Indices should be saved in batches of 25 (or perhaps this is configurable?) where the most recent, incomplete batch is named latest instead of with an ID
When a batch completes its set:
an index ID is generated
a new latest batch is created and the ID of the last batch is added to its previous metadata key
Searches always start from the latest batch and then download subsequent batches via linked list
Indices should be created for post types, slugs, and tags
Should there also be an index for all posts?
Should the individual post data be available on the API endpoint?
latestinstead of with an IDlatestbatch is created and the ID of the last batch is added to itspreviousmetadata keyIndices should be created for post types, slugs, and tags