RFR (XXS) 8080535: (ch) Expected size of Character.UnicodeBlock.map is not optimal

[Ivan Gerasimov]

Hi everyone!

What about this variant:
http://cr.openjdk.java.net/~igerasim/8080535/02/webrev/

No named constant.
A comment in the code about initial capacity.
In the test, we use the same constant to check if it were sufficient to 
hold the final number of entries.

If someone evil shrinks the initial capacity in the code, the test will 
not be able to detect it, though it seems unlikely.

// Please ignore the change to test/TEST.groups in the webrev; it is a 
leftover from a fix for JDK-8080330

Sincerely yours,
Ivan