Tom Lane writes:
> Yeah, I was just looking at the IfSupported variant. In the structure
> I just suggested (separate ProcessSlowUtility function), we could make
> that work by having switch cases for some statements in both functions,
I've done it the way you propose here, and then in the Slow
sepgql: Use getObjectIdentity rather than getObjectDescription.
KaiGai Kohei, based on a suggestion from Álvaro Herrera
Branch
--
master
Details
---
http://git.postgresql.org/pg/commitdiff/b620fdabba3fd3c00587fb586f74eb7dc6eca223
Modified Files
--
contrib/sepgsql/database.c
Minor wording corrections for object-access hook stuff.
KaiGai Kohei
Branch
--
master
Details
---
http://git.postgresql.org/pg/commitdiff/d017bf41a32d08885f00a274603ed2e50816fe7f
Modified Files
--
src/backend/catalog/objectaccess.c |2 +-
src/include/catalog/objectaccess.
sepgsql: Enforce db_procedure:{execute} permission.
To do this, we add an additional object access hook type,
OAT_FUNCTION_EXECUTE.
KaiGai Kohei
Branch
--
master
Details
---
http://git.postgresql.org/pg/commitdiff/f8a54e936bdf4c31b395a2ab7d7bc98eefa6dbad
Modified Files
--
c
pgindent: fix downloading of BSD indent binary
Also fix accessing pgentab binary and tar.
Gurjeet Singh
Branch
--
master
Details
---
http://git.postgresql.org/pg/commitdiff/8daa4e960e15eb31e141585968c91cc666889585
Modified Files
--
src/tools/pgindent/pgindent | 19 ++
pgindent: improve error messages
per suggestion from Gurjeet Singh
Branch
--
master
Details
---
http://git.postgresql.org/pg/commitdiff/5003f94f661d3eff0c44774576fe2f1ae202da86
Modified Files
--
src/tools/pgindent/pgindent | 33 +
1 files cha
Clean up the mess around EXPLAIN and materialized views.
Revert the matview-related changes in explain.c's API, as per recent
complaint from Robert Haas. The reason for these appears to have been
principally some ill-considered choices around having intorel_startup do
what ought to be parse-time
Fix sporadic rebuilds for .pc files
The build of .pc (pkg-config) files depends on all makefiles in use, and
in dependency tracking mode, the previous coding ended up including
/dev/null as a makefile. Apparently, on some platforms the modification
time of /dev/null changes sporadically, and so t
