Is it correct that I only need ‘taxonomy’ as entry in the first line of the csv file?
The export contains the full hierarchy of taxonomy entries. What should be in the import file? Can I add only the last entry, in my case the name of the street or did I needthe full hierarchy?
If there is an existing item with taxonomy entry and I plan to add some more taxonomy entries to this item. Can I add only the new entries in the upload file? Are they added to the existing ones or replace they the existing ones.
These are great questions, let’s answer them. BTW @Maria_Cecilia we should open an issue in the Wiki with these topics to remind us to clarify them in the documentation.
The first line of the CSV (the header) is mainly to help you know which column you are referring to in the Importer Mapping step. Those strings that come after the first “|” are settings of the current metadata that will only be used by the Importer if in the Mapping step you choose the “Create metadatum” option. This way they will know how to configure a new metadata (it is a “taxonomy”-type one, “multiple”, with “display_no” selected, etc, etc). So if your metadata already exists and you select it in the Mapping step, nothing new happens.
Sadly, when dealing with hierarchical taxonomies, you must include the full hierarchy path. Otherwise we wouldn’t be able to separate child terms with the same name (which is possible).
If you are performing an item update via CSV (considering you have a column with `special_item_id`), then you must add existing terms (the ones exported via CSV Exporter). We have a logic for only appending attachments, but not metadata values (yet).
Also don’t forget to play a bit with Bulk Editing tools. You can use it append taxonomy terms instead of using the CSV Importer.
I see the problem with identical terms in taxonomy hierarchies. In my case I can avoid them. The levels in taxonomies here are 4-5 steps deep und some terrms are 12-20 signs long. This will maximize the error rate while editing the csv file in Calc/Excel.
How are other institutions handling this problem?
In my case the editors are 75-80 years old and some of them will not be able to edit the items in Tainacan. My very first idea is to write a macro for Calc/Excel that replaces the taxonomy terms by the path. This is not ideal because the list of taxonomy terms is dynamic and we start with 30-40 and I expect that we will get more than 100 later.
Uhm… to be honest we usually leave more user friendly input to the UI and tend to consider CSV editing a more advanced flow, so I don’t know of many situations like these.
I do know that some people usually avoid nesting too much. When Taxonomies get too hierarchical, one way to go is to really separate them in more than one metadata, if that can make sense, obviously.
As for your particular situation I don’t think writing a script (with the help of AI) to parse the CSV and fill that hierarchy would be too hard… You could offer as an input the exported version of the hierarchy that the CSV Vocabulary Exporter offers. I know it is not ideal if you are aiming to keep that Taxonomy growing over time, but it is what I can figure by now
