jimomura@lsuc.UUCP (10/28/87)
If you've been working with the Color Computer version of
Microware C and have tried to write a program reading irregular
amounts of data through an SCF device, you've probably run into
this problem. The original compiler spec in the manuals says
that getstat(1,path) returns -1 for no data and 0 for data waiting.
This is true. However, it also says that read(path,buffer,nnn)
will return the number of bytes actually read -- which seems
to imply that it will return having read fewer bytes than those
specified by nnn. This doesn't work. It'll sit there waiting
for all the bytes to arrive. Now, the assembly language I$GetStt
call returns the actual numbers waiting to be read, which if you
use it, means you can specify what you need to receive from
read(), but no way is provided in the standard libraries. So,
I've written a short program which can be used in place of
getstat(1,path). It's patterned after _gs_rdy() which comes in
the 68K Microware compiler, so using it should also enhance
portability. Having read all this preamble, you'll be disappointed
when you see how short the program is, but, that's life. ;-)
/* gs_rdy.c
* 1987/10/26
* Public Domain
*/
/* By Jim Omura, Byte Information Exchange 'jimomura'
* 2A King George's Drive
* Toronto, Ontario
* M6M 2G9
*/
/* _gs_rdy()
*
* Usage:
*
* int _gs_rdy(path)
*
* int path;
*
* Returns the number of bytes waiting to be read.
*
* To be used in place of "getstat(1,path)" in order
* to work with read().
*/
int _gs_rdy(path)
int path;
{
int waiting;
#asm
leas -2,s stack points at "waiting"
lda 7,s load "path" -- actually ldd 6 and shift left 8.
ldb #1
os9 $8d I$GetStt
bcs errhand
lda #0 clear upper byte to make sense.
std 0,s
bra cleanexit
errhand ldd #-1
std 0,s
cleanexit leas 2,s Restore stack pointer.
#endasm
return(waiting);
}
--
Jim Omura, 2A King George's Drive, Toronto, (416) 652-3880
ihnp4!utzoo!lsuc!jimomura
Byte Information eXchange: jimomura