jdries%doppler@faatcrl (J. Francis Dries, III) (10/26/90)
I was hacking up a program to convert Sun Rasterfiles to GIF files when
I ran into a little problem. Almost all the non-Intel world (or so I'm
told) writes ints MSByte first, whereas Borland's TC++ is
reading/writing ints/longs (I need longs, since I'm working with 32bit
ints from the Sun) LSByte first.
My Question is, is there a way to reverse the order, quickly or cleanly?
I could read in each char at a time and OR it onto the long and SHL all
4 on, but that seems kinda sloppy to me. I would think that the is a
flag that can be set, or at least a built in macro...
Any ideas?
---
J. Francis Dries, III
...!rutgers!faatcrl!doppler!jdries jdries%doppler@faatcrl
"You can put a bullet in my head, but you can't kill a word I said"
-- Mike Muirdmurdoch@watstat.waterloo.edu (Duncan Murdoch) (10/26/90)
In article <i6TmR3w163w@doppler> jdries%doppler@faatcrl (J. Francis Dries, III) writes: >I was hacking up a program to convert Sun Rasterfiles to GIF files when >I ran into a little problem. Almost all the non-Intel world (or so I'm >told) writes ints MSByte first, whereas Borland's TC++ is >reading/writing ints/longs (I need longs, since I'm working with 32bit >ints from the Sun) LSByte first. > >My Question is, is there a way to reverse the order, quickly or cleanly? >I could read in each char at a time and OR it onto the long and SHL all >4 on, but that seems kinda sloppy to me. I would think that the is a >flag that can be set, or at least a built in macro... The i486 has a new instruction BSWAP that does exactly what you want in 1 clock (if the data is already in a register), but I suppose that doesn't help you much. I'd do it by loading the two word sized parts, exchanging the bytes of each, and writing them out again. Something like: mov ax, word ptr [data] mov dx, word ptr [data+2] xchg al,ah xchg dl,dh mov word ptr[data],dx mov word ptr[data+2],ax
teittinen@cc.helsinki.fi (10/29/90)
In article <i6TmR3w163w@doppler>, jdries%doppler@faatcrl (J. Francis Dries, III) writes: > Almost all the non-Intel world (or so I'm > told) writes ints MSByte first, whereas Borland's TC++ is > reading/writing ints/longs (I need longs, since I'm working with 32bit > ints from the Sun) LSByte first. > > My Question is, is there a way to reverse the order, quickly or cleanly? > > Any ideas? If my memory serves me right (I don't have the manual with me right now) there is a library routine named swab() which does the byte swapping for you. I think that it is very interesting to glimpse through the manuals once in a while and find out some interesting and useful library functions... -- E-Mail: teittinen@finuh.bitnet ! "Studying is the only way teittinen@cc.helsinki.fi ! to do nothing without Marko Teittinen, student of computer science ! anyone blaming you" -me
dfoster@jarthur.Claremont.EDU (Derek R. Foster) (10/31/90)
In article <i6TmR3w163w@doppler> jdries%doppler@faatcrl (J. Francis Dries, III) writes: >I was hacking up a program to convert Sun Rasterfiles to GIF files when >I ran into a little problem. Almost all the non-Intel world (or so I'm >told) writes ints MSByte first, whereas Borland's TC++ is >reading/writing ints/longs (I need longs, since I'm working with 32bit >ints from the Sun) LSByte first. > >My Question is, is there a way to reverse the order, quickly or cleanly? In the case of ints, look up the "swab" function. I don't know about longs, though. Derek Riippa Foster