Hi,

we have done quite a few migrations from EPiServer CMS 4, 5 and 6 to CMS 7 and the approach we have taken is to build a custom migration tool that uses the built in events ContentImporting and ContentImported. We use the normal import package and our tool helps us with the heavy lifting. I don't know if this helps you but the following commented code example will explain the main building blocks.

// The migration library will scan the assemblies of the bin folder
// looking for classes marked with the MigrationInitializer attribute
// and a public static void Initialize method which it will happily call.
[MigrationInitializer]
public static class MigrationDefinitions {
  public static void Initialize() {
    // Register your mappings using the static Register method of the 
    // MapperRegistry class. This method accepts a variable parameter list
    // if IPageMapping instances
    MapperRegistry.Register(

      // Hopefully you will not need to implement the IPageMapper interface
      // yourself, but use the PageMapper fluent API instead. The starting
      // point of the API is the Define method which accepts a name for the
      // mapping
      PageMapper.Define("Mapping name")

        // The Map method defines a mapping for a page type. The new strongly 
        // typed content type is specified as the type argument and the name 
        // of the old page type is given as the first argument. 
        .Map<NewsListPage>("[Old] NewsArchive",

          // The second argument to Map is an Action<SourcePage, TNew> callback
          // in which you specify how properties are mapped. The SourcePage
          // class contains the properties of the page being migrated and has
          // a couple of helpful methods you can use to get the values. 
          // This example uses the GetValueWithFallback method which falls 
          // back to the next property in the list when a previous value does
          // not exist or is null
          (s, d) => d.Heading = s.GetValueWithFallback<string>("Header", "PageName"))

        .Map<NewsPage>("[Old] NewsItem",
          (s, d) => {
            d.Heading = s.GetValue<string>("Header");
            // Sometimes you'll need to clean up content from the old pages
            // This is best done via extension methods. A few are defined in
            // the library which may or may not be of use to you
            d.Introduction = s.GetValue<string>("Ingress").CleanupForIntroduction();
            d.MainBody = s.GetValue<XhtmlString>("Bread").CleanupForMainBody();
        })

        // You may leave the Action argument if no properties are mapped
        .Map<ContainerPage>("[Old] Folder")

        // The Default method defines the mapping for all pages where a mapping
        // cannot be found.
        .Default<ArticlePage>(
          (s, d) => {
            d.Heading = s.GetValueWithFallback<string>("MainHeader", "Header", "PageName");
            d.Introduction = s.GetValue<XhtmlString>("Ingress").CleanupForIntroduction();
            d.MainBody = s.GetValue<XhtmlString>("MainBody").CleanupForMainBody();
        }));
  }
}

/Marcus

Ah, a customized mapper. That's nice. I might use this if the standard convert function is not enough. Thanks :)