Inline "Sed" script

[Sean OB_1]

Howdy.

Years ago I had a script that essentially did inline sed on files.

With it you'd call the script and give the sed command, it would then run against the file(s) you specify and do the sed and save the file.

Does anyone have something like this that they'd be willing to post?

TIA.

Sean

[Sanjay_6]

Hi Sean,

Is this what you are looking for.

sed one-liner

http://www.student.northpark.edu/pemente/sed/sed1line.txt

Hope this helps.

Regds

[Sanjay_6]

Hi,

Here is something using awk,

http://www.student.northpark.edu/pemente/awk/awk1line.txt

http://www.sap-basis-abap.com/unix/awk-one-liner-tips.htm

Hope this helps.

Regds

[Rodney Hills]

"perl" can do that and more-

perl -p -i -e 's/this/that/g' allmyfiles/*

Will replace the text in all the files under directory allmyfiles.

Change option -i to -i.bu and it will save a backup of each file with an extenstion of .bu.

HTH

-- Rod Hills

[Victor Fridyev]

Try this

#!/usr/bin/sh
TMF=/tmp/$$
PRGN=$(basename $0)
if [ $# -ne 3 ]; then
echo "$PRGN:USAGE $PRGN FILENAME CURR CHANGE"
exit
fi
if [ -s $1 ];then
cat $1| sed "s?$2?$3?g" > $TMF
if ! diff $1 $TMF 1>/dev/null ;then
cp $TMF $1
fi
else
echo "$PRGN:File $1 NOT FOUND"
fi
rm $TMF

[Stuart Abramson]

##
# sed_files.ksh SDA 12/16/03
#
# Change a string in a file and recreate the file with:
#
# string1 -> string2
#
USAGE="Usage: sed_files.ksh [FILE]"
#
# Example:
# sed_files.ksh OLD NEW *.ksh
#

if (( $# < 3 ))
then
print $USAGE
exit 1
fi
OLD=$1
NEW=$2
shift 2
#
for F in $*
do
if [[ -f $F ]]
then
# We have a regular file
file $F | grep -q text
if (( $? == 0 ))
then
# We have a text file
grep -q $OLD $F
if (( $? == 0 ))
then
# We have a text file with "string1" in it
cp -p $F $F.old
sed "s/$OLD/$NEW/g" $F > $F.new
mv $F.new $F
rm $F.old
fi
fi
fi
done