Comments on: MySQL encoding http://fschiettecatte.wordpress.com/2007/05/18/mysql-encoding/ Thoughts from the edge of the 'net Fri, 16 May 2008 20:51:24 +0000 http://wordpress.org/?v=MU By: François Schiettecatte http://fschiettecatte.wordpress.com/2007/05/18/mysql-encoding/#comment-2212 François Schiettecatte Tue, 28 Aug 2007 11:42:57 +0000 http://fschiettecatte.wordpress.com/2007/05/18/mysql-encoding/#comment-2212 Joseph Thanks, that would be my understanding too, but the documentation is silent on the issue. I would assume that they have done the "right thing" though. Joseph

Thanks, that would be my understanding too, but the documentation is silent on the issue. I would assume that they have done the “right thing” though.

]]> By: Joseph Riesen http://fschiettecatte.wordpress.com/2007/05/18/mysql-encoding/#comment-2207 Joseph Riesen Tue, 28 Aug 2007 08:29:46 +0000 http://fschiettecatte.wordpress.com/2007/05/18/mysql-encoding/#comment-2207 Thanks for the information! I was looking for these assorted options (to force a development server to talk UTF-8) and found your page here. FYI, read about the space requirements for UTF-8 at: http://dev.mysql.com/doc/refman/5.0/en/charset-unicode.html Specifically, it states: "Tip: To save space with UTF-8, use VARCHAR instead of CHAR. Otherwise, MySQL must reserve three bytes for each character in a CHAR CHARACTER SET utf8 column because that is the maximum possible length. For example, MySQL must reserve 30 bytes for a CHAR(10) CHARACTER SET utf8 column." In other words, you _are_ correct in stating that a CHAR(20) in UTF-8 will require 60 bytes, but a VARCHAR will continue to reserve only the number of bytes necessary to hold the string. In other words, a VARCHAR(20) _could_ use up to 60 bytes of storage, but any lower-ASCII characters will still only take up one byte. Or at least, so I understand it. YMMV. =) Thanks for the information! I was looking for these assorted options (to force a development server to talk UTF-8) and found your page here.

FYI, read about the space requirements for UTF-8 at:
http://dev.mysql.com/doc/refman/5.0/en/charset-unicode.html

Specifically, it states:
“Tip: To save space with UTF-8, use VARCHAR instead of CHAR. Otherwise, MySQL must reserve three bytes for each character in a CHAR CHARACTER SET utf8 column because that is the maximum possible length. For example, MySQL must reserve 30 bytes for a CHAR(10) CHARACTER SET utf8 column.”

In other words, you _are_ correct in stating that a CHAR(20) in UTF-8 will require 60 bytes, but a VARCHAR will continue to reserve only the number of bytes necessary to hold the string. In other words, a VARCHAR(20) _could_ use up to 60 bytes of storage, but any lower-ASCII characters will still only take up one byte.

Or at least, so I understand it. YMMV. =)

]]> By: noel http://fschiettecatte.wordpress.com/2007/05/18/mysql-encoding/#comment-863 noel Sun, 03 Jun 2007 06:31:52 +0000 http://fschiettecatte.wordpress.com/2007/05/18/mysql-encoding/#comment-863 aww man I was dealing with this very issue last year and ran across the same article. One thing I read though was it's better to avoid char in UTF and just make everything varchar because varchar won't use more bytes than neccessary on standard ascii/latin chars. Of course you lose the performance boost of setting all fields to char but the only place I ever saw anyone do that was at that one rss search engine company... Here's the article I read: http://www.oreillynet.com/onlamp/blog/2006/01/turning_mysql_data_in_latin1_t.html It includes a few handy extras too like: httpd.conf: AddCharset UTF-8 .utf8 AddDefaultCharset UTF-8 php.ini default_charset = "utf-8" my.cnf character-set-server=utf8 default-collation=utf8_unicode_ci Also if you don't want everything connecting to the mysqld to get set to utf8, you can fire off a query immediately after connecting: "SET NAMES utf8" . This is good if you have some legacy crap you can't/won't/don't want to change. aww man I was dealing with this very issue last year and ran across the same article. One thing I read though was it’s better to avoid char in UTF and just make everything varchar because varchar won’t use more bytes than neccessary on standard ascii/latin chars. Of course you lose the performance boost of setting all fields to char but the only place I ever saw anyone do that was at that one rss search engine company…

Here’s the article I read:
http://www.oreillynet.com/onlamp/blog/2006/01/turning_mysql_data_in_latin1_t.html

It includes a few handy extras too like:

httpd.conf:
AddCharset UTF-8 .utf8
AddDefaultCharset UTF-8

php.ini
default_charset = “utf-8″

my.cnf
character-set-server=utf8
default-collation=utf8_unicode_ci

Also if you don’t want everything connecting to the mysqld to get set to utf8, you can fire off a query immediately after connecting: “SET NAMES utf8″ .

This is good if you have some legacy crap you can’t/won’t/don’t want to change.

]]>