ИСТИНА

We present a binary storage for nested data structures with arrays support and its application to nested hstore. We consider this storage as a step forward true json data type. Recently, PostgreSQL got json data type, which basically is a string storage with validity checking for stored values and some related functions. To be a real data type, it has to have a binary representation, development of which could be a big project if started from scratch. Hstore is a popular data type, we developed years ago to facilitate working with semi-structured data in PostgreSQL. It is mature and widely used data type with indexing support. Our idea is to implement a binary storage for nested data structures with array support and use it to develop nested hstore data type as well as for json. The prototype was presented at PGCon-2013 and it proves the benefit of using such storage for hstore both in performance and functionality. Now we present an improved storage with types support, which makes possible to use it as storage layer for json. Also, we show, that hstore can get a great performance gain (an order of magnitude) from new GIN fast scan algorithm.