Implement OpenSearch/Elasticsearch document source/destination/corpus

[johann-petrak]

(See https://github.com/opensearch-project/OpenSearch)

Make sure the Python elasticsearch packages are fully compatible with OpenSearch.

Destination:

• handles how Document instances are getting stored and indexed.
  • always store the actual document json (possibly with filtered annotation sets)
  • in addition add fields for searching based on config options
  • this may include annotation in context text fields (after filtering/checking for overlaps)
  • store config information within the same index?

Source:

• iterator of documents for some search

Corpus:

• store as for destination
• retrieve by document index
• also use as source