On Mon, May 17, 2021 at 4:23 PM Amit Kapila <amit.kapil...@gmail.com> wrote: > > On Tue, May 11, 2021 at 11:28 AM Bharath Rupireddy > <bharath.rupireddyforpostg...@gmail.com> wrote: > > > > On Mon, May 10, 2021 at 7:39 PM Tom Lane <t...@sss.pgh.pa.us> wrote: > > > > > > Bharath Rupireddy <bharath.rupireddyforpostg...@gmail.com> writes: > > > > On Mon, May 10, 2021 at 12:00 PM Tom Lane <t...@sss.pgh.pa.us> wrote: > > > >> Yeah, this error message seems outright buggy. However, it's a minor > > > >> matter. Also, some people think "positive" is the same thing as > > > >> "non-negative", so maybe we need less ambiguous wording? > > > > > > > Since value 0 can't be considered as either a positive or negative > > > > integer, I think we can do as following(roughly): > > > > > > > if (value < 0) "requires a zero or positive integer value" > > > > if (value <= 0) "requires a positive integer value" > > > > > > I was thinking of avoiding the passive voice and writing > > > > > > "foo must be greater than zero" > > > > +1 for "foo must be greater than zero" if (foo <= 0) kind of errors. > > But, we also have some values for which zero is accepted, see below > > error messages. How about the error message "foo must be greater than > > or equal to zero"? > > > > +1 for your proposed message for the cases where we have a check if > (foo < 0). Tom, Michael, do you see any problem with the proposed > message? We would like to make a similar change at another place [1] > so wanted to be consistent. > > [1] - > https://www.postgresql.org/message-id/CALj2ACWGB9oHCR5ygkc8u6_QDqecObf9j2MxtOgsjZMMKsLj%3DQ%40mail.gmail.com
Thanks all for your inputs. PSA v2 patch that uses the new convention. With Regards, Bharath Rupireddy. EnterpriseDB: http://www.enterprisedb.com
v2-0001-Disambiguate-error-messages-that-use-non-negative.patch
Description: Binary data
