pgsql: Improve error reporting for DROP FUNCTION/PROCEDURE/AGGREGATE/RO
Improve error reporting for DROP FUNCTION/PROCEDURE/AGGREGATE/ROUTINE.
These commands allow the argument type list to be omitted if there is
just one object that matches by name. However, if that syntax was
used with DROP IF EXISTS and there was more than one match, you got
a "function ... does not exist, skipping" notice message rather than a
truthful complaint about the ambiguity. This was basically due to
poor factorization and a rats-nest of logic, so refactor the relevant
lookup code to make it cleaner.
Note that this amounts to narrowing the scope of which sorts of
error conditions IF EXISTS will bypass. Per discussion, we only
intend it to skip no-such-object cases, not multiple-possible-matches
cases.
Per bug #15572 from Ash Marath. Although this definitely seems like
a bug, it's not clear that people would thank us for changing the
behavior in minor releases, so no back-patch.
David Rowley, reviewed by Julien Rouhaud and Pavel Stehule
Discussion: /messages/by-id/15572-ed1b9ed09503de8a@postgresql.org
Branch
------
master
Details
-------
https://git.postgresql.org/pg/commitdiff/bfb456c1b9656d5b717b84d833f62cf712b21726
Modified Files
--------------
src/backend/parser/parse_func.c | 396 ++++++++++++++++++---------
src/include/parser/parse_func.h | 4 +-
src/test/regress/expected/drop_if_exists.out | 29 ++
src/test/regress/sql/drop_if_exists.sql | 24 ++
4 files changed, 328 insertions(+), 125 deletions(-)
Hi,
On 2019-03-21 15:52:23 +0000, Tom Lane wrote:
Improve error reporting for DROP FUNCTION/PROCEDURE/AGGREGATE/ROUTINE.
These commands allow the argument type list to be omitted if there is
just one object that matches by name. However, if that syntax was
used with DROP IF EXISTS and there was more than one match, you got
a "function ... does not exist, skipping" notice message rather than a
truthful complaint about the ambiguity. This was basically due to
poor factorization and a rats-nest of logic, so refactor the relevant
lookup code to make it cleaner.Note that this amounts to narrowing the scope of which sorts of
error conditions IF EXISTS will bypass. Per discussion, we only
intend it to skip no-such-object cases, not multiple-possible-matches
cases.Per bug #15572 from Ash Marath. Although this definitely seems like
a bug, it's not clear that people would thank us for changing the
behavior in minor releases, so no back-patch.David Rowley, reviewed by Julien Rouhaud and Pavel Stehule
Discussion: /messages/by-id/15572-ed1b9ed09503de8a@postgresql.org
I now get:
/home/andres/src/postgresql/src/backend/parser/parse_func.c: In function ‘LookupFuncWithArgs’:
/home/andres/src/postgresql/src/backend/parser/parse_func.c:2285:5: warning: this statement may fall through [-Wimplicit-fallthrough=]
switch (objtype)
^~~~~~
/home/andres/src/postgresql/src/backend/parser/parse_func.c:2336:4: note: here
case FUNCLOOKUP_AMBIGUOUS:
^~~~
which seems like a somewhat righteous complaint? I'd just add a break to
silence it (which can't be reached, because all paths ought to error
out).
Greetings,
Andres Freund
On Sat, 23 Mar 2019 at 15:33, Andres Freund <andres@anarazel.de> wrote:
I now get:
/home/andres/src/postgresql/src/backend/parser/parse_func.c: In function ‘LookupFuncWithArgs’:
/home/andres/src/postgresql/src/backend/parser/parse_func.c:2285:5: warning: this statement may fall through [-Wimplicit-fallthrough=]
switch (objtype)
^~~~~~
/home/andres/src/postgresql/src/backend/parser/parse_func.c:2336:4: note: here
case FUNCLOOKUP_AMBIGUOUS:
^~~~which seems like a somewhat righteous complaint? I'd just add a break to
silence it (which can't be reached, because all paths ought to error
out).
hmm. Surprised your compiler can't see that's not possible. I guess we
just need a break on line 2335.
--
David Rowley http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
David Rowley <david.rowley@2ndquadrant.com> writes:
On Sat, 23 Mar 2019 at 15:33, Andres Freund <andres@anarazel.de> wrote:
I now get:
/home/andres/src/postgresql/src/backend/parser/parse_func.c:2285:5: warning: this statement may fall through [-Wimplicit-fallthrough=]
hmm. Surprised your compiler can't see that's not possible. I guess we
just need a break on line 2335.
Yeah, this no doubt is connected to what I noticed in commit 41c912cad:
Testing with gcc 8.0.1 (Fedora 28's current version), I found that
I also had to put explicit "break"s after elog(ERROR) or ereport(ERROR);
apparently, for this purpose gcc doesn't recognize that those don't
return. That seems like possibly a gcc bug, but it's fine because
in most places we did that anyway; so this amounts to a visit from the
style police.
Will add the break shortly.
regards, tom lane