From: | Kyotaro Horiguchi <horikyota(dot)ntt(at)gmail(dot)com> |
---|---|
To: | pgsql-hackers(at)lists(dot)postgresql(dot)org |
Subject: | Some doubious error messages and comments |
Date: | 2021-04-28 08:36:33 |
Message-ID: | 20210428.173633.1525059946206239295.horikyota.ntt@gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Hello.
0001: I found some typos in a error message and a comment.
multirangetypes.c: 1420
> errmsg("range_intersect_agg must be called with a multirange")));
This "range_intersect_agg" looks like a typo of "multirange_..".
operatorcmds.c:303
> * Look up a join estimator function ny name, and verify that it has the
"ny" looks like a typo of "by".
0002: The following messages are substantially same and are uselessly
split into separate messages. I'm not sure any compiler complains
about using %zu for int, explicit casting would work in that case.
be-secure-gssapi.c:351
> (errmsg("oversize GSSAPI packet sent by the client (%zu > %zu)",
> (size_t) input.length,
> PQ_GSS_RECV_BUFFER_SIZE - sizeof(uint32))));
be-secure-gssapi.c:570
> (errmsg("oversize GSSAPI packet sent by the client (%zu > %d)",
> (size_t) input.length,
> PQ_GSS_RECV_BUFFER_SIZE)));
0003: The messages below seems to be a bit unclear. I'm not sure they
worth doing.
conversioncmds.c: 130
errmsg("encoding conversion function %s returned incorrect result for empty input",
This is not wrong at all, but another message just above is saying
that "encoding conversion function %s must return type %s". Why
aren't we explicit here, like this?
"encoding conversion function %s must return zero for empty input"
typecmds.c:4294
> if (requireSuper)
> if (!superuser())
> ereport(ERROR,
> errmsg("must be superuser to alter a type")));
Where, requireSuper varies depending on the set of operations but the
description looks like describing general behavior. I'm not sure but
something like the following might be better?
+ errmsg("must be superuser to perform all operations")));
+ errmsg("some of the operations require superuser privilege")));
Any opinions or suggestions?
regards.
--
Kyotaro Horiguchi
NTT Open Source Software Center
Attachment | Content-Type | Size |
---|---|---|
0001-Fix-typos.patch | text/x-patch | 1.5 KB |
0002-Consolidate-substantially-same-messages.patch | text/x-patch | 946 bytes |
0003-Clarify-merror-messages.patch | text/x-patch | 1.4 KB |
From | Date | Subject | |
---|---|---|---|
Next Message | Aleksander Alekseev | 2021-04-28 08:40:27 | Re: Better sanity checking for message length words |
Previous Message | David Rowley | 2021-04-28 08:24:17 | Re: Result Cache node shows per-worker info even for workers not launched |