Re: json indexing and data types

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Merlin Moncure <mmoncure(at)gmail(dot)com>
Cc: Kaare Rasmussen <kaare(at)jasonic(dot)dk>, PostgreSQL General <pgsql-general(at)postgresql(dot)org>
Subject: Re: json indexing and data types
Date: 2015-12-03 04:04:52
Message-ID: 9371.1449115492@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Merlin Moncure <mmoncure(at)gmail(dot)com> writes:
> On Wed, Dec 2, 2015 at 12:03 AM, Kaare Rasmussen <kaare(at)jasonic(dot)dk> wrote:
>> As json essentially only has three basic data types, string, int, and
>> boolean, I wonder how much of this - to index, search, and sort on
>> unstructured data - is possible.

> I feel your pain. jsquery is superb for subdocument searching on
> *specific* subdocuments but range searching is really limited.

Yeah. The problem here is that a significant part of the argument for
the JSON/JSONB datatypes was that they adhere to standards (RFC 7159 in
particular). I can't see us accepting a patch that changes them into
JSON-plus-some-PG-enhancements.

For cases where you know that specific sub-fields can be expected to be
of particular datatypes, I think you could get a lot of mileage out of
functional indexes ... but you'd have to write your queries to match the
indexes, which could be painful.

(Having said that, it sure looks to me like JSON's idea of a number is
float/numeric, not merely int. Are you sure you need more capability
in that department, and if so what exactly?)

regards, tom lane

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message mdaswani 2015-12-03 04:31:22 Support for hardware tokens for server/replication private key
Previous Message Dennis 2015-12-03 03:54:23 Re: loading data into cluster - can I daisy-chain streaming replication?