Here is an example: http://mysqlserverteam.com/indexing-json-documents-via-virtu...

The optimizer will match JSON expressions to virtual columns w/indexes. So that means you can keep your queries in this form and have them indexed:

EXPLAIN SELECT * FROM features WHERE feature->"$.properties.STREET" = 'MARKET'\G

Is there a way to index JSON in MySQL without creating one index per unique field?

The mapping is not strictly one index per field, but a virtual column expression must map to a MySQL type. So depending on the use-case you may be able to use more complicated expressions to limit indexes.

But to expand your question, there are two FRs we are looking at:

1) Inverted (aka FULLTEXT) indexes on JSON documents 2) Multi-value indexes

i believe you have to create a virtual column on specific document fields and create a secondary index on that similar to what you do with Postgres' expression indices, I do not think MySQL has a general JSON GIN index yet