Extract Chinese Characters from a string

Does anyone have a suggestion for extracting Chinese characters from a text string? I have some documents imported from Word that have Chinese translations in the same paragraph as the English text. I am trying to write a routine to split the Chinese into a separate attribute, but it is proving tricky to reliably detect the Chinese characters. Any Ideas?
SystemAdmin - Tue Nov 27 23:25:20 EST 2012

Re: Extract Chinese Characters from a string
SystemAdmin - Tue Dec 18 07:42:31 EST 2012

One solution may be to use the isValidChar() DXL perm in combination with the Latin1 codepage, since Chinese characters will not have a Latin1 codepage entry.

I've attached a sample which splits the Object Text attribute into attributes called "Chinese" and "English". Its not perfect, handling of space and full-stop characters would need to be added.
Attachments

attachment_14921119_findNonLatin.dxl