Thread: pgsql-server: Fix obviously-utterly-untested noTocComments code.

pgsql-server: Fix obviously-utterly-untested noTocComments code.

From
tgl@svr1.postgresql.org (Tom Lane)
Date:
Log Message:
-----------
Fix obviously-utterly-untested noTocComments code.

Modified Files:
--------------
    pgsql-server/src/bin/pg_dump:
        pg_backup_archiver.c (r1.95 -> r1.96)

(http://developer.postgresql.org/cvsweb.cgi/pgsql-server/src/bin/pg_dump/pg_backup_archiver.c.diff?r1=1.95&r2=1.96)

Re: pgsql-server: Fix obviously-utterly-untested

From
Philip Warner
Date:
At 05:44 AM 31/08/2004, Tom Lane wrote:
Fix obviously-utterly-untested noTocComments code.

It was extensively tested with a variety of DBs.

I *think* the problem was that the patch may have been against a too-old source (8.0b1), which still had _printTocHeader and _printTocData. eg. the code:

-       if
(AH->noTocComments)
-               return;
-

should have been in _printTocHeader, whereas it ended up in _printTocEntry.

Can this be easily avoided in the future? More context and/or just use latest CVS?


----------------------------------------------------------------
Philip Warner                    |     __---_____
Albatross Consulting Pty. Ltd.   |----/       -  \
(A.B.N. 75 008 659 498)          |          /(@)   ______---_
Tel: (+61) 0500 83 82 81         |                 _________  \
Fax: (+61) 03 5330 3172          |                 ___________ |
Http://www.rhyme.com.au          |                /           \|
                                 |    --________--
PGP key available upon request,  |  /
and from pgp.mit.edu:11371       |/

Re: pgsql-server: Fix obviously-utterly-untested

From
Bruce Momjian
Date:
Philip Warner wrote:
> At 05:44 AM 31/08/2004, Tom Lane wrote:
> >Fix obviously-utterly-untested noTocComments code.
>
> It was extensively tested with a variety of DBs.
>
> I *think* the problem was that the patch may have been against a too-old
> source (8.0b1), which still had _printTocHeader and _printTocData. eg. the
> code:
>
> -       if (AH->noTocComments)
> -               return;
> -
>
> should have been in _printTocHeader, whereas it ended up in _printTocEntry.
>
> Can this be easily avoided in the future? More context and/or just use
> latest CVS?

Ah, I remember having trouble applying the patch so I had to do it
manually.  I looked at the context around the patch, saw a matching
comment, and put it there.  It did seem quite high to me in the function
but I figured the author knew the ramifications.

--
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073