Re: quoting psql varible as identifier - Mailing list pgsql-hackers

From Robert Haas
Subject Re: quoting psql varible as identifier
Date
Msg-id 603c8f071001221120ra2b6c7cne5105e8882960f18@mail.gmail.com
Whole thread Raw
In response to Re: quoting psql varible as identifier  (Pavel Stehule <pavel.stehule@gmail.com>)
Responses Re: quoting psql varible as identifier
Re: quoting psql varible as identifier
List pgsql-hackers
On Fri, Jan 22, 2010 at 7:19 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:
> here is new variant. Add scan_state flag "valid" and enhance
> protection against execution broken statement.

This doesn't make sense to me.  You've changed the way \set is handled
- in a way that doesn't seem particularly appropriate to me - but most
of the other backslash commands are unmodified - but then there's this
hack at the bottom that overrides the return value if
psql_scan_is_valid() fails.  And then there's this:

-                               /* we need not do psql_scan_reset() here */
+                               psql_scan_reset(scan_state);

It's extremely unclear what the rationale for this change is.

Basically, you need to either improve the comments in here so that
someone can understand what is going on, or you need to submit it with
some detailed documentation explaining the rationale behind each
change and why it is right, or more than likely both.  But I think the
whole approach is likely off-base and you need to go back and think
about a cleaner way to get this done.

...Robert


pgsql-hackers by date:

Previous
From: David E. Wheeler
Date:
Subject: Review: listagg aggregate
Next
From: Tom Lane
Date:
Subject: Re: Git out of sync vs. CVS