mark at indymedia.org
Thu Jul 28 12:14:03 PDT 2005
seems like now might be a good time to "force" all sf-active
sites & databases to be utf8.. but at least make it an easily
i changed radio.indymedia.org to be a utf-8 site, it was easy.
i wanted to change the db to utf8 as well so that e.g.
searches and everything work seamlessly:
run a sql command for each table:
Alter table webcast modify column article text character
set utf8, modify column summary text character
set utf8 [etc.]
note, if your database has something aside from latin1, this
probably won't work right. it'll get garbled. there are
various docs online to help you deal w/ this.. so at least
make sure you backup the old database first..!
you also might as well set the default character set to utf8
for each table and the database as a whole, so if something
is added in the future it will be utf8.
add this to get_connection() in db_class.inc:
mysql_query('SET CHARACTER SET utf8');
and you could also add
mysql_query('SET NAMES utf8');
change the <meta> header for your web and admin pages to
put this in your httpd.conf:
so your server is sending out the correct header.
remove any utf_encode() from syndication classes. also remove
any translation being done for windows characters in posts.
this stuff will all be utf-8 now, so it's not necessary.
if you are aggregating feeds onto your site, remove any
utf_decode() functions from these.
that's it. well of course there are some other considerations
re: multibyte text but this is basic functionality.
this stuff can easily be placed in "if" statements based on
sfactive.cfg but might be worth forcing utf8 on everyone at
More information about the imc-sf-active