Hi, Database is MySQL 4.1.5-gamma, running on Linux. All charset variables are set to default 'latin1', except 'character_set_system', which is utf8.
A database was created and running in this environment, but the data it gets (from php) were in utf8. Mysqldump is used without any options to export database to sql text file. When importing back into the same database some characters (non latin,cyrillic in fact) get broken. These are exactly 4 characters that broke, but it is no fun, you know. Any combination of charset setings and database creation both from mysql client and server did not work over this problem. Now my question is: what is the encoding that mysqldump puts data in? I cannot decode it with anything! To me it looks like UTF8, but not quite. Somebody mentioned on this list before, that when nothing is specified when dumping data that is not in the default character set, mysqldump insert something "other"?! What is this exactly? It seems strange, because importing back only few characters are broken, other non-latin are intact. But it is not working OK, in fact it made the data corrupted. Exporting back after import confirm this, since the corrupted characters are replaced by one and the same combination - 0xFFED0 0x3F - for capital letters (0x0410, 0x41D), and 0xFFED1 0x3F for small letters 0x0441 and 0x044F. I know that database and dump should be with specified correct charset, but I look into way to recover in this case, since data are already corrupted and original database is not available. Thanks, Delyan -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
