static void CrawlLinks()
{
//takes the next page from the stack and extracts all of the links, adds non visted & allowed pages to the stack
Link page = toVist.Pop(); //get the next page to view
//page may have been viewed already (duplicate links on the same page), if not crawl if allowed
if(!LinkVisted(page) && CanVisit(page))
{
Console.WriteLine(page.linkString);
visited.Add(page);
try
{
String html = new WebClient().DownloadString(page.linkString); //download the HTML
MatchCollection regexLinks = Regex.Matches(html, @"((<a.*?>.*?</a>)|(<A.*?>.*?</A))", RegexOptions.Singleline); //get the ahrefs
//put all the unvisited links onto the stack
for(int i = 0; i < regexLinks.Count; i++)
{
Link link = new Link(ReturnPageLink(regexLinks[i].Value), page.path, page.linkString);
if (link.linkString != null) { toVist.Push(link); }
}
}
catch(WebException ex)
{
//maybe not found or server error, no need to fail if so
}
}//if the link hasn't already been visited
}
static List<Link> visited = new List<Link>(); //the links that have been visited
#endregion Fields
#region Methods
static Boolean CanVisit(Link link)
{
//takes a link and determines whether we can visit it, based on the same domain rule and by obeying robots.txt
if(link.linkString.Contains(seed) && !isDisallowed(link)) //check whether this link is from the correct domain
{
List<String> robots = new List<string>();
//get a list of all the disallowed links from robots.txt, if it hasn't been visited
if (!RobotsChecked(link.robots))
{
try
{
robots = Regex.Split(new WebClient().DownloadString(link.robots), "\r\n").ToList();
robotsDownloaded.Add(link.robots);
}
catch(System.Net.WebException ex) //may not be found
{
return true; //if no robots.txt then assume we can visit
}
disallowed.AddRange((from line in robots
where line.Contains("Disallow")
select link.path + @"/" + line.Replace(@"Disallow:","").Replace(" ","").Substring(1)).ToList());
}
//check whether this passed in link is in the disallowed list
var exists = (from lnk in disallowed
where link.linkString.Contains(lnk)
select lnk).Any();
return !exists;
}
else
{
return false; //a different domain or is disallowed
}
}
static Boolean isDisallowed(Link link)
{
return (from lnk in disallowed
where lnk == link.path
select lnk).Any();
}
static Boolean LinkVisted(Link link)
{
return (from lnk in visited
where link.linkString == lnk.linkString
select lnk).Any();
}
