Benchmarking different algo to speed up search.

[dhruv-1001]

Description

Right now we have 4 administrative boundaries. State, District, Sub District, Village.
After project setup, master list for this can be found under server/geojson-data as PARSED_MASTER_LOCATION_NAMES.json which is derived from MASTER_LOCATION_NAMES.json.

We have an API that does fuzzy search on all levels, and returns the matches.
Right now the implementation is using Fuse, which is really slow. (implementation inside server/location.search.js
We need to benchmark the search with different algo(s) that uses trie data structure. Do some research on what provides us with best results.

References:
https://gist.github.com/smhanov/94230b422c2100ae4218
https://docs.rs/rsmorphy/latest/rsmorphy/dawg/index.html
https://johnresig.com/blog/javascript-trie-performance-analysis/
https://github.com/gotenxds/WordGraphs

[SakshumGusain]

Sir, I'm interested in taking up this ticket

[dhruv-1001]

Hey @SakshumGusain, sure! go ahead. Let me know, if you have any doubts.

[akshatcoder-hash]

we can use separate trie for each level (state, district, subDistrict, village) to store the respective names allowing independent searches at each level