public void initiateSeedToFrontier()
{
List<HTMLLink> seed = new List<HTMLLink>();
HTMLLink stack = new HTMLLink(); // XML loader fejler på disse links
stack.hrefLink = new Uri("http://stackoverflow.com/questions/10984336/net-using-using-blocks-vs-calling-dispose");
HTMLLink ww2 = new HTMLLink();
ww2.hrefLink = new Uri("http://en.wikipedia.org/wiki/World_War_II");
HTMLLink mma = new HTMLLink();
mma.hrefLink = new Uri("http://en.wikipedia.org/wiki/MMA");
HTMLLink newz = new HTMLLink();
newz.hrefLink = new Uri("http://raid1.newz.dk/western-digital-annoncerer-10-tb-harddisk");
HTMLLink china = new HTMLLink();
china.hrefLink = new Uri("http://en.wikipedia.org/wiki/Republic_of_China_%281912%E2%80%9349%29");
seed.Add(ww2);
seed.Add(mma);
seed.Add(china);
//seed.Add(newz);
foreach (HTMLLink seedUri in seed)
{
frontier.Enqueue(seedUri);
}
}
public List<HTMLLink> FindLinks(HTMLLink currentWebsite)
{
List<HTMLLink> listOfLinks = new List<HTMLLink>();
MatchCollection linkMatches;
string hostUrl = currentWebsite.hrefLink.Scheme + "://" + currentWebsite.hrefLink.Host;
string protocolAsString = hostUrl.Replace(currentWebsite.hrefLink.Host, "");
StreamReader streamReader = new StreamReader(getDataStream(currentWebsite.hrefLink));
while (!streamReader.EndOfStream)
{
linkMatches = Regex.Matches(streamReader.ReadLine(), @"(<a.*?>.*?</a>)", RegexOptions.Singleline);
foreach (Match link in linkMatches)
{
string htmlString = link.Groups[1].Value;
Match linkUrl = Regex.Match(htmlString, @"href=\""(.*?)\""", RegexOptions.Singleline);
if (linkUrl.Success)
{
string temporaryString = "";
HTMLLink htmlLink = new HTMLLink();
htmlLink.title = Regex.Replace(htmlString, @"\s*<.*?>\s*", "", RegexOptions.Singleline);
temporaryString = linkUrl.Groups[1].Value;
if (temporaryString.StartsWith("http"))
{
//link is correctly formatted
}
else if (temporaryString.StartsWith("//"))
{
temporaryString = temporaryString.Remove(0, 2);
temporaryString = protocolAsString + temporaryString;
}
else if (temporaryString.StartsWith("/"))
{
temporaryString = hostUrl + temporaryString;
}
else temporaryString = hostUrl + "/" + temporaryString;
//Få dette til at fjerne alle uønskede links og filtyper .avi, .php, etc.
Match wrongUrl = Regex.Match(temporaryString, @"\.php", RegexOptions.Singleline);
if (!temporaryString.Contains("#") && temporaryString.Length > 0 && !wrongUrl.Success)
{
htmlLink.hrefLink = new Uri(temporaryString);
listOfLinks.Add(htmlLink);
}
}
}
}
return listOfLinks;
}
/// <summary>
/// Filters the urls by the information collected in the robots.txt file
/// </summary>
/// <param name="link">The link to filter</param>
/// <returns>true if the link should be crawled, false otherwise</returns>
public bool FilterUrlsByRobotsTxt(HTMLLink link)
{
if (disallowedList.Contains(link.hrefLink.AbsolutePath))
{
if (allowedList.Contains(link.hrefLink.AbsolutePath))
{
return true;
}
else return false;
}
else
{
return true;
}
}