Please help me in understanding the use of "Filter unknown properties" in developer studio>Pipeline>Index adapter. As per the documentation, If enabled, it will remove source properties from endeca records so that endeca record can contain exclusively endeca dimensions and properties. Does it mean endeca removes auto mapped source properties?
Thanks, Mohan
I don't recognize the term "auto mapped source properties" and don't think it's really a standard way to refer some aspect of Forge, so I'm just guessing here, but I assume you are referring to Forge's behavior where it leaves source properties in the Endeca record even if they haven't been explicitly mapped to an Endeca property. If that's what you mean, then yes, Endeca will remove those properties from the record before sending it to Dgidx so you are only left with properties you've explicitly defined in Dev Studio and created mappings for in your pipeline.
In particular, after Endeca maps source properties to Endeca properties using the property mapper, the source properties are still left there on the record as part of your final Endeca record that is sent to Dgidx. For example, if you have a CSV file like this:
ID,Name,Inventory,Color,PartNumber
1456,Hammer,34,Black,1234
...and you have defined the following Endeca properties in Developer Studio: product.name, record.spec, product.color, product.partnumber
...and you have these mappings in your property mapper:
ID -> record.spec
Name -> product.name
Color -> product.color
PartNumber -> product.partnumber
When you run your pipeline, if you leave "Filter unkown properties" as false, you'll end up with this record:
Record 1456
-----------
record.spec: 1456
product.name: Hammer
product.color: Black
product.partnumber: 1234
Inventory: 34
...but if you check that checkbox, you'll get the same record but without the Inventory property when it finally goes into the index.
