GFR: Make scale for large record groups/collections

[DominicBM]

The current approach currently doesn't scale for large record groups/collections. Currently, we first query the internal API for all the series in an RG/Collection, and then we make an array of all the series' NAIDs to pass as a query to the public API. For large record groups/collections, we can't pass hundreds of NAIDs in a URL parameter, so this errors.

We have to do it this way because (1) the public API fields for parentRecordGroup and parentCollection are not searchable (they are not included on the field white list), but (2) the internal API doesn't give us all the data we need, since it is designed for brief results. So we use the internal API to get the list of NAIDs, and then query them explicitly instead of querying by parent description.

In the future, with the white list issue resolved, we'd hope to just search "https://catalog.archives.gov/api/v1?rows=10000&description.series.parentRecordGroup.recordGroupNumber=" to get all the necessary metadata fields for all the series in a record group at once.

[DominicBM]

Right now, this is mitigated by limiting series results to the first 750, which omits some RGs (about 50).