If fields were added to or removed from an object type definition and no cleanup of the type was made, the file exporter exports removed fields. In some cases, the exporter breaks. This happens for example, if a new field was added to a type without a cleanup and specific objects of that type were edited. So some objects of that type will have data about the added field in the database and some not.
This happens in the following exporters:
CSV
JSON
XML
In the XLSX exporter, this behavior was fixed by
FYI:
Fields removed from the type. When exporting the objects, these fields are no longer displayed.
New fields are assigned to the type. When exporting the objects these fields are displayed.
In both cases the 'clean' - function does not change the initial situation. It does not matter if a 'clean' is performed.
New fields are assigned to the type and a default value is set for the field. The values are only taken over during 'clean'.
If the default value of an already existing field is changed afterwards. If the 'clean' - function cannot be executed anymore.
Should the default value in the type should always overwrite the empty value? Or should empty values be allowed and the default values are only used for the fallback?
Merged and cherry-picked to the version-1.4 branch.