Hi Marc,

The way the redirect service works is that when a page is requested, before the page is returned, the url is looked up in the database. This might be something like www.example.com\abc. If it finds this and there is a new url to redirect to, it will redirect to that url and not display the umbraco 404.

If I understand you correctly, are you saying that I should let the redirect service code run, but if it does not find, then I should let umbraco redirect to a proper 404 page using the TryFindContent.

I am still a bit confused on how TryFindContent works. Once I implement it, how would I get it to redirect to a 404 page?

Thanks for the help, Saied

Marc,

Interestingly, I was just looking at the code that handles the redirects from old urls to new urls and it does use the TryContentFinder. The ApplicationStarting event has this line:

ContentLastChanceFinderResolver.Current.SetFinder(new NotFoundContentFinder());

How can I modify NotFoundContentFinder to redirect to a custom 404 page( a node that I would create under the root of each site for example) if it does not find an entry in the database. Here is the code:

    public class NotFoundContentFinder : IContentFinder
{
    private static Guid SystemUserId = new Guid("a23d1aec-b91a-4fa8-8e9e-f65966c68389");

    public bool TryFindContent(PublishedContentRequest contentRequest)
    {
        if (!contentRequest.Is404)
        {
            return false;
        }

        var domainName = HttpContext.Current.Request.Url.Authority;
        var slug = HttpContext.Current.Request.Path;

        if (slug.StartsWith("/error.aspx", StringComparison.OrdinalIgnoreCase))
        {
            slug = HttpUtility.UrlDecode(HttpContext.Current.Request.QueryString.ToString()).Replace("404;", "");
            slug = slug.Substring(slug.IndexOf(domainName) + domainName.Length);
        }
        if (slug.StartsWith(":80", StringComparison.OrdinalIgnoreCase))
        {
            slug = slug.Substring(3);
        }
        if (slug.StartsWith(":443", StringComparison.OrdinalIgnoreCase))
        {
            slug = slug.Substring(4);
        }

        var redirectSet = RedirectFromLookup(domainName, slug) || RedirectSctFiles(domainName, slug) || RedirectBullydogFiles(domainName, slug);
        if (redirectSet)
        {
            HttpContext.Current.Server.ClearError();
        }
        return redirectSet;
    }

    private static bool RedirectFromLookup(string domainName, string slug)
    {
        Redirect redirect;
        try
        {
            var svc = new RedirectService();
            redirect = svc.Add(new Redirect
            {
                Slug = slug,
                SiteName = domainName,
                IsActive = true,
                RedirectType = RedirectType.Permanent
            }, SystemUserId);
        }
        catch (Exception )
        {
            //TODO: log ex
            return false;
        }

        if (redirect == null || redirect.RedirectType == null || string.IsNullOrWhiteSpace(redirect.TargetUrl))
        {
            return false;
        }

        switch (redirect.RedirectType.Value)
        {
            case RedirectType.Permanent:
                HttpContext.Current.Response.RedirectPermanent(redirect.TargetUrl);
                HttpContext.Current.Response.End();
                break;
            case RedirectType.Temporary:
                HttpContext.Current.Response.Redirect(redirect.TargetUrl);
                HttpContext.Current.Response.End();
                break;
        }
        return true;
    }

    private static bool RedirectSctFiles(string domainName, string slug)
    {
        if (domainName.Contains("sctflash.com") && slug.StartsWith("/sctfiles/", StringComparison.OrdinalIgnoreCase))
        {
            slug = "/sctfiles/" + slug.Substring(10);
            HttpContext.Current.Response.RedirectPermanent("http://cdn.derivesystems.com" + SlugWithRandomQueryString(slug));
            HttpContext.Current.Response.End();
            return true;
        }
        return false;
    }

    private static bool RedirectBullydogFiles(string domainName, string slug)
    {
        if (domainName.Contains("bullydog.com") && slug.StartsWith("/updateagent/", StringComparison.OrdinalIgnoreCase))
        {
            HttpContext.Current.Response.RedirectPermanent("http://dv-prod-wamp01.cloudapp.net" + SlugWithRandomQueryString(slug));
            HttpContext.Current.Response.End();
            return true;
        }
        return false;
    }

    private static string SlugWithRandomQueryString(string slug)
    {
        if (string.IsNullOrWhiteSpace(slug))
        {
            return slug;
        }
        var parts = slug.Split("?".ToCharArray(), StringSplitOptions.None);
        if (parts.Length == 0)
        {
            return slug;
        }
        return string.Format("{0}?{1}", parts[0], Guid.NewGuid());
    }
}

UPDATE:

Marc,

After looking at it some more, I thought if the redirectSet returns false, meaning it did not find an entry in the database or lookup, I could redirect to a not-found page that I create in my content structure on my root. What do you think?

  if (redirectSet)
            {
                HttpContext.Current.Server.ClearError();
            }
            else
            {
                HttpContext.Current.Response.RedirectPermanent("/not-found", true);
            }

My only issue is what if the page changes name, like from not-found to uh-oh. I guess I could store the 404 page in a appsetting or is there another option?

Hi Marc,

I did the following steps and was wondering if it was correct. I put the following in the ApplicationStarting event:

ContentFinderResolver.Current.InsertTypeBefore<ContentFinderByNotFoundHandlers, NotFoundContentFinder>();

ContentLastChanceFinderResolver.Current.SetFinder(new PageNotFoundContentFinder());

If I understand the above correctly, does the InsertTypeBefore method insert the NotFoundContentFinder before the PageNotFoundContentFinder so that runs first. If the NotFoundContentFinder does not return true, it then goes to the PageNotFoundContentFinder. Is this correct?

Another question I had was is there a default content finder? What I mean is, it looks like what is happening is:

1. Run the custom redirect first. If redirect found, return.
2. If page exists in website (not 404), go to it - where does this come in (is there a default for this?)
3. if no redirect found and 404 is thrown, run the Last Chance finder.

No. 2 is what is confusing. I know that no matter what page I go to, t will always run the code for number one first. So if it does not find a redirect and the page exists, I was just wondering if there was a default content finder that runs next that takes the user to the page if it exists.

Have you had any luck running the Page Not Found Manager along the 301 Url Tracker. Now that you have helped me get this working, it would be nice to use the 301 Url Tracker to be able to setup redirects through Umbraco, but I am not sure if that would break the redirect one. Also, I wouldn't want the 301 Url Tracker to handle 404s because I have the Page Not Found Manager for that.

Thanks for the great help, Saied

Hi Saied

Yes there are a number of default IContentFinders that ship with Umbraco, these form the 'Umbraco pipeline' because each one is executed one after another until a piece of content is found.

You can see the code here:

https://github.com/umbraco/Umbraco-CMS/blob/82954b6b6f303d6ae2d704e4d606840ada841a57/src/Umbraco.Web/WebBootManager.cs

ContentFinderResolver.Current = new ContentFinderResolver(
                ServiceProvider, LoggerResolver.Current.Logger,
                // all built-in finders in the correct order, devs can then modify this list
                // on application startup via an application event handler.
                typeof(ContentFinderByPageIdQuery),
                typeof(ContentFinderByNiceUrl),
                typeof(ContentFinderByIdPath),

                // these will be handled by ContentFinderByNotFoundHandlers so they can be enabled/disabled
                // via the config file... soon as we get rid of INotFoundHandler support, we must enable
                // them here.
                //typeof (ContentFinderByNiceUrlAndTemplate),
                //typeof (ContentFinderByProfile),
                //typeof (ContentFinderByUrlAlias),

                // note: that one should run *after* NiceUrlAndTemplate, UrlAlias... but at the moment
                // it cannot be done - just make sure to do it properly in v8!
                typeof(ContentFinderByRedirectUrl),

                // implement INotFoundHandler support... remove once we get rid of it
                typeof(ContentFinderByNotFoundHandlers)
            );

You can see the last one in the default list is 'ContentFinderByNotFoundHandlers' and we inserted your custom redirect one before this.

But before your Redirect are other finders, eg ContentFinderByNiceUrl, that will find content for existing Umbraco Content, before it gets a chance to fall through to your redirect, if that makes sense ?