🚀 go-pugleaf

You need to do it differently in Java.  

private short swap2(short s)
    {
        short tmp1, tmp2;

        tmp1 = (s & 0xFF) << 8;
        tmp2 = (s >> 8) & 0xFF;
        return (tmp1 | tmp2);
    }

and call it with
sval = swap2(sval);

I think that will work, and a similar procedure would work for 4 byte
numbers (int in Java).  Depending on what you are doing be careful
with the sign (+/-).  

To convert between floats and ints see Float.floatToIntBits and
Float.intBitsToFloat (the JDK documentation explains the format of the
int).

On Tue, 16 Mar 1999 09:45:23 -0700, Doug Bath <dbath@agt.net> wrote:

>Greetings
>
>I am trying to convert an old C++ app to java and I have hit a snag. I
>am writing a class to read seismic data in SEGY format which consists of
>short integers, long integers, and floats in IBM and IEEE format. These
>values can be in 'big endian' or 'little endian' byte order.
>In my C++ app I wrote simple byte swap routines for 2 byte or 4 byte
>values that looke like
>
>void TSegyFile::swap2(unsigned char  *str)
>/*      byte swap 16 bit integers */
>{
>    unsigned char tmp;
>
>    tmp=str[0];
>    str[0]=str[1];
>    str[1]=tmp;    //    wrote a similar function for 4 byte values
>}
>
>To call the function I cast the short integer to an unsigned char as
>
>swap2((unsigned char*)&sval);
>
>When I try to do something similar in java
>private void swap2(byte[] s)
>    {
>        byte tmp;
>
>        tmp=s[0];
>        s[0]=s[1];
>        s[1]=tmp;
>    }
>
>and call it with
>swap2((byte[]) sval);
>
>I get a message telling me I can't make such a  cast.
>
>I haven't even got to the point where I have to cast a float to a long
>int to do the IBM to IEEE conversion but I am already dreading it.
>
>Any insights would be most appreciated.
>
>
>Doug Bath
>

Chris Hill
Chris244@aol.com