Add PubTator implementation

This will implement ONLY TextAnnotator

https://www.ncbi.nlm.nih.gov/research/pubtator/api.html

Be sure to implement the throttle

Note that certain ID transformations will need to be transformed

Note: we may also want to extend the TextAnnotator API to accept PMIDs as well as texts; pubtator could then return cached results.

Start by locating the existing TextAnnotator API and reading the linked PubTator API documentation. Define the implementation boundary around TextAnnotator, then identify the required throttle behavior and ID transformations. Done means PubTator-backed text annotation is implemented with those requirements addressed; the PMID caching idea is explicitly outside this scope.

Written by the indexing model from the issue text.