2010/10/17 Terry Laurenzo <tj@laurenzo.org>:
> Hi all -
> I independently started some work on a similar capability as was contributed
> back in August by Joey Adams for a json datatype. Before starting, I did a
> quick search but for some reason didn't turn this existing thread up.
> What I've been working on is out on github for
> now: http://github.com/tlaurenzo/pgjson
> When I started, I was actually aiming for something else, and got caught up
> going down this rabbit hole. I took a different design approach, making the
> internal form be an extended BSON stream and implementing event-driven
> parsing and serializing to the different formats. There was some discussion
> in the original thread around storing plain text vs a custom format. I have
> to admit I've been back and forth a couple of times on this and have come to
> like a BSON-like format for the data at rest.
Reading your proposal, I'm now +1 for BSON-like style. Especially JS
engine's capabilities to map external data to the language
representation are good news. I agree the mapping is engine's task,
not data format task. I'm not sure if your BSON-like format is more
efficient in terms of space and time than plain text, though.
I like as simple design as we can accept. ISTM format, I/O interface,
simple get/set, mapping tuple from/to object, and indexing are minimum
requirement. Something like JSONPath, aggregates, hstore conversion
and whatsoever sound too much.
Regards,
--
Hitoshi Harada