note 110075 deleted from language.variables.basics by googleguy

Note Submitter: Olivier Miakinen ---- Answer to previous remark (2012-06-20). Actually, the regular expression *is* correct, but it applies to bytes (as correctly stated in 'Note'), not to characters. In your example $Ελληνικά, although the Ε is Unicode U+0395, which is greater than U+00FF, its UTF-8 representation uses two bytes CE and 95, both of them in range 80-FF. Similarly, the λ is U+03BB, but its UTF-8 representation uses two bytes CE and BB. The complete UTF-8 representation for Ελληνικά is : CE 95 CE BB CE BB CE B7 CE BD CE B9 CE BA CE AC All this perfectly conforms to the regular expression given.