I want to deal with several data, i.e., data.*.txt with following structure
MSG|20010102|123 125 4562 409|SEND
MSG|20010102|120 230|SEND
MSG|20010102|120 204 5071|SEND
MSG|20010103|2 11 1098 9810|SEND
......

index file index.txt is
11
201
298
100
......

What I want to do is:
1) extract the record in index.txt
2) search it in data files $3, if this colomn contains the index number, then delete this record, else save without change

I have many questions:
1)Since each time I extract a record in index file, search it and delete records in data files, I have to save to original files for next index number. how can I save the file?
2) how to delete records using awk

i think it needs shell and awk scripts,
I try by myself, but i could not do any further.
while read record
do
awk ......
down < index.txt

thanks a lot.