awk or sed 1

[columb]

I'm tidying up some old code (not mine!) and I've come across

#Remove tabs (\t or \011) and 'squeeze' out line feeds (\n or \012)
cat $tmp/pbecsu.tmp.$$|tr "\011" " "|tr -s "\012"|grep PBEC|grep -v typopn >> $tmp/pbec.tmp.$$

rm -f $tmp/pbecsu.tmp.$$

#Put quotes around each field and remove spaces, except for first 2 records
sed -e '2,$ s/^/"/; 2,$ s/;/","/g; 2,$ s/$/"/; 3,$ s/ //g' $tmp/pbec.tmp.$$ > $FNAME

Ignoring the bit about the first two records (I'm sorting that out elsewhere) and the way it only deletes spaces from the third line on I'm looking to combine all this into one awk or sed statement. If it's awk then

awk '/PBEC/ && ! /typopn/ { .... }' input > output

but I can't work out how to put a sed type statement in the braces. On the other hand, if I use sed then I can't work out how to do the grep like statements. I've been buried in the lemur book for hours and I'm getting nowhere. Any pointers please.

Ceci n'est pas une signature
Columb Healy

 

[PHV]

Have a look at the sub and gsub functions.

Hope This Helps, PH.
FAQ219-2884
FAQ181-2886

 

[columb]

Thanks PH - if you knew how much money (in terms of my time) you have saved my company in this and many other posts you would be sending me a bill!

Ceci n'est pas une signature
Columb Healy