Re: [HACKERS] CREATE DOMAIN json vs built-in json

2012-03-20 Thread Tom Lane
Hitoshi Harada writes: > I've noticed our plv8 regression test now fails. It has CREATE DOMAIN > json AS text ... and validates text via v8's JSON.parse(), which was > working before introducing built-in json type. The test itself can be > solved simply by creating schema, but my surprise is tha

Re: [HACKERS] CREATE DOMAIN json vs built-in json

2012-03-20 Thread Abel Abraham Camarillo Ojeda
On Tue, Mar 20, 2012 at 2:47 AM, Abel Abraham Camarillo Ojeda wrote: > On Tue, Mar 20, 2012 at 2:44 AM, Hitoshi Harada wrote: >> I've noticed our plv8 regression test now fails.  It has CREATE DOMAIN >> json AS text ... and validates text via v8's JSON.parse(), which was >> working before introdu

Re: [HACKERS] CREATE DOMAIN json vs built-in json

2012-03-20 Thread Abel Abraham Camarillo Ojeda
On Tue, Mar 20, 2012 at 2:44 AM, Hitoshi Harada wrote: > I've noticed our plv8 regression test now fails.  It has CREATE DOMAIN > json AS text ... and validates text via v8's JSON.parse(), which was > working before introducing built-in json type.  The test itself can be > solved simply by creatin

[HACKERS] CREATE DOMAIN json vs built-in json

2012-03-20 Thread Hitoshi Harada
I've noticed our plv8 regression test now fails. It has CREATE DOMAIN json AS text ... and validates text via v8's JSON.parse(), which was working before introducing built-in json type. The test itself can be solved simply by creating schema, but my surprise is that we allow a domain whose name i