Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!uxc.cso.uiuc.edu!tank!uwvax!astroatc!brown
From: brown@astroatc.UUCP (Vidiot)
Newsgroups: comp.binaries.ibm.pc.d
Subject: Re: Combine script (was Re: More silence in comp.binaries.ibm.pc)
Message-ID: <2507@astroatc.UUCP>
Date: 3 Aug 89 18:22:54 GMT
References: <8472@bsu-cs.bsu.edu> <2487@astroatc.UUCP> <26918@srcsip.UUCP> <5154@umd5.umd.edu>
Reply-To: brown@astroatc.UUCP (Vidiot)
Organization: Vidiot's Hangout
Lines: 29

In article <5154@umd5.umd.edu> dschuetz@umd5.umd.edu (David Schuetz) writes:
<
<As long as we're talking about combine, could somebody please tell me what
<the "sed '/^END/,/^BEGIN/d'" (or whatever) does?  I figured that it 
<deleted everything not within the BEGIN/END block of the file, but in 
<some tests (just sending the result to output rather then piping it to
<uudecode), I found that it didn't make any changes at all, and that
<uudecode didn't give a damn whether or not there was article header
<garbage in the file.

Look very carefully at the script again.  It is telling sed that the first
line number of the range starts with END in the left column.  The last
line number of the range starts with BEGIN in the left column.  The header
information is NOT thrown away because sed will find the first instance
of END and the end of the first file.  The BEGIN is right before the text
starts of the second file, and so on.  What sed ends up doing is throwing
away the end garbage of one file through the header information of the next
file.  Remember all of the posted pieces are placed into one file first.

Try reversing the END and BEGIN and see what happens to the output.  It will
be like keeping the bath water and throwing out the baby.

I hope this helps explain what is happening.  Just remember that the first
BEGIN is ignored.
-- 
	        harvard\     att!nicmad\
Vidiot            ucbvax!uwvax..........!astroatc!brown
	        rutgers/  decvax!nicmad/
	ARPA/INTERNET: brown%astroatc.UUCP@spool.cs.wisc.edu