Mixed Indexes not getting used for Partial/regex match queries

[pushkarnagpal]

@dwitry
Storage: Janusgraph with HBase as backend (0.4.x)
Gremlin Tinkerpop version : 4.3.4
cypher-for-gremlin version: 1.0.1

I want to utilize the mixed indexes for performance of the following scenarios :

1. For the partial match ( =~ ) queries:

match (p:party) where p.partyType =~ 'organization' return p limit 1

The query does work but I want to verify if this translates to a gremlin query that utilizes an index for large datasets. But the function cypherRegex() function errors out in translation.

Translating to textContains() would make use of the mixed indexes here

2. Queries using the CONTAINS clause

This clause translates to containing() function which does not utilize any indexes as mentioned in this documentation here:
https://docs.janusgraph.org/v0.4/index-backend/text-search/#tinkerpop-text-predicates
Same for STARTS WITH and ENDS WITH clauses.

Translating to textContains() instead would make use of mixed indexes to improve query performance here.

Reference Issue:
#294