[BUGS] BUG #4402: Column expansion: date comes wthout a cast
The following bug has been logged online:
Bug reference: 4402
Logged by: Joachim Unger
Email address: [EMAIL PROTECTED]
PostgreSQL version: 8.3
Operating system: Windwos XP Professional 32 Bit
Description:Column expansion: date comes wthout a cast
Details:
The column expansion lists date types without a date cast.
Here is an example where I'm using NEW inside a trigger:
[...]
ins_sql := 'INSERT INTO ' || quote_ident(new_child_table) || '('
||col_names(new_child_table) || ') VALUES ' || NEW;
RAISE NOTICE 'INSERT: %',ins_sql;
EXECUTE ins_sql;
[...]
Execution of the generated SQL throws an error:
ERROR: column "adate" is of type date but expression is of type integer
LINE 1: INSERT INTO table_2008_1(id,adate) VALUES (1,2008-01-30)
HINT: You will need to rewrite or cast the expression.
That hint describes exactly what is missing:(1,DATE '2008-01-30')
--
Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs
Re: [BUGS] BUG #4398: Backslashes get escaped despite of backslash_quote=off
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hello Tom, On 4 Sep 2008 at 12:53, Tom Lane wrote: > > "Rainer Glueck" <[EMAIL PROTECTED]> writes: > > How can backslashes be stopped from escaping themselves in like-statements > > Use ESCAPE ''. There isn't any other way provided. > ... > if you have no escape character then you have no way to quote % and _ Yes, that's true indeed. And since I learned that the escape clause is recommended to cope with that matter not only by PostgreSQL, but also by Transact-SQL, I'm going to adapt my database connector this way, withdraw this topic and thank you for pushing me a little bit more to enlightenment. :) Regards, Rainer -BEGIN PGP SIGNATURE- Version: idw's PGP-Frontend 5.1.0.0 / 4-2006 + PGP 6.5.8 Comment: KEY: 0x231D96D8 - FP: 9E01C5D7 7041D0C4 804F4705 76216606 231D96D8 iQA/AwUBSMDeV3YhZgYjHZbYEQKv3gCgx6UWMbVLsym6hb8/DG91JMNX3OAAoK9M NrpA0F3AGjPeqacOks2AVtJe =8SXN -END PGP SIGNATURE- -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
Re: [BUGS] BUG #4374: pg_restore does not restore public schema comment
Tom Lane wrote: > Craig Ringer <[EMAIL PROTECTED]> writes: > > The issue actually appears to be that the comment on default schema like > > `public' isn't dumped in the first place. > > Hmm ... there is a specific exclusion in _printTocEntry to suppress > dumping of either the CREATE SCHEMA command for "public" or its comment. > I don't quite recall the reasoning [ digs... ] Oh, here it is: > http://archives.postgresql.org/pgsql-committers/2008-01/msg00212.php > You can probably find some related discussion in the archives right > before that, but the commit message is clear enough. > > On the whole, since "public" is a quasi-built-in object, it doesn't seem > like a very good idea to depend on being able to alter its properties. > Why not attach your version comment to some other, entirely user-defined > object? "CREATE SCHEMA version" perhaps ... I have added a comment in the source code about this. -- Bruce Momjian <[EMAIL PROTECTED]>http://momjian.us EnterpriseDB http://enterprisedb.com + If your life is a hard drive, Christ can be your backup. + -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
Re: [BUGS] BUG #4377: casting result of timeofday() to timestamp fails in some timezones
Todd A. Cook wrote: > Tom Lane wrote: > > > > In the meantime, why aren't you just using clock_timestamp()? > > timeofday() is deprecated. > > I am using clock_timestamp() now. The use of timeofday() was from the era > when 8.0 was hot stuff. > > BTW, the word "deprecated" does not appear on the docs page where timeofday() > is listed (http://www.postgresql.org/docs/8.3/static/functions-datetime.html), > and there doesn't seem to be anything in the context of the 3 occurrences of > "timeofday" that would imply it is deprecated. Uh, we will probably never remove timeofday() but the new functions are clearer. -- Bruce Momjian <[EMAIL PROTECTED]>http://momjian.us EnterpriseDB http://enterprisedb.com + If your life is a hard drive, Christ can be your backup. + -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
Re: [BUGS] BUG #4402: Column expansion: date comes wthout a cast
Joachim Unger escreveu:
> The column expansion lists date types without a date cast.
>
And all of the text types too.
> Here is an example where I'm using NEW inside a trigger:
> [...]
> ins_sql := 'INSERT INTO ' || quote_ident(new_child_table) || '('
> ||col_names(new_child_table) || ') VALUES ' || NEW;
> RAISE NOTICE 'INSERT: %',ins_sql;
> EXECUTE ins_sql;
> [...]
>
Why aren't you expanding it yourself? Currently, PostgreSQL doesn't have
a function that does the quoting-if-needed automagically.
--
Euler Taveira de Oliveira
http://www.timbira.com/
--
Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs
