Question please around hierarchy when exporting data from an adapter to dataset.
Instructions below are from the connector manual:
1. Importing Product Data Into A Data Set
2. Within the Connector application there is the ability to import product records into a data set, based on exporting data from a selected adaptor. When this occurs a data import process is called within the Connector that will attempt to import product records into a data set. Listed below are steps that this Product import process will follow:

1. Call adaptor's Products data export to retrieve product records from its configured data source.
2. For all existing product records within the data set the isUpdated field to N. This denotes that the product records haven't been updated.
3. Iterate through each product record retrieved from the adaptor, perform the following steps for each record:
1. Check if the product record's keyProductID field has value set, if not skip importing the record into the data set.
2. Check if a previous product record from the adaptor has already been read that contains the same keyProductID, if another imported record does then skip importing the record.
3. Check if an existing product record already exists within the data set, if a matching record is not found then add a product record to the data set.

This indicates that when data is exported from a Micronet adapter to the dataset for attachments in this example, if the keyProductId exists from a previous export and an Incremental export is run, the record is ignored/skipped.

In my case an incremental export was run, after one of the records had been changed (different product attachment name), it was then overwitten by the export back to original data (when it was supposed to be skipped as existing).

Is this correct process, or do I have a workflow issue?