/**
* Parse XML that contains a Sitemap Index. Example Sitemap Index:
*
* <p><?xml version="1.0" encoding="UTF-8"?> <sitemapindex
* xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"> <sitemap>
* <loc>http://www.example.com/sitemap1.xml.gz</loc> <lastmod>2004-10-01T18:23:17+00:00</lastmod>
* </sitemap> <sitemap> <loc>http://www.example.com/sitemap2.xml.gz</loc>
* <lastmod>2005-01-01</lastmod> </sitemap> </sitemapindex>
*
* @param url - URL of Sitemap Index
* @param nodeList
*/
private SiteMapIndex parseSitemapIndex(URL url, NodeList nodeList) {
LOG.debug("Parsing Sitemap Index");
SiteMapIndex sitemapIndex = new SiteMapIndex(url);
sitemapIndex.setType(SitemapType.INDEX);
// Loop through the <sitemap>s
for (int i = 0; i < nodeList.getLength() && i < MAX_URLS; i++) {
Node firstNode = nodeList.item(i);
if (firstNode.getNodeType() == Node.ELEMENT_NODE) {
Element elem = (Element) firstNode;
String loc = getElementValue(elem, "loc");
// try the text content when no loc element
// has been specified
if (loc == null) {
loc = elem.getTextContent().trim();
}
try {
URL sitemapUrl = new URL(loc);
String lastmod = getElementValue(elem, "lastmod");
Date lastModified = SiteMap.convertToDate(lastmod);
// Right now we are not worried about sitemapUrls that point
// to different websites.
SiteMap s = new SiteMap(sitemapUrl, lastModified);
sitemapIndex.addSitemap(s);
LOG.debug(" {}. {}", (i + 1), s);
} catch (MalformedURLException e) {
LOG.trace("Don't create an entry with a bad URL", e);
LOG.debug("Bad url: [{}]", loc);
}
}
}
sitemapIndex.setProcessed(true);
return sitemapIndex;
}
/**
* Set when this URL was last modified.
*
* @param lastModified
*/
public void setLastModified(String lastModified) {
this.lastModified = SiteMap.convertToDate(lastModified);
}
