom.Node;
import org.xml.sax.InputSource;

import jakarta.annotation.Resource;

/**
 * The {@code HtmlTransformer} class is responsible for transforming HTML responses
 * during the crawling process. It extracts data, identifies child URLs, and handles
 * character set encoding.
 * <p>
 * This class extends {@link AbstractTransformer} and utilizes various helper classes

    public void test_init_withIngestFactory() {
        ComponentUtil.register(ingestFactory, "ingestFactory");
        indexUpdater.init();
        assertNotNull(indexUpdater);
    }

    // Test destroy when crawling is not finished
    public void test_destroy_notFinished() {
        final List<Crawler> crawlerList = new ArrayList<>();
        final TestCrawler crawler = new TestCrawler();
        crawlerList.add(crawler);

        // We just verify that the crawl completes with EXIT_OK
        assertTrue("Crawler should complete", result == Constants.EXIT_OK);
    }

    // Test concurrent crawling
    public void test_doCrawl_concurrentWebAndData() {
        final List<String> executionOrder = Collections.synchronizedList(new ArrayList<>());

        WebFsIndexHelper mockWebHelper = new WebFsIndexHelper() {

        }
        return false;
    }

    /**
     * Create the parameters for extraction.
     * @param responseData The response data.
     * @param crawlingConfig The crawling configuration.
     * @return The parameters for extraction.
     */
    protected Map<String, String> createExtractParams(final ResponseData responseData, final CrawlingConfig crawlingConfig) {

        transformer.addTransformationRule("<[^>]+>", ""); // Remove HTML tags
        transformer.addTransformationRule("\\s+", " "); // Normalize whitespace

        // Simulate crawling response
        ResponseData responseData = new ResponseData();
        responseData.setUrl("http://example.com/page.html");
        responseData.setParentUrl("http://example.com/");

    public static final String ERRORS_design_file_is_unsupported_type = "{errors.design_file_is_unsupported_type}";

    /** The key of the message: Failed to create a crawling config at a wizard. */
    public static final String ERRORS_failed_to_create_crawling_config_at_wizard = "{errors.failed_to_create_crawling_config_at_wizard}";

    /** The key of the message: This feature is disabled. */

                    buf.append(e.getCause().getMessage());
                }
                logger.info(buf.toString());
            } else if (logger.isDebugEnabled()) {
                logger.debug("Crawling Access Exception at {}", url, e);
            }
        }

        // request header
        for (final Header header : requestHeaderList) {
            httpRequest.addHeader(header);
        }

to learn?'

  `Well, there was Mystery,' the Mock Turtle replied, counting
off the subjects on his flappers, `--Mystery, ancient and modern,
with Seaography:  then Drawling--the Drawling-master was an old
conger-eel, that used to come once a week:  HE taught us
Drawling, Stretching, and Fainting in Coils.'

  `What was THAT like?' said Alice.

  `Well, I can't show it you myself,' the Mock Turtle said:  `I'm

Caio Marcelo de Oliveira Filho <******@****.***> 1514852203 -0800

Russ Cox <******@****.***> 1410149331 -0400