How to read unsigned 4-byte numbers?

Hello,

In java.io.DataInput, there are methods readUnsignedByte (1-byte read), readUnsignedShort (2-byte read) but there is no readUnsignedInt(4-byte read). If I use readInt() it's a signed value. Any thoughts as to the easiest way I can read an unsigned 4-byte value? OF COURSE I'll store it as a long, but I'm lost. I could do it myself, but that's definitely not what I want to do, as you can imagine.

Tell me they've fixed this in 1.5. Or 5.0. Or whatever.

Thanks!

Rob

wouldn't that work:
?

Probably not since readLong will read 8 bytes from the DataInput source.

What about amending Stefan's suggestion as follows:

I'm sure you can do something clever with the bit shifting operators instead of the second line but I'll leave that to you, if you like.

Jules

Ahh, yeah! D'uh.

I have an ulterior reason for understanding this particular problem. I'll discuss in another thread.

Thanks, RK

[B][Julian]: What about amending Stefan's suggestion as follows:
[/B]
Still two problems. One is that 'l' is a truly horrible name for a variable, since it looks so much like 1.

The other is that the code doesn't work. Try:

or

or

[ August 10, 2004: Message edited by: Jim Yingst ]

Ha! Good point.

Three cheers for testing - hip, hip...

Jules

Originally posted by Jim Yingst:
[B][Julian]: What about amending Stefan's suggestion as follows:
long l = (long)data.readInt(); if (l < 0 ) l += Integer.MAX_VALUE;[/B]
Still two problems. One is that 'l' is a truly horrible name for a variable, since it looks so much like 1. The other is that the code doesn't work. Try:
long val = (long) data.readInt(); if (val < 0 ) val = val + 2 * (Integer.MAX_VALUE + 1L);
or
long val = (long) data.readInt(); if (val < 0 ) val = val - 2 * Integer.MIN_VALUE;
or
long val = (long) data.readInt(); if (val < 0 ) val |= 0X100000000;

[ August 10, 2004: Message edited by: Jim Yingst ]

I believe I understand that last technique (that adding "1" to the 9th bit will force that set bit and all the higher-order ones over to zero, right?).

However, I'd appreciate some explanation of your first two techniques. I presume you're exploiting certain details of two's complement notation, but I haven't figured out just what you're doing there.

will not work, 2nd statement does nothing for Integer.MIN_VALUE<=val<0.

Correct:

or

[ August 14, 2004: Message edited by: Vlado Zajac ]
[ August 14, 2004: Message edited by: Vlado Zajac ]