As I get close to 100 posts on this blog, one thing I’ve noticed is that I don’t cover as many CAD topics as you might expect. I’m not sure why; FME works well with CAD formats, we have plenty of CAD users, and I do have a background of CAD use. So this post is for FME users marked with the curse of those who draft, design and edit…..
When I describe a dynamic FME workspace I often say that “it lets you write data, even when you don’t know what the incoming layers or attributes will be”.
This perhaps causes more confusion than understanding, so, in order to clarify that statement, this post documents a recent workspace I created for a data translation project.
This use case is intended to show a real world scenario for FME dynamic workspaces, and how to integrate dynamic functionality with batching tools like a Dataset Fanout.
Mapping FME Feature Types (layers) is greatly complicated when there is no direct correlation between the source and destination schemas. Hundreds of tests and connections can result in a workspace worthy of the “FME Chamber of Horrors”. The SchemaMapper transformer can help by testing the features against mappings defined in a lookup table. Or, as Kenny Rogers might say, “I just dropped a SchemaMapper in, to see what condition my Attr Value Field was in”.
The SchemaMapper transformer is - rightly or wrongly - infamous for its complexity. But in a previous SchemaMapper post I showed how simple it was for mapping attributes from one name to another, and now I’ll demonstrate basic techniques for mapping Feature Types.