Examples with CrawlerException org.asqatasun.crawler.exception.CrawlerException used on opensource projects

Example 1 with CrawlerException

use of org.asqatasun.crawler.exception.CrawlerException in project Asqatasun by Asqatasun.

the class AsqatasunCrawlJob method initializeCrawlContext.

/**
     * 
     * @param url
     * @param crawlParameterSet
     * @param heritrixFileName
     * @return
     */
private File initializeCrawlContext(Collection<String> urlList, Set<Parameter> crawlParameterSet, String heritrixFileName) {
    buildOutputDirectory();
    BufferedReader in = null;
    FileWriter fw = null;
    try {
        LOGGER.debug("crawlConfigFilePath: " + crawlConfigFilePath + " for copy");
        String filepath = crawlConfigFilePath + "/" + heritrixFileName;
        DocumentBuilderFactory docFactory = DocumentBuilderFactory.newInstance();
        DocumentBuilder docBuilder = docFactory.newDocumentBuilder();
        Document doc = docBuilder.parse(filepath);
        if (LOGGER.isDebugEnabled()) {
            LOGGER.debug("filepath : " + filepath);
            for (Parameter param : crawlParameterSet) {
                LOGGER.debug(param.getParameterElement().getParameterElementCode() + " " + param.getValue());
            }
        }
        doc = setOptionToDocument(urlList, crawlParameterSet, doc);
        //write the content into xml file
        TransformerFactory transformerFactory = TransformerFactory.newInstance();
        Transformer transformer = transformerFactory.newTransformer();
        DOMSource source = new DOMSource(doc);
        String resultFileName = currentJobOutputDir.getPath() + "/" + heritrixFileName;
        StreamResult result = new StreamResult(new File(resultFileName));
        transformer.transform(source, result);
    } catch (IOException | ParserConfigurationException | SAXException ex) {
        LOGGER.error(ex);
        throw new CrawlerException(ex);
    } catch (TransformerConfigurationException ex) {
        LOGGER.error(ex);
        throw new CrawlerException(ex);
    } catch (TransformerException ex) {
        LOGGER.error(ex);
        throw new CrawlerException(ex);
    } finally {
        if (in != null) {
            try {
                in.close();
            } catch (IOException ex) {
                LOGGER.error(ex);
                throw new CrawlerException(ex);
            }
        }
        if (fw != null) {
            try {
                fw.close();
            } catch (IOException ex) {
                LOGGER.error(ex);
                throw new CrawlerException(ex);
            }
        }
    }
    return new File(currentJobOutputDir.getPath() + "/" + heritrixFileName);
}