Unix shell / SED oneliner to delete lines

[pcdas]

I want an one-liner to delete the TOP 5 lines or BOTTOM 5 lines from a text file. An early reply will be highly appreciated.

 

[dstxaix]

Try the following command to delete the first file lines.

sed -n '1,5d' inputfile > outputfile

Try the below command to delete last 5 lines

sed -e :a -e '$d;N;2,5ba' -e 'P;D' inputfile > outputfile

 

[SamBones]

One liner in Korn Shell...

sed -n "6,$(($(( $(wc -l<inputfile) )) - 5))p" inputfile

A little more generalized...

# X is lines to remove from front and rear
X=5

sed -n "$((X+1)),$(($(( $(wc -l<inputfile) )) - $X))p" inputfile

 

[Annihilannic]

A simpler way to delete the first 5 lines:

tail +6 inputfile > outputfile

And a not quite as simple way to delete the last 5:

head -n $(( $(wc -l<inputfile) - 5)) inputfile > outputfile

Sam, did you mean to include the extra set of $(( ))?

Annihilannic.

 

[feherke]

Hi

My "I don't want to think right now" solution :

tail +6 /input/file | tac | tail +6 | tac > /output/file

Feherke.

http://rootshell.be/~feherke/

 

[pcdas]

My sincere thanks to Dstxaix, Sam & Anihilannic for giving me different solutions. I tried all the different solutions and solved my problem. However, I liked the general solution provided by Sam ! You guys are fantastic !!

 

[SamBones]

Hi Anni,

Yes, I did mean to include a second set of $(( )). It might not be needed, but I usually put that around anything that needs to be handled arithmetically.

Yup, you're right, this works fine...

# X is lines to remove from front and rear
X=5

sed -n "$((X+1)),$(( $(wc -l<inputfile) - $X))p" inputfile

 

[stefanwagner]

My nobrainer: $d deletes the last line - just repeat it 5 times. sed -i modifies the file in place:

for dummy in {1..5}; do sed -i '$d' file ; done

don't visit my homepage:

http://home.arcor.de/hirnstrom/bewerbung

 

[Annihilannic]

And just for fun, a generalised awk solution:

awk -v l=5 'NR>l{print a[NR%l]}{a[NR%l]=$0}' inputfile

Annihilannic.