Re: [HACKERS] Another bug in pg_operator.h

Bruce Momjian <maillist@candle.pha.pa.us> writes:
>> Tom Lane <tgl@sss.pgh.pa.us> writes:
>> I will also commit a new regression test script
>> that looks for all the test conditions that I used to locate these
>> problems, in hopes that no new bugs of this ilk will creep in.

> What I did was to make a file in the include/catalog directory called
> template1_check.sql and pg_attribute_check.sql.  These are SQL
> statements that check various catalogs and report problems where things
> are missing joins.  Perhaps we could put something in there, or move
> these to the regression directory and include them in your stuff.

I saw those but it wasn't clear to me when they would get applied or
whether they were hand-generated or derived from something else.  So
I went and made a new regression test, because I think I comprehend
those.  If you want to fold the opr_sanity regress test into one of
the sql files in include/catalog, go right ahead.  (Or, maybe those
files should be pushed over to regression testing?  I dunno.)

> I see "@" means "on" sometimes, and "contained" sometimes, and they use
> "@" for both uses for the point/path combination.  Looks like a problem
> that "@" applies to point/path, and "on" and "contained" are both valid.
> However, they seem to mean the same thing.  Are on_ppath and
> pt_contained_path doing the same thing.  Thomas could help here.

I would expect that on_ppath checks to see if the point is on (touches)
the path, whereas pt_contained_path checks to see if the point is within
the area enclosed by the path.  But I haven't looked to see if that's
what the author of the code thought...
        regards, tom lane