I don't understand what -K and /u mean.
The pickaxe says -K "specifies the code set to be used", but to be used for what? Specifies the code set of the source code? It affects to the internal representation of strings? It affects regexps?
Besides those questions, I played a bit around to try to figure this out without luck:
I saved a simple Ruby test script in UTF-8. This test script has a simple regexp with an accented character that filters external test data in Latin-1 converted to UTF-8 with iconv. The matching succeeded no matter whether I used -Ku or /u. In any of the four combinations it succeeded. Why?
As you see I don't have the ideas ordered about this, if it is more simple to explain the picture instead of answering that please feel free.
-- fxn