I found it was easier to concatenate the UNION at the macro call site,
rather than the macro body. This eliminates the extra macro. Patch
attached and applied.
---
pgman wrote:
Modified patch attached and applied.
Bruce Momjian pgman@candle.pha.pa.us writes:
I found it was easier to concatenate the UNION at the macro call site,
rather than the macro body. This eliminates the extra macro. Patch
attached and applied.
Yeah, that looks like a much safer idea --- avoids making assumptions
about the
Modified patch attached and applied.
Rather than create the lists in psql, I used UNION SELECT 'KEYWORD' to
pass the keyword to the backend to be added to the query list.
This was already being done for schema names, and was easy and efficient
to add. My addition is even simpler because it
I will adjust this patch based on later feedback emails.
Your patch has been added to the PostgreSQL unapplied patches list at:
http://momjian.postgresql.org/cgi-bin/pgpatches
It will be applied as soon as one of the PostgreSQL committers reviews
and approves it.
On Fri, Feb 03, 2006 at 01:59:19PM -0500, Bruce Momjian wrote:
I will adjust this patch based on later feedback emails.
I have another version of the patch. Minor style errors are corrected in it.
As far as I can see, open items are:
- Should all the
const char* foocompletion = { BAR,
On Fri, Feb 03, 2006 at 04:15:32PM -0500, Tom Lane wrote:
- Should all the
const char* foocompletion = { BAR, BAZ };
be changed to
char* foocompletion = { BAR, BAZ };
Removing const decoration sure seems like the wrong direction to be
going in. Why can't you still have const
On Fri, Feb 03, 2006 at 04:15:32PM -0500, Tom Lane wrote:
Joachim Wieland [EMAIL PROTECTED] writes:
- Should all the
const char* foocompletion = { BAR, BAZ };
be changed to
char* foocompletion = { BAR, BAZ };
Removing const decoration sure seems like the wrong direction to be
Neil,
thanks for your review. I accepted what you wrote for all items I don't
mention in this reply.
On Sat, Jan 07, 2006 at 10:59:27PM -0500, Neil Conway wrote:
+ #define COMPLETE_WITH_MALLOCED_LIST(list) \
+ do { COMPLETE_WITH_LIST((const char**) list); free(list); list
= (char**)
Hi,
psql's tab completion has the following problem:
If we have the following syntax for example:
SET SESSION AUTHORIZATION user;
SET SESSION AUTHORIZATION DEFAULT;
After SET SESSION AUTHORIZATION, the tab completion can offer a list of
roles or the string constant DEFAULT. However it can't
