Does FlexSearch support dynamic document descriptors for multiple document schemas? (a lot different schemas that may grow up in future)

[ngdangtu-vn]

I usually get a lot excel files with different table structures and required to convert it into searchable list and embedded it into a website. I give this context to focuse on the unpredictable json schemas that my website could have. Therefore I look for options to set multiple document descriptors and runtime (not in inital stage new Document) so I can keep adding new json entries to FlexSearch.

So far I try to work around with this solution (for SPA):

Divide all the lists into pages. Each page only handle a one or two list so it can reduce amount of work on FlexSearch

1. Save all data into an IndexedDb
2. Load specific list base on the id provided from the current page
3. Initialize new Document w/ proper document descriptors
4. Ready the search
5. On page leaving, destroy the current doc instance and go back to step 2

What do you think about my solution? Is it a good idea to keep destroy doc instance? Or do we have a proper way to reuse the doc instance?