update docs to explain how to delete input levels

Author: stevenwinship

doc/sphinx-guides/source/api/changelog.rst

@@ -13,7 +13,7 @@ v6.8
 - For POST /api/files/{id}/metadata passing an empty string ("description":"")  or array ("categories":[]) will no longer be ignored. Empty fields will now clear out the values in the file's metadata. To ignore the fields simply do not include them in the JSON string.
 - For PUT /api/datasets/{id}/editMetadata the query parameter "sourceInternalVersionNumber" has been removed and replaced with "sourceLastUpdateTime" to verify that the data being edited hasn't been modified and isn't stale.
 - For GET /api/dataverses/$dataverse-alias/links the Json response has changed breaking the backward compatibility of the API.
-- For PUT /api/dataverses/$dataverse-alias/inputLevels custom input levels that had been previously set will no longer be deleted.
+- For PUT /api/dataverses/$dataverse-alias/inputLevels custom input levels that had been previously set will no longer be deleted. To delete input levels send an empty list (deletes all), then send the new/modified list.
 - For GET /api/externalTools and /api/externalTools/{id} the responses are now formatted as JSON (previously the toolParameters and allowedApiCalls were a JSON object and array (respectively) that were serialized as JSON strings) and any configured "requirements" are included.
 
 v6.7

doc/sphinx-guides/source/api/native-api.rst

@@ -1106,7 +1106,8 @@ Update Collection Input Levels
 
 Updates the dataset field type input levels in a collection.
 
-Please note that this endpoint does not change previously updated input levels of the collection page, so if you want to keep the modify exisitng ones, you will need to include them in the JSON request body.
+Please note that this endpoint does not change previously updated input levels of the collection page, so if you want to add new levels or modify existing ones, you will need to include them in the JSON request body.
+In order to delete input levels you must call this API with an empty list to delete all of the input levels, then call this API with the new list of input levels.
 
 If one of the input levels corresponds to a dataset field type belonging to a metadata block that does not exist in the collection, the metadata block will be added to the collection.