Re: postgres_fdw - should we tighten up batch_size, fetch_size options against non-numeric values?

From: Fujii Masao <masao(dot)fujii(at)oss(dot)nttdata(dot)com>
To: Bharath Rupireddy <bharath(dot)rupireddyforpostgres(at)gmail(dot)com>
Cc: Kyotaro Horiguchi <horikyota(dot)ntt(at)gmail(dot)com>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: postgres_fdw - should we tighten up batch_size, fetch_size options against non-numeric values?
Date: 2021-06-30 12:23:52
Message-ID: 8ca543af-4164-08e2-4496-2c05690886db@oss.nttdata.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 2021/05/20 1:01, Bharath Rupireddy wrote:
> Thanks for the comments. I added separate messages, changed the error
> code from ERRCODE_SYNTAX_ERROR to ERRCODE_INVALID_PARAMETER_VALUE and
> also quoted the option name in the error message. PSA v3 patch.

Thanks for updating the patch!

+ (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
+ errmsg("invalid numeric value for option \"%s\"",
+ def->defname)));

In reloptions.c, when parse_real() fails to parse the input, the error message
"invalid value for floating point option..." is output.
For the sake of consistency, we should use the same error message here?

- (errcode(ERRCODE_SYNTAX_ERROR),
- errmsg("%s requires a non-negative integer value",
+ (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
+ errmsg("invalid integer value for option \"%s\"",

IMO the error message should be "invalid value for integer option..." here
because of the same reason I told above. Thought?

Regards,

--
Fujii Masao
Advanced Computing Technology Center
Research and Development Headquarters
NTT DATA CORPORATION

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message David Rowley 2021-06-30 12:24:19 Re: Use pg_nextpower2_* in a few more places
Previous Message Amit Kapila 2021-06-30 12:15:08 Re: Allow streaming the changes after speculative aborts.