I agree with Joel that it's a shame to call Chinese characters by the Japanese term for "Chinese characters." Of course, it is also a shame that Unicode has sanctified use of "CJK Ideographs" although many students of Chinese culture would prefer "logographs": from their very beginning, most Chinese characters were not "pictures," as Chinese scholars recognized long, long ago.
As for the GREP metacharacter "~K", I guess it goes back to IDCS3 (for search only, not replace) and was paired with "^K" in IDCS4 Text searches. Who knew? I wish I did. Maybe David Blatner did but chose not to mention it in his October 2009 InDesign Secrets piece on using GREP to find Japanese with Unicode text ranges. The interesting point there, of course, is that the ranges are discontinuous.
Is there a way to find out just what ranges "~K" covers? A quick test suggests in IDCS4, "~K" doesn't find the 42,711 characters added in CJK Extension B in Unicode 3.1 (2001) -- though you can find them with "[\x{20000}-\x{2A6DF}]+".
David
