Re: ERROR: value too long for type character
От
Stephan Szabo
Тема
Re: ERROR: value too long for type character
Дата
Msg-id
20020521174657.R28822-100000@megazone23.bigpanda.com
Ответ на
Re: ERROR: value too long for type character (Doug Fields)
Список
Дерево обсуждения
PostgreSQL 7.1 forces sequence scan when there is no reason Denis Perchine <dyp@perchine.com>
Re: PostgreSQL 7.1 forces sequence scan when there is no reason Tom Lane <tgl@sss.pgh.pa.us>
Re: PostgreSQL 7.1 forces sequence scan when there is no reason Denis Perchine <dyp@perchine.com>
Re: PostgreSQL 7.1 forces sequence scan when there is no reason Tom Lane <tgl@sss.pgh.pa.us>
Re: PostgreSQL 7.1 forces sequence scan when there is no reason Denis Perchine <dyp@perchine.com>
ERROR: value too long for type character varying(255) Doug Fields <dfields-pg-general@pexicom.com>
Re: ERROR: value too long for type character varying(255) Neil Conway <nconway@klamath.dyndns.org>
Re: ERROR: value too long for type character varying(255) Jan Wieck <janwieck@yahoo.com>
Re: ERROR: value too long for type character Doug Fields <dfields-pg-general@pexicom.com>
Re: ERROR: value too long for type character Stephan Szabo <sszabo@megazone23.bigpanda.com>
On Tue, 21 May 2002, Doug Fields wrote: > > > > Do you know how I can tell it to "silently truncate" the data as I believe > > > it used to? > > > >Is there a reason you can't just use TEXT? It will be stored the same > >internally... > > The only reason being that I don't want to waste space storing arbitrarily > long pieces of information. > > I was not aware that TEXT is just as (in)efficient as VARCHAR, though. If > that is truly the case, I will consider using TEXT in some places, but only > when I am very careful in the INSERT and UPDATE clauses so as not to store > too long things. > > >But the most obvious way I can think of to do what you're asking is > >to define a rule to replace the value inserted with a call to > >substring(). > > So, basically: > > 1) There's absolutely no way to get back the 7.1.3 "silent truncation" > behavior (which is, IMO, stupid, I like the non-standard truncation behavior) You could do this via a text column and a before trigger which might work better than a rule, but I haven't tried either, so...
В списке pgsql-general по дате отправления