Shell Programming and Scripting

Python alternative:

Code:

#!/usr/bin/python
#name:test.py
import glob
for file in glob.glob("*.trf"):
     r = open(file)
     print header = r.readline().strip()
     r.close()

usage:

Code:

./test.py > outputfile

#!/usr/local/bin/bash ;# ksh 2nd choice if bash not installed.
HDRSTR="unique string in every header"
DIR=/path/to/data/dir
OFL=/path/to/output/file
TMPFL=/tmp/zzz
cat /dev/null > $OFL
cd $DIR
for i in ls
do
cat $i >> $TMPFL
done
cat $TMPFL|sort -u > $OFL
CHK=`cat -n $OFL|grep $HDRSTR|cut -f1 -d' '`
TTL=`cat $OFL|wc -l`
cat $OFL|head -n $(($CHK-1)) > $OFL
cat $OFL|tail -n $(($TTL-$CHK)) >>$OFL

Okay, this hasn't been tested. It looks right, but the -1 on the penultimate line may not be needed... Hmmm.

Have fun.

Code:

sed -n "1!p" file2006* > newfile

awk '{if(fn != FILENAME)NR=1; if(NR != 1 )print $0;fn=FILENAME} ' filename*

Note that any of these solutions that use shell-globbing instead of find will break on large numbers of files... I used find since I figured there might be thousands of those autogenerated files.