It looks like this have something to do with the UrlRewriteModule. If I disables the UrlRewriteModule I see that the link is correct, contains the correct language code in the esplanguage parameter.

Someone have any idees on how to fix this problem?

We've had the same problem and reported this, but I've haven't found a solution so far. We can't use the CurrentPage.LinkUrl at all, most of the time the link is in english and not swedish although we have set swedish as main language and the page is in swedish, and other times the link is in mixed languages. In our case, the url is in english even if we have a property of type pagereference, were we specifically select a swedish page, and state that the page should be gotten in the swedish language. We have gotten around this problem by always using the urlresolver to get the link in the correct language. Although, it would be nice to simply be able to use the linkurl. I hope someone finds the problem soon.

var urlResolver = ServiceLocator.Current.GetInstance<UrlResolver>();
var url = urlResolver.GetVirtualPath(contentLink, languageBranch);

Hi.

I was able to solve this problem by creating a custom UrlRewriter, CustomHierarchicalUrlRewriteProvider that extends HierarchicalUrlRewriteProvider. In the HierarchicalUrlRewriteProvider there is a method ConvertToExternalInternal, it contains amongst other thing thes two lins of code:

if (UrlRewriteContext.Languages.Count == 1 || LanguageSelection.IsHostLanguageMatch(currentLanguageBranch))

    currentLanguageBranch = string.Empty;

After experiencing with the method explained by Timea Eriksson i found out that clearing the currentLanguageBranch caused the url to be generated with the wrong pagenames.

So i made a override of the method in my UrlRewiter replicating the behaviour of the original function except thos two lines, therby fixing the problem.

Does anyone know the concequences of removing those two lines?

Thanks for the report. We will have a look to see if we can fix this.