We are using the nightly build of Courier (v2.11.122).

 

We implemented a Custom PropertyDataResolverProvider.

Because we have a Custom datatype called "LinkPicker" to be able to create a hyperlink to another content node in Umbraco. To enable the translation of the linked node in the other environment where this gets extracted, we add the Guid from the linked node as extra field.

 

In the PackagingProperty method we save the LinkPicker as json via the following code:

 

public override void PackagingProperty(Item item, ContentProperty propertyData)        
        {
            //1. Get the LinkPicker
            var linkPicker = JsonHelper.DeserializeJsonString<UmbrellaLinkPicker>(propertyData.Value.ToString());
            if (linkPicker != null && linkPicker.ContentId > 0)
            {
                //1. Get the Guid from Umbraco With Type Content
                var umbracoGuid = PersistenceManager.GetDefault(this.ExecutionContext)
                    .GetUniqueId(linkPicker.ContentId, NodeObjectTypes.Document);
               //2. Get the new Value (json) for the linkPicker along with the Guid from Umbraco
               var courierLinkPicker = new CourierLinkPicker { UmbracoGuid = umbracoGuid, LinkPicker = linkPicker };
               // The new value is the LinkPicker with the Guid of the Linked Content
               // When extracting, we can find the new environment's NodeId from this Guid
               propertyData.Value = JsonHelper.SerializeJsonString(courierLinkPicker);
               
               //3. Add the dependency for the linked Content because Umbraco won't do this automatically
                item.Dependencies.Add(umbracoGuid.ToString(), _linkPickerProviderId);
          }
          base.PackagingProperty(item, propertyData);
}

The output of the PackagingProperty method would be something like:

 

<ContentProperty>
            <Alias>alternativeLink</Alias>
            <DataType>6b72ada3-9b53-45a1-93ff-42bb13bdadd7</DataType>
            <PropertyEditorAlias>Umbrella.LinkPicker</PropertyEditorAlias>
            <DBType>Nvarchar</DBType>
            <Value xsi:type="xsd:string">{&#xD;
  "umbracoGuid": "40d25257-746a-4520-b7f6-16e899c7fd00",&#xD;
  "linkPicker": {&#xD;
    "type": "content",&#xD;
    "contentId": 1171,&#xD;
    "contextId": 0,&#xD;
    "mediaId": 0,&#xD;
    "newWindow": false,&#xD;
    "url": null,&#xD;
    "title": "widget group",&#xD;
    "email": null,&#xD;
    "iconCssClass": ""&#xD;
  }&#xD;
}</Value>
        </ContentProperty>

 

On Extracting we use this code:

 

 

public override void ExtractingProperty(Item item, ContentProperty propertyData)
        {
            // Get the LinkPicker and its linked Guid (wrapper in CourierLinkPicker)
            var courierlinkPicker = JsonHelper.DeserializeJsonString<CourierLinkPicker>(propertyData.Value.ToString());
            if (courierlinkPicker != null && courierlinkPicker.LinkPicker != null && courierlinkPicker.UmbracoGuid != Guid.Empty)
            {              
                    // Extract LinkPicker
                    // Resolve the Umbraco Guid to this new environment's NodeId with type Document
                   var newContentNodeId = PersistenceManager.GetDefault(this.ExecutionContext).GetNodeId(courierlinkPicker.UmbracoGuid, NodeObjectTypes.Document);
            var umbrellaLinkPicker = courierlinkPicker.LinkPicker;            
            if (newContentNodeId > 0)
            {
                // Update the linked Content with the new ContentId
                umbrellaLinkPicker.ContentId = newContentNodeId;
            }
}

 

Now here's the catch:

 

When the linked node is of the DocumentType Page it works as excepted, the "local" nodeId gets retreived.

 

We use the concept of subnodes to enable the Back Office user to add content blocks to a Page. 

We call them Widgets.

 

When the linked node is a Widget the returned nodeId is 0 (zero).

 

What we know is that Page has a MVC View and Controller.

 

The widgets don't, because it's the Page has the responsibility to render the widgets.

! Also we test the extracting functionality by deleting the DB, thus simulating that the new environment has no content, documentypes, etc yet.

So maybe the issue is that the target node does not exist on the target environment at the moment that the source page is extracted??

The target node is added as dependency to the source node, so I would expect that this would prevent that from happening?

 

Many thanks in advance for any help regaring this issue!

 

Kind Regards,

Jan.