Are you sure the file is still there? Because you are using FeatureCaching, you can remove the extracted zip using the explorer without workbench knowing it.

Hi niels, yeah I see it popping up in my explorer when running the workspace. Maybe it has to do with my settings of the 2nd FeatureReader (reading XML). I clicked 'Feature Types to Read' and got an error.

However, reading the same XML with a seperate XML reader works fine:

This should really work. The error is because the Dataset is an attribute instead of a value, so it can't scan for a schema, so you can't choose from scanned featuretypes.

Did you set the FeatureReaders OutputPorts to Single Output Port?

Okay, hence it is not possible to read XML in the middle of a workflow ?

I set my FeatureReaders to SingleOutput port but that does not make a difference.

It depends 😃 If your schema is fixed, it is perfectly doable:

• In the FeatureReader XML reader, select a sample file in Dataset.
• In the Parameters, select the elements to match. (Assuming you use Feature Paths.)
• Set Output Ports to Single Output Port. *
• Test if this works as expected.
• Now change FeatureReader XML reader Dataset to path_windows.

* You can choose to set Output Ports to One per Feature Type, but then you need to select the sample file every single time you make a change to the FeatureReader, to populate the outputports. And that quickly becomes annoying in my book.

Another way of dynamically processing XML is to use a Text Reader (Read Whole File at Once = Yes) and process the attribute with the XML using XMLFragmenters.