Re: I18n problems: Working with double-byte languages

Ricardo Junior wrote:

Hi people ! I have a query regarding double-byte languages on PHP. I need to sort a index localized from English to Korean. My idea is: * get the Korean strings from file; * convert them to UTF8; * insert them in a Oracle database set to work with UTF8; * set "NLS_LANG" with "Alter session" SQL command to this variable match with Korean language; * get the strings sorted by Oracle with a "SELECT * FROM table SORT BY field ASC" SQL command; I've tried to convert the Korean string to UTF8 with utf8_encode function, but this function converted each byte from double-byte string to its relative in UTF8 chars. Reverting this string encoded to UTF8 with utf8_decode function, the browser can display successfully the Korean chars (because the individual bytes of the double-byte string will be as them were before), but Oracle can't sort the strings properly because that UTF8 chars converted by utf8_encode function were not relative to a double-byte char, but relative to a 2 single bytes chars! How can I convert a double-byte string to UTF8 properly??? Really thanks !!!!

UTF-8 works but EUC-KR may be better. Anyway, take a look at mbstring, iconv, gettext modules. -- Yasuo Ohgaki