Export Excel/CSV: Multiple table relations

Before this release, export definitions with data format Excel or CSV were limited to use a maximum of 2 data items (tables).
This was due to the unstructured output making it difficult to support more than 1 data item relation.

There are of course still limitations due to the restrictions on the output format.
While multiple data relations between records will be allowed, it will still not be possible to have multiple arrays for different data items (1 to n relation) related to the same data item, since this can’t be properly translated to a flattened structure. Trying to set multiple arrays on data items in this way will result in the following error message.

However, it will be allowed to use the array option on a data item that also has a parent data item with the array option checked.
Below you will find some examples to illustrate the possibilities with this functionality.

Why is this functionality introduced as optional beta feature?
This is because there are scenario’s where this new handling of table relations for unstructured exports could break existing configurations.
This will allow users to check if existing configurations in export definitions still work when this feature is enabled, and will allow to make modifications where necessary without having to worry about breaking active definitions with the new release.

One example of breaking functionality is that with the current solution only records are exported when both records in the parent table and child table can be found if the definition uses 2 data items.
In database terms, the relation between the records is seen as an inner join.
This is in contrast with the structured exports (JSON/XML) where these relations are handled as left outer join. i.e. the parent record is not required to also have child records.