Dataflow connection to Lakehous table created by EventHandler

[Ian_dw]

I am trying to connect to a Lakehouse table, created by an EventHandler process, where records are posted via the Azure Hub.

When I use a dataflow one of my 'textual' fields is missing from the list, although it is clearly visible & readable via the Lakehouse explorer.

When I look at another table created a different way, 'text' fields are highlighted with an 'abc' tag, to show it's a text field.

Looking at the field of the "Event" table, which is missing, it is tagged as '{}'. Am i right to assume it might refer to an object instead??, and if so, how can I get Fabric to use the field as a text one (it contains JSON data). or alternatively, what is needed to be sent to the Azure Hub, to force it to ba a Json text string??

[miguel]

Does the field show up at all in Dataflows? or its just not showing up at all?

[Ian_dw]

Hi Miguel, the field is completely missing in the Dataflow, but the fields either side of it are there. In the Lakehouse explorer the field 'IS' there but is marked with a {} next to it's field name. For other Textual fields, these are marked in the explorer as 'abc'.  If I use pyspark.sql the data seems to look like an object, which can be expanded by clicking the small triangle. The Event handler process itself, allows you to convert the 'field' to Text, but when data hits the Lakehouse table, it seems to error, when trying to convert the data.  I am wondering if the Source code which pushes the event, needs this field somehow being sent is converted into a 'Textual' format. However to do this I have to get to the code, which I cannot see, so obviously a Fabric solution would be better..