The character map is not used during routing of a request request (that is it will not be used to e.g. map 'ä' to 'a' during segment parsing). It is used when an url segment is created. So e.g. given that you have a regexp definied in

UrlSegmentOptions.ValidUrlCharacters 

that does not allow for character 'ä' it will then use the charater map to see if the unallowed character should be replaced with something e.g. 'ä' => 'a' when the url segment is created

Thanks John for clarifing that.

Is there a similar character map for the segment parsing? Or any other way to replas characters before segment parsing?

There is no similar map used at routing. What you probably could do is hook up an event handler to 

EPiServer.Web.Routing.IContentRouteEvents.RoutingContent

On the eventargs (of type RoutingEventArgs) you could access the path as args.RoutingSegmentContext.RemainingPath and update it (e.g. by replacing 'ä' => 'a')

Thanks again Johan!

I will look in to that possible solution.

This is the code for my sulotion based hon Johans sugestion. If anybody in the future should need it =)

[InitializableModule]
[ModuleDependency(typeof(EPiServer.Web.InitializationModule),
typeof(EPiServer.Commerce.Initialization.InitializationModule))]
public class EventInitialization : IInitializableModule
{
public void Initialize(InitializationEngine context)
{
var contentRouteEvents = ServiceLocator.Current.GetInstance<IContentRouteEvents>();
contentRouteEvents.RoutingContent += RoutingEvents_RoutingContent;
}

public void Uninitialize(InitializationEngine context)
{

}

private static void RoutingEvents_RoutingContent(object source, RoutingEventArgs e)
{
e.RoutingSegmentContext.RemainingPath = e.RoutingSegmentContext.RemainingPath.
Replace("ö", "o").
Replace("Ö", "o").
Replace("å", "a").
Replace("Å", "a").
Replace("ä", "a").
Replace("Ä", "a");
}
}