Exclude nested fields with large data under Collection/Upload datasets

[yuanzhou]

As a result of hubmapconsortium/entity-api#767, we agreed upon a short-term solution to exclude the files, metadata, and ingest_metadata fields nested under Collection.datasets and Upload.datasets in entity-api. This will resolve the page loading issue in ingest-ui.

On the search-api side, the large files list would still cause failures when indexing Collections and Uploads with such datasets. Confirmed by IU and Harvard teams, they are not using files, metadata, and ingest_metadata fields nested under Collection.datasets and Upload.datasets from the search documents. So we'll be excluding them to reduce the document size.

https://portal.hubmapconsortium.org/browse/collection/14f93f4c97996eefd5779f6a29d3f13c can be used for verifying the results.