Re: bug in 7.4.2, with Handling of Double Quotation Marks

Поиск
Список
Период
Сортировка
От Stephan Szabo
Тема Re: bug in 7.4.2, with Handling of Double Quotation Marks
Дата
Msg-id 20040405165823.W19491@megazone.bigpanda.com
обсуждение исходный текст
Ответ на bug in 7.4.2, with Handling of Double Quotation Marks  (Alexander S <sasha@in.crimea.ua>)
Список pgsql-bugs
On Wed, 31 Mar 2004, Alexander S wrote:

> Bug in 7.4.2.. Concern with handling of Double  Quotation Marks (").
> Escape of  double quotation marks can't solve the problem.
>
> ----------
> CREATE TABLE public.test
> (
>   my_column varchar NOT NULL,
>   CONSTRAINT my_primary PRIMARY KEY (my_column)
> ) WITHOUT OIDS;
> -----------
> CREATE TABLE public.ref_test
> (
>   my_column varchar[]
> ) WITH OIDS;
> -----------
> CREATE FUNCTION public."t-f_ref_test"()
> RETURNS trigger AS
> 'BEGIN
> FOR i IN 1..30 LOOP
>     IF NEW."my_column"[i] not in (select "my_column" from public.test) THEN
>         RAISE EXCEPTION \'error: my_column[] --> test.my_column \';
>     END IF;
> END LOOP;
> RETURN NEW;
> END;'
>   LANGUAGE 'plpgsql' VOLATILE;
> ----------
> CREATE TRIGGER t_ref_test
>   BEFORE INSERT OR UPDATE
>   ON public.ref_test
>   FOR EACH ROW
>   EXECUTE PROCEDURE public."t-f_ref_test"();
> ----------
> insert into test values (' here \"some_text\" ');
> ----------
> insert into ref_test values (' {here \"some_text\"} ');
>
> But  here  postgres answers:  error: my_column[] --> test.my_column.
> So I think this exception should not appear here and this is bug.
> Instead I can only insert:
> --
> insert into ref_test values (' {here \\"some_text\\"} ');       -- but
> this is not exactly that i want to have.

What is the specific problem you're having, is it needing the double
backslashes on insert? is it the output format backslash escaping
quotation marks?


В списке pgsql-bugs по дате отправления:

Предыдущее
От: Peter Eisentraut
Дата:
Сообщение: Re: Bug #641: LIBPQ Linking with the static library on the WIN32 platform fails connections with "unknown host name".
Следующее
От: Paul Tillotson
Дата:
Сообщение: Re: Renaming a table leaves orphaned implicit sequences which