From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Ashwin Agrawal <aagrawal(at)pivotal(dot)io> |
Cc: | Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Declared but no defined functions |
Date: | 2019-07-08 19:38:13 |
Message-ID: | 11451.1562614693@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Ashwin Agrawal <aagrawal(at)pivotal(dot)io> writes:
> Do we wish to make this a tool and have it in src/tools, either as part of
> find_static tool after renaming that one to more generic name or
> independent script.
Well, the scripts described so far are little more than jury-rigged
hacks, with lots of room for false positives *and* false negatives.
I wouldn't want to institutionalize any of them as the right way to
check for such problems. If somebody made the effort to create a
tool that was actually trustworthy, perhaps that'd be a different
story.
(Personally I was wondering whether pgindent could be hacked up to
emit things it thought were declarations of function names. I'm
not sure that I'd trust that 100% either, but at least it would have
a better shot than the grep hacks we've discussed so far. Note in
particular that pgindent would see things inside #ifdef blocks,
whether or not your local build ever sees those declarations.)
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Daniel Gustafsson | 2019-07-08 19:42:23 | Assertion for logically decoding multi inserts into the catalog |
Previous Message | Konstantin Knizhnik | 2019-07-08 19:32:38 | Re: [HACKERS] Cached plans and statement generalization |