Print chess symbols using UnicodeBlock?

Question

With jdk12, came Chess symbols (source):

Unicode 11.0.0 introduced the following new features that are now included in JDK 12

[...] 4 blocks for the following existing scripts:

Georgian Extended

Mayan Numerals

ndic Siyaq Numbers

Chess Symbols

With that in mind, I tried to print those characters with the following code, to test the functionality and use those later in a little chess game:

Character.UnicodeBlock block = Character.UnicodeBlock.CHESS_SYMBOLS;
for (int i = 0; i < 1114112; i++) {
    char unicode = (char) i;
    if(Character.UnicodeBlock.of(unicode) == block) {
        System.out.println(unicode);
    }
}

However, it is not printing anything. The code works if I replace CHESS_SYMBOLS with, for instance, ARABIC. I have java 12.0.1.

Question: Why isn't the above code printing anything ?

https://stackoverflow.com/questions/7665120/displaying-chess-pieces-with-unicode-in-eclipse-using-java — Steve Smith, May 14 '19 at 12:53
@StephenC I'm using Intellij, if I try `System.out.println("♔");` it does print the symbol, so I believe it is strictly code-related — Paul Lemarchand, May 14 '19 at 12:56
@SteveSmith It is not printing rectangles, it is printing nothing at all. Also I'd like to use `UniblockCode` to do the printing, or at least understand why it doesn't work. — Paul Lemarchand, May 14 '19 at 12:57
You are only checking the first 65536 characters there are 1,114,112 Unicode characters. — JGNI, May 14 '19 at 13:03
@JGNI I edited the code, still not printing though. I also tried `Character.MAX_VALUE`. — Paul Lemarchand, May 14 '19 at 13:04
@SteveSmith If `System.out.println("♔");` does work then it must not be a charset issue, or maybe something has to be specified in code ? — Paul Lemarchand, May 14 '19 at 13:20
Maybe the Windows charset doesn't contain all the unicode characters? https://coderanch.com/t/538165/java/Displaying-UniCode-characters — Steve Smith, May 14 '19 at 13:23

score 8 · Accepted Answer · answered May 14 '19 at 13:29

Some chess symbol characters exist in the Miscellaneous Symbols block, but you are specifically checking for 16-bit char values in a different block. The Chess Symbols block contains zero characters with 16-bit values; it starts at U+1FA00, and ends at U+1FA6F.

By casting to char, you are trimming all values above U+FFFF to their lowest 16 bits; for example, if i is 0x1fa60, casting it to a char will make it 0xfa60, which prevents your block check from succeeding.

To make your code work, you need to stop assuming that all codepoints are 16-bit values. You can do that by changing this:

char unicode = (char) i;

to this:

int unicode = i;

Of course, when changing `char unicode = (char) i;` to `int unicode = i;`, there is no need for two distinct variables at all. — Holger, May 14 '19 at 15:41

Karol Dowbecki · Answer 2 · 2019-05-14T13:41:42.470

Unfortunately Character.UnicodeBlock doesn't have methods to tell what is the beginning and ending value for code points within the block. In Unicode 11 the chess symbols block runs from U+1FA00 to U+1FA6D.

Java uses UTF-16 and surrogate pairs to represent characters over U+10000. In this case code point U+1FA00 will be represented as two char values: U+D83E (high surrogate) and U+DE60 (low surrogate).

You should use Character.toChars() to correctly print the code point which is always an int:

Character.UnicodeBlock block = Character.UnicodeBlock.CHESS_SYMBOLS;
for (int i = 0; i < 1114112; i++) {
    if (Character.UnicodeBlock.of(i).equals(block)) {
        System.out.println(Character.toChars(i));
    }
}

Print chess symbols using UnicodeBlock?

2 Answers2