byte array
in [Java Programming]
|
From: Arne Vajhøj on 28 Mar 2010 08:33 On 27-03-2010 23:06, Mike Amling wrote: > Mike Schilling wrote: >> Peter Duniho wrote: >>> You can fix the code in at least one of two ways: >>> >>> � cast the literal to byte, forcing the conversion: "myArray[0] = >>> (byte)0xc9;" >>> � specify a literal that can fit in a byte: "myArray[0] = >>> 0xffffffc9;" >> >> How intuitive that an 8-bit literal doesn't fit into an 8-bit byte, >> but a 32-bit literal does ;-) >> >> A bit more seriously, I presume that 99% of people would use the cast >> rather than try to type the right number of 'f's. > > To avoid the counting of Fs, you can also use > > myByteArray[0]=0xC9-256; > > to assign 0x80..0xFF to a byte. My suggestion for easiest and most readable code is to cast all byte values from 0x00 to 0xFF. And occasionally spend some minutes cursing whoever decided that byte was unsigned. Arne
From: Lew on 29 Mar 2010 15:06 Peter Duniho wrote: >> You can fix the code in at least one of two ways: > >> cast the literal to byte, forcing the conversion: "myArray[0] = >> (byte)0xc9;" >> specify a literal that can fit in a byte: "myArray[0] = >> 0xffffffc9;" > Mike Schilling wrote: > How intuitive that an 8-bit literal doesn't fit into an 8-bit byte, but a > 32-bit literal does ;-) > Neither of the literals in Pete's example is an eight-bit literal. > A bit more seriously, I presume that 99% of people would use the cast rather > than try to type the right number of 'f's. And the other 1% should have done so. Code isn't just about making the computer do the right thing, it's about documenting that thing for future maintenance programmers. -- Lew
From: Daniel Pitts on 29 Mar 2010 16:51 On 3/28/2010 5:33 AM, Arne Vajh�j wrote: > On 27-03-2010 23:06, Mike Amling wrote: >> Mike Schilling wrote: >>> Peter Duniho wrote: >>>> You can fix the code in at least one of two ways: >>>> >>>> � cast the literal to byte, forcing the conversion: "myArray[0] = >>>> (byte)0xc9;" >>>> � specify a literal that can fit in a byte: "myArray[0] = >>>> 0xffffffc9;" >>> >>> How intuitive that an 8-bit literal doesn't fit into an 8-bit byte, >>> but a 32-bit literal does ;-) >>> >>> A bit more seriously, I presume that 99% of people would use the cast >>> rather than try to type the right number of 'f's. >> >> To avoid the counting of Fs, you can also use >> >> myByteArray[0]=0xC9-256; >> >> to assign 0x80..0xFF to a byte. > > My suggestion for easiest and most readable code is to cast > all byte values from 0x00 to 0xFF. > > And occasionally spend some minutes cursing whoever > decided that byte was unsigned. you mean signed, right? The only unsigned primitive numeric data type in Java is char. -- Daniel Pitts' Tech Blog: <http://virtualinfinity.net/wordpress/>
From: Arne Vajhøj on 29 Mar 2010 19:46 On 29-03-2010 16:51, Daniel Pitts wrote: > On 3/28/2010 5:33 AM, Arne Vajh�j wrote: >> On 27-03-2010 23:06, Mike Amling wrote: >>> Mike Schilling wrote: >>>> Peter Duniho wrote: >>>>> You can fix the code in at least one of two ways: >>>>> >>>>> � cast the literal to byte, forcing the conversion: "myArray[0] = >>>>> (byte)0xc9;" >>>>> � specify a literal that can fit in a byte: "myArray[0] = >>>>> 0xffffffc9;" >>>> >>>> How intuitive that an 8-bit literal doesn't fit into an 8-bit byte, >>>> but a 32-bit literal does ;-) >>>> >>>> A bit more seriously, I presume that 99% of people would use the cast >>>> rather than try to type the right number of 'f's. >>> >>> To avoid the counting of Fs, you can also use >>> >>> myByteArray[0]=0xC9-256; >>> >>> to assign 0x80..0xFF to a byte. >> >> My suggestion for easiest and most readable code is to cast >> all byte values from 0x00 to 0xFF. >> >> And occasionally spend some minutes cursing whoever >> decided that byte was unsigned. > you mean signed, right? Yes. :-( Arne
From: Kevin McMurtrie on 29 Mar 2010 21:28 In article <4bb13bdb$0$283$14726298(a)news.sunsite.dk>, Arne Vajh�j <arne(a)vajhoej.dk> wrote: > On 29-03-2010 16:51, Daniel Pitts wrote: > > On 3/28/2010 5:33 AM, Arne Vajh�j wrote: > >> On 27-03-2010 23:06, Mike Amling wrote: > >>> Mike Schilling wrote: > >>>> Peter Duniho wrote: > >>>>> You can fix the code in at least one of two ways: > >>>>> > >>>>> � cast the literal to byte, forcing the conversion: "myArray[0] = > >>>>> (byte)0xc9;" > >>>>> � specify a literal that can fit in a byte: "myArray[0] = > >>>>> 0xffffffc9;" > >>>> > >>>> How intuitive that an 8-bit literal doesn't fit into an 8-bit byte, > >>>> but a 32-bit literal does ;-) > >>>> > >>>> A bit more seriously, I presume that 99% of people would use the cast > >>>> rather than try to type the right number of 'f's. > >>> > >>> To avoid the counting of Fs, you can also use > >>> > >>> myByteArray[0]=0xC9-256; > >>> > >>> to assign 0x80..0xFF to a byte. > >> > >> My suggestion for easiest and most readable code is to cast > >> all byte values from 0x00 to 0xFF. > >> > >> And occasionally spend some minutes cursing whoever > >> decided that byte was unsigned. > > you mean signed, right? > > Yes. > > :-( > > Arne As somebody who occasionally writes data processing code, I'll join you in some cursing. I also curse not being able to allocate arrays of object instances. Sun's GC algorithms barely keep pace with high-end use. You'd think this would be a priority feature to help application servers and number crunchers. -- I won't see Google Groups replies because I must filter them as spam
First
|
Prev
|
Pages: 1 2 Prev: Writing jnlp program for both sandbox and all-permissions Next: BufferedReader vs NIO Buffer |