How do I control the order in which feature types are written? There are several ways to control the order in which the features are written.
Order the readers:
Provided there are no "blocker" transformers, the features will be written in the order they are read. This only applies to the reader level, not the feature types. You don't have control over the order the individual feature types in a reader are read.
Order the writers:
The writer feature type order cannot be controlled; however, the writer order can be controlled. This means that if you have two feature types in your workspace and you want one feature type written before another, you can add a writer for each feature type. The feature type belonging to the first writer in the Navigator pane will be written first, and the feature type belonging to the second writer will be written next.
There are two MapInfo mid/mif writers in this workspace. Bus stops will be written first, and bus routes second.
Hold-up features in the pipeline using FeatureHolder:
You can use the FeatureHolder transformer to hold back features. These features will be written last. In the example of parent/child tables in a database, you might hold back the parent features with FeatureHolder, write the children, and then write the parent table. A Sorter can also be used to order output at an individual feature level.
How to Write Feature Types in a Specific Order
Was this article helpful?
As of January 14th, 2026, comments on knowledge base articles have been closed. To make sure questions don’t get missed and to enable more community support, we’ve moved discussions to the FME Community. If you have a question or a comment about this article, please create a new post or create a support ticket.