Quote:
Originally Posted by Chompy
Thanks for pointing that out haha, didn't realize. 
|
Ah, well, it's not that you can't mix lower and upper case in ranges, it's just that B < a lexicographically, so the correct range would be [B-a], which includes B, C, ... Z, [, \, ], ^, _ , `, and finally a.
I think what you wanted for [a-B] is actually [ABa-z].
e: I've done some testing and transliteration to e.g.
PHP Code:
[a-z] -> getascii(c) in |getascii("a"), getascii("z")|
works properly with arbitrary bounds (as long as they are properly parsed*) and even "broken" ranges like [a-B].
*Consider [[-]] and [\[-\]].