[TYPO3-english] Charset encoding problem since 5.3.6

[Jigal van Hemert]

Hi,

On 8-7-2011 15:18, David ARNOULT - Edition & Internet Operations wrote:
> Despite my concern, it seems that MySQl-TYPO3 connection is in latin1 mode
> for all my languages, Chinese included...

It happened quite often in the past that the BE forceCharset was set to 
utf-8, but that the database tables and the connection was still set to 
latin1. This resulted in utf-8 encoded data in latin1 tables.

Years ago I made a script [1] to "convert" such databases to utf-8, 
including the wrongly encoded data. The "conversion" itself is done in 
MySQL, so it is not influenced by connection charsets, etc.

It was pretty hard for some people to see if this problem existed in the 
database. Some time ago I made a script to detect the problem. It is 
however still experimental! It cannot detect the problem with 100% 
certainty.

Both scripts are meant to run from a subdirectory of your TYPO3 
installation (e.g. from fileadmin). They pick up the database 
credentials automatically.

Needless to say backup your database before running the db_utf8_fix 
script for real! It is in test mode by default. You have to change a 
constant in the beginning of the script to let it work for real.

[1] http://dcbjht.home.xs4all.nl/typo3/db_utf8_fix.zip
[2] http://dcbjht.home.xs4all.nl/typo3/db_utf8_test.zip

-- 
Kind regards / met vriendelijke groet,

Jigal van Hemert.