protected void processFullURL(
final List<BotLink> linksForProcessing, final Link tkLink, final String u) {
String scheme = "";
String host = "";
String path = "";
String query = "";
final Matcher m = SIMPLE_LINK.matcher(u);
while (m.find()) {
if (m.groupCount() >= 2) {
scheme = m.group(1).trim();
final String tmp = m.group(2).trim();
final Matcher m2 = SIMPLE_LINK2.matcher(tmp);
while (m2.find()) {
if (m2.groupCount() >= 2) {
host = m2.group(1).trim();
// At this point we should have a path
// Remove the 'query' section if available
final String tmp2 = m2.group(2).trim();
if (tmp2.indexOf('?') > 0) {
final String wQuery = tmp2.substring(tmp2.indexOf('?') + 1);
path = tmp2.substring(0, tmp2.indexOf('?'));
query = wQuery;
} else {
path = tmp2;
}
} // End of the if //
}
}
} // End of the while
if (scheme.length() > 0 && host.length() > 0) {
// Create a link for for further processing //
final BotLink link = new BotLink();
link.setHost(host);
if (path.length() > 0) {
link.setPath("/" + path);
} // End of the if //
link.setScheme(scheme);
link.setQuery(query);
link.setLink(tkLink);
logger.info("Attempt to process and add to queue / link , link=" + link);
linksForProcessing.add(link);
} // End of the if //
} // End of method //
protected void writeFileAndSave(
final BotLink link,
final Session session,
final Metadata metadata,
final String document,
final String txtOnlyDoc) {
// Link added to processing, write the information to file
final TextHelpers text = new TextHelpers();
final String dir = text.baseDirectory(link);
final String fpathText = text.mangleText(link);
final String fpathFull = text.mangleFull(link);
final File chkdir = new File(OctaneCrawlerConstants.CRAWLER_HOME + "/" + dir);
final boolean res = chkdir.mkdirs();
final long wstart = System.currentTimeMillis();
final File txtfile = new File(OctaneCrawlerConstants.CRAWLER_HOME + "/" + fpathText);
final File fullfile = new File(OctaneCrawlerConstants.CRAWLER_HOME + "/" + fpathFull);
try {
// Write text version and full file //
new IO<Void>()
.w(
txtfile.getAbsolutePath(),
new Fx<PrintWriter>() {
public void $(final PrintWriter o, final int idx) {
o.println(txtOnlyDoc);
}
});
new IO<Void>()
.w(
fullfile.getAbsolutePath(),
new Fx<PrintWriter>() {
public void $(final PrintWriter o, final int idx) {
o.println(document);
}
});
final long diff = System.currentTimeMillis() - wstart;
logger.info(
"Result after directory and writes / mk:" + chkdir + " res=" + res + " procTime=" + diff);
final BotCrawlerDAO dao = new BotCrawlerDAO();
final BotCrawlerLink persistLink = new BotCrawlerLink();
persistLink.setHost(link.getHost());
persistLink.setUrl(String.valueOf(link.toBuilder()));
persistLink.setTitle(metadata.get("title"));
persistLink.setDescr(metadata.get("description"));
persistLink.setPath(fpathText);
persistLink.setLinkcount(link.getNumberLinks());
persistLink.setSource(Thread.currentThread().getName());
if (NullRef.hasValue(link.getStatusline())) {
persistLink.setStatusline(link.getStatusline());
persistLink.setStatus(link.getCode());
} // Check the status line //
if (link.getLink() != null) {
if (link.getLink().getText() != null) {
persistLink.setLinktext(link.getLink().getText());
}
}
dao.createLink(session, persistLink);
} catch (final Throwable e) {
e.printStackTrace();
} // End of try - catch //
} // End of method write file and persist //
public List<BotLink> parse(
final BotLink origLink, final URIBuilder lastBuilder, final String document) {
final SessionFactory sf = (SessionFactory) ctx.getBean("sessionFactory");
final Session session = sf.openSession();
try {
final InputStream input = new ByteArrayInputStream(document.getBytes());
final LinkContentHandler linkHandler = new LinkContentHandler();
final ContentHandler textHandler = new BodyContentHandler();
final ToHTMLContentHandler toHTMLHandler = new ToHTMLContentHandler();
final TeeContentHandler teeHandler =
new TeeContentHandler(linkHandler, textHandler, toHTMLHandler);
final Metadata metadata = new Metadata();
final ParseContext parseContext = new ParseContext();
final HtmlParser parser = new HtmlParser();
parser.parse(input, teeHandler, metadata, parseContext);
final String titleOfPage = metadata.get("title");
// For analytical data, ignore pages that don't have titles
if (!NullRef.hasValue(titleOfPage)) {
logger.warn("Warning, invalid title for page, EXITING logic, link=" + origLink);
return null;
}
// Continue with parsing //
final List<BotLink> linksForProcessing = new ArrayList<BotLink>();
final Set<String> urls = new HashSet<String>();
int fullLinkCount = 0;
for (final Link link : linkHandler.getLinks()) {
fullLinkCount++;
}
int linkcount = 0;
// Loop through the links on the page
// And add a set number to the queue.
final Random rchk = new Random(System.currentTimeMillis());
final List<Link> linksFromPage = linkHandler.getLinks();
Collections.shuffle(linksFromPage);
for (final Link link : linksFromPage) {
// Add a 30% chance of adding this link
final double rval = rchk.nextDouble();
final boolean okToAdd = rval > 0.65;
if (okToAdd && link.getUri() != null) {
linkcount++;
if (linkcount > MAX_LINKS_PAGE) {
// Only process a given number of links on a page //
break;
} // End of if max reached
final String fullURL = this.fullURL(link, lastBuilder, urls);
if (fullURL != null) {
try {
this.processFullURL(linksForProcessing, link, fullURL);
} catch (final Throwable te) {
te.printStackTrace();
}
}
} // End of the if //
} // End of the for through the links //
// Parse the available URLS //
logger.info(
"In Web Parser for "
+ lastBuilder
+ " # availableNumberOfLinks="
+ urls.size()
+ " fullLinkCount="
+ fullLinkCount);
// Persist the current link //
origLink.setNumberLinks(fullLinkCount);
this.writeFileAndSave(origLink, session, metadata, document, textHandler.toString());
processLinksForQueue(linksForProcessing);
return linksForProcessing;
} catch (final Throwable e) {
e.printStackTrace();
} finally {
if (session != null) {
session.close();
}
} // End of the try - catch //
return null;
} // End of the method //
