🚀 go-pugleaf

>>>>> On 26 Jul 92 22:11:56 GMT, merritt@climate.gsfc.nasa.gov (John H. Merritt) said:

John> In article <1992Jul26.033811.21614@nsisrv.gsfc.nasa.gov>, merritt@climate.gsfc.nasa.gov (John H. Merritt) writes:
John> |> Explain why the following 'nawk' program produces:
John> [ output deleted ]

John> Already there is misunderstanding about what I want in the way of answers.
John> I want a solution to getting the same output when seeing the output
John> at the terminal or redirecting it to a file.  
John> I WANT THE PIPE TO REMAIN.

Then your solution is to pipe everything through the pipe and it will
come out at the same time.  (Unfortunately you don't want to sort your
begin and end parts though.)

Your problem is illustrated here as a data flow diagram

	      BEGIN and END ----------> OUTPUT1
	    / 					\
data -> [awk] 					 > terminal output
	    \ MIDDLE --------> [cat] -> OUTPUT2 /

As soon as you have two processes (when awk forks the cat process as a
child) the branches get out of step.

Sort will take a long time to run as it has to read all its input
before starting to output. 


John> If there is something like 'fflush' in AWK, that would be great.

There isn't. Besides awk isn't producing the output of sort. 

John> So far 'close' was the closest thing I found, but the exact syntax eludes
John> me -- if it is the right thing at all.

That lets you close a file, but I don't think you can close a pipe.

I think you will have to settle for three lots of output, (BEGIN,
MIDDLE and END say) and join them together when the all processes have
exited.

Awk might exit before the cat or sort pipeline has however, so you
must move the pipe outside of awk.

Make your awk program do this

	BEGIN	do begin stuff > begin.tmp

	MIDDLE  do middle stuff to standard out 

	END 	do end stuff > end.tmp


Then run it like this

nawk -f awkfile  | sort > sort.tmp && cat begin.tmp sort.tmp end.tmp > finaloutput


John> -- 
John> \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
John> John H. Merritt --> merritt@climate.gsfc.nasa.gov
John> "I am generally intolerant of ignorance,
John> but I have made an exception in your case."




--
 Regards,			  pd@x.co.uk           IXI Ltd.
	Paul Davey		  pd@ixi.uucp          62-74 Burleigh St.
				  ...!uunet!ixi!pd     Cambridge,  U.K.
 "These _are_ interesting times"  +44 223 462 131      CB1  1OJ