Today is about getting an automated pinger setup on our blog so
you can notify indexers like Google Blogsearch about your new
posts.
We will set it up with a simple event handler that checks if the
blog has any urls to ping. We will add the urls to the blog root
itself by using the related links control:
This creates a simple xml structure which is saved on our
document. So everytime we publish a new blogpost, we will simple
look up the serviceUrls property recursively and ping all the urls
in there.
First, a little snippet to get the value recursively:
private string GetValueRecursively(string alias, int
nodeId)
{
Document n = new Document(nodeId);
Property p = n.getProperty(alias);
if (p != null &&
!string.IsNullOrEmpty(p.Value.ToString()))
return
p.Value.ToString();
else if (n.Parent != null)
return
GetValueRecursively(alias, n.Parent.Id);
return string.Empty;
}
So now we can just pass the alias of the property and the Id of
the document where we should start looking. If nothing is found it
will continue up the content tree untill it finds anything or runs
out of parent documents to look in.
As soon as we've done this, we will query the related links xml.
It is structure with a parent <links> element containing
child <link> elements with attributes on storing urls, names
and other meta data. We will also setup the blog name and url which
will be send off to the pinged urls.
string blogName = GetValueRecursively("blogName",
sender.Id);
string blogUrl = umbraco.library.NiceUrlFullPath(sender.Id);
string currentDomain =
HttpContext.Current.Request.ServerVariables["SERVER_NAME"].ToLower();
if (!UmbracoSettings.UseDomainPrefixes)
blogUrl = "http://"
+ currentDomain + blogUrl;
foreach (XmlNode link in xd.SelectNodes("//link [@type =
'external']"))
{
string ping =
link.Attributes["link"].Value;
//Log.Add(LogTypes.Debug, sender.Id, ping + "
n:" + blogName + " u:" + blogUrl);
PingService(ping, blogName, blogUrl);
}
This is all put together in an event handler that will fire off
after any documents with the contenttype alias = "BlogPost" has
been published. The actual ping is handled in the method
PingService which simply puts together a chunk of xml and sends it
off to the url as a normal HttpWebRequest.
Done, we can now ping any service that supports the xmlrpc blog
pings.