Representing numbers without casting in Java

Question

Does a more natural way to represent those numbers without the use of casting exist in Java?

float tolerance = (float) 8.0e-7;
byte[] data = new byte[]{(byte) 0xFF , (byte) 0x15};

Not using casting on those instances just doesn't work at all, and it just looks weird when you are defining constants of a primitive type and you have to use casting to represent it (it is like you are saying that Java doesn't support those data types naturally which is ridiculous).

I understand that for the float number you can represent it like this:

float tolerance = 0.0000008f;

but it is just almost unreadable in this form.

`0x15` will standalone without the cast. `0xff` needs the cast because bytes are signed in java and gets interpreted as `-1` — kjsebastian, Feb 27 '17 at 09:42
@conscells Followup: So `-0x01` would actually work without cast as replacement for `(byte) 0xFF`. Or just `-1`. — Andreas, Feb 27 '17 at 09:46
@conscells I actually didn't know that there is cases that I don't need to cast bytes, that was new information for me. — Mohamed Talaat Harb, Feb 27 '17 at 11:26

score 2 · Accepted Answer · edited May 23 '17 at 12:25

2

For float values, you can simply express within the literal that you want a float, and not a double (which requires the cast):

float tolerance =  8.0e-7F;

There is no literal for bytes though (see here).

edited May 23 '17 at 12:25

Community

1
1

answered Feb 27 '17 at 09:40

GhostCat

137,827
25
176
248

You can even drop the `.0` when using `f`. – Andreas Feb 27 '17 at 09:42
Ah, I see. Thanks. But he asked this way; so I will keep it for the moment. – GhostCat Feb 27 '17 at 09:43
I wish a similar solution existed for bytes but thank you for your answer @GhostCat – Mohamed Talaat Harb Feb 27 '17 at 11:24

Representing numbers without casting in Java

1 Answers1