SED script for searching text in a file and copying to new file

[gessi2000]

Hello,

I am no SED Guru and therefore I need short help for writing a SED script which searches for a RegExp in a file and copyies from that RegExp to another RegExp content to a new file like:

$sed -n -e /RegExp1/,/RegExp2/p mytextfile

So now I get the content from RegExp1 to RegExp2 to Stdout which is OK.

But I don't want to have the lines with RegExp1 and RegExp2 within the output but to start 1 line after RegExp1 to copy and stop 1 line before RegExp2.
Can anyone tell me what RegExp to use so I get lines RegExp1+1(line) to RegExp2-1(line) to the output?

[tpl]

sed '1d;$d' <mytextfile

[sathiya]

sed -n '/regexp1/,/regexp2/w anotherfile' filename

Refer here for more examples: Unix Sed Tutorial: How To Write to a File Using Sed

[drl]

Hi.

...Can anyone tell me what RegExp to use so I get lines RegExp1+1(line) to RegExp2-1(line) to the output?

I would say that it is not the regular expression that can do this, but rather how the RE is combined and used in a command. Here are a few ways:

#!/usr/bin/env bash

# @(#) s1	Demonstrate range extraction, ed, sed, cgrep, perl.

# Infrastructure details, environment, commands for forum posts. 
set +o nounset
LC_ALL=C ; LANG=C ; export LC_ALL LANG
echo ; echo "Environment: LC_ALL = $LC_ALL, LANG = $LANG"
echo "(Versions displayed with local utility \"version\")"
c=$( ps | grep $$ | awk '{print $NF}' )
version >/dev/null 2>&1 && s=$(_eat $0 $1) || s=""
[ "$c" = "$s" ] && p="$s" || p="$c"
version >/dev/null 2>&1 && version "=o" $p ed sed cgrep perl
set -o nounset
echo

FILE=${1-data1}

echo " Data file $FILE:"
cat $FILE

line1="qux"
line2="grault"

rm -f t1
ed $FILE 2>/dev/null <<EOF 
/$line1/+1,/$line2/-1w t1
q
EOF

echo
echo " Extracted data from ${line1}+1 to ${line2}-1 with ed:"
cat t1

rm -f t1
echo
echo " Results with sed:"
# See http://www.faqs.org/faqs/editor-faq/sed/ section 4.15
# ( "b" means "branch" )
sed -n "/${line1}/,/${line2}/{
/${line1}/b
/${line2}/b
p
}" $FILE >t1
cat t1

rm -f t1
echo
echo " Results using 2 sed commands:"
# First command could also easily be the corresponding awk command.
sed -n "/${line1}/,/${line2}/p" $FILE |
sed '1d;$d'

echo
echo " Windowing copy with cgrep (expecting trailing line):"
cgrep -D -I2 -w "${line1}" "${line2}" $FILE

echo
echo " Results with perl, print within a range:"
perl ./p1 "${line1}" "${line2}" $FILE

exit 0

producing:

% ./s1

Environment: LC_ALL = C, LANG = C
(Versions displayed with local utility "version")
OS, ker|rel, machine: Linux, 2.6.26-2-amd64, x86_64
Distribution        : Debian GNU/Linux 5.0 
GNU bash 3.2.39
GNU Ed 0.7
GNU sed version 4.1.5
cgrep - (local: ~/executable/cgrep May 29 2009 )
perl 5.10.0

 Data file data1:
foo
bar
baz
qux
quux
corge
grault
garble
warg
fred
plugh
xyzzy
thud

 Extracted data from qux+1 to grault-1 with ed:
quux
corge

 Results with sed:
quux
corge

 Results using 2 sed commands:
quux
corge

 Windowing copy with cgrep (expecting trailing line):
quux
corge
grault

 Results with perl, print within a range:
quux
corge

What we might learn from this is:
1) it is best to supply a sample of the data, along with expected output. That avoids each responder needing to create a data file and provides a common reference.

2) tpl's solution is clever, I had not thought about running 2 seds (assuming that is what is being implied).

3) I could not coerce my often-favorite, cgrep, into doing the right thing with only one call.

4) the perl code is not shown, but could be provided if needed.

5) see the URL noted in the script for the sed FAQ that provides many solutions to common problems.

6) many people find awk easier to use and easier to read compared to sed.

Best wishes ... cheers, drl