Re: A parsing question
| От | Tom Lane |
|---|---|
| Тема | Re: A parsing question |
| Дата | |
| Msg-id | 1154086.1591224944@sss.pgh.pa.us обсуждение исходный текст |
| Ответ на | A parsing question (Michael Nolan <htfoot@gmail.com>) |
| Список | pgsql-general |
Michael Nolan <htfoot@gmail.com> writes:
> Recently I was typing in a query in PG 10.4.
> What I MEANT to type was: Where xyz >= 2400
> What I actually typed was: Where xyz >- 2400
> The latter was interpreted as 'where xyz > -2400', but I'm wondering if it
> shouldn't have thrown an error on an unrecognized operator '>-'
No, we do it that way intentionally, because the SQL standard requires us
to parse such cases in this way. The CREATE OPERATOR man page has some
details about what is and isn't a valid operator name ... and '>-' isn't
one. (The rules there probably look a bit bizarre, but the intention
is to ensure that operators found in the SQL spec can be parsed without
requiring spaces between them.)
regards, tom lane
В списке pgsql-general по дате отправления: