@Override
public boolean shouldVisit(WebURL url) {
// Don't crawl non-HTML pages
String href = url.getURL().toLowerCase(); // Turns http://www.ics.uci.edu/SomePage.PHP ->
// http://www.ics.uci.edu/somepage.php
if (FILTERS.matcher(href).matches()) // filter using file extension
return false;
// Only crawl within the domain of the seed URL
String currentUrlDomain = URLHelper.getDomain(url.getURL());
String seedUrlDomain = URLHelper.getDomain(this.params.getSeedUrl());
if (currentUrlDomain == null || !currentUrlDomain.endsWith(seedUrlDomain)) return false;
// Don't crawl the same pages too many times (avoid infinite loops)
if (!stats.intendToVisit(url.getURL())) return false;
return true;
}
public boolean intendToVisit(String url) {
// Determine if the page has been visited too many times (might be infinite loop)
String urlWithoutQuery = URLHelper.removeQuery(url);
if (urlWithoutQuery == null) return true;
// Get the current visit-intent count
int count = 1;
if (this.pagesToCrawl.containsKey(urlWithoutQuery))
count = this.pagesToCrawl.get(urlWithoutQuery);
if (count >= 20)
return false; // visit single page at most 20 times (with different query strings)
// Update with another intent
this.pagesToCrawl.put(urlWithoutQuery, count + 1);
return true;
}
