We are using a admin import based pipeline for standing up environments quickly. We have been running on 4.7
We recently wanted to update to Neo4j 5.x, and ran into a couple of issues
After changing to the updated import syntax, we still ran into a few issues as it turns out our raw data has some rows where their fields are blank for columns where the datatype is set to point or datetime.
I guess previous versions allowed ignoring these columns for those rows when --ignore-empty-strings was set to 'true'
However that seems not to work now on 5.1 where admin import complains for every field that is set to '' where the datatype set to datetime or point in the header
Is there any trivial way to tell admin-import to ignore these fields, or allow the datetime and point datatypes to be nullable?
I am able to "work around" the problem by removing the datatypes for these fields, and rather do post processing, but that will be a significant pipeline change. Alternatively by "fixing" all the source data (even more work)