Обсуждение: PostgreSQL Naming Rules

Robert Buckley, 28.10.2011 12:49:
> according to this article
> http://www.informit.com/articles/article.aspx?p=409471, the naming of
> tables, and fields is restricted to 63 characters and must start with
> an underscore or letter. Nothing is however said about in which
> character set.
>
> Am I allowed to name a table field < Änderung_1 >. The Ä is a german
> letter contained within the UTF8 character set.

It does work, but I have had some tools (JDBC based) that choked on those tables. And I had some troubles running
pg_dump/pg_restorewith those tables. But as I did that merely out of curiosity, I didn't bother to find a way on how to
dealwith them. 

Thomas

Robert Buckley wrote:
> according to this article http://www.informit.com/articles/article.aspx?p=409471, the naming of
> tables, and fields is restricted to 63 characters and must start with an underscore or letter. Nothing
> is however said about in which character set.
>
> Am I allowed to name a table field < Änderung_1 >. The Ä is a german letter contained within the UTF8
> character set.

You can, but it's a really bad idea to have non-ASCII names.

Yours,
Laurenz Albe

On Friday, October 28, 2011 3:49:58 am Robert Buckley wrote:
> Hi,
> according to this
> article http://www.informit.com/articles/article.aspx?p=409471, the naming
> of tables, and fields is restricted to 63 characters and must start with
> an underscore or letter. Nothing is however said about in which character
> set.
>
> Am I allowed to name a table field < Änderung_1 >. The Ä is a german letter
> contained within the UTF8 character set.


http://www.postgresql.org/docs/9.1/interactive/sql-syntax-lexical.html#SQL-
SYNTAX-IDENTIFIERS

"SQL identifiers and key words must begin with a letter (a-z, but also letters
with diacritical marks and non-Latin letters) or an underscore (_). Subsequent
characters in an identifier or key word can be letters, underscores, digits
(0-9), or dollar signs ($). Note that dollar signs are not allowed in identifiers
according to the letter of the SQL standard, so their use might render
applications less portable. The SQL standard will not define a key word that
contains digits or starts or ends with an underscore, so identifiers of this form
are safe against possible conflict with future extensions of the standard. "

>
> yours,
>
>
> Rob

--
Adrian Klaver
adrian.klaver@gmail.com

On 10/28/2011 06:49 PM, Robert Buckley wrote:

> according to this article
> http://www.informit.com/articles/article.aspx?p=409471, the naming of
> tables, and fields is restricted to 63 characters and must start with an
> underscore or letter. Nothing is however said about in which character set.
>
> Am I allowed to name a table field < Änderung_1 >. The Ä is a german
> letter contained within the UTF8 character set.

It'll work, but sadly you will at some point run into badly written
tools that mangle anything but 7-bit-ascii text.

psql on the Windows command line needs some extra love to behave
correctly; you may have to issue a chcp command to change the console to
unicode before starting psql if you're working with data that cannot be
represented in the charset your version of Windows uses. That won't be
an issue for umlauts etc, but if you wanted (say) hebrew table names on
a German windows you'd probably have to change the codepage.

Other than that, PostgreSQL and its tools should be fine. You are likely
to run into badly written tools elsewhere that mangle anything that
isn't 7-bit ascii, though, so be careful.

In general, if PostgreSQL doesn't like a table identifier you can
"double quote" it to get it to accept it. Table identifiers in
double-quotes also have their case preserved rather than automatically
being lower-cased like unquoted identifiers. Because the downcasing
rules for anything except 7-bit ascii are less than clear, I'd recommend
double-quoting all identifiers that use characters outside that range.

--
Craig Ringer