| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Peter Eisentraut <peter_e(at)gmx(dot)net> |
| Cc: | pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: parsenodes vs. primnodes |
| Date: | 2007-01-31 00:10:28 |
| Message-ID: | 13579.1170202228@sss.pgh.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Peter Eisentraut <peter_e(at)gmx(dot)net> writes:
> If something from primnodes.h (XmlExpr) needs something from
> parsenodes.h (TypeName), should I just move the former to the latter,
> or is there some major semantic dividing line between the two? Or
> maybe TypeName should really be a primnode?
Anything in primnodes should never need a TypeName. Post-parse-analysis
representation of type info is as a pg_type Oid. If you think you need
something else, let's have a discussion about why. (Perhaps you need to
split XmlExpr into two different node types, one for raw grammar output
and one for parse-analysis output?)
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tom Lane | 2007-01-31 00:32:39 | Re: Logging Lock Waits |
| Previous Message | Simon Riggs | 2007-01-30 23:48:20 | Logging Lock Waits |