separate

[ovince]

hi All

what is the easyest way (if posible with one command line) to separate large files into subfiles starting form one header to another. For example,

#head 1
23434
23434
34324
#head 2
345345
#head 3
34543
34545
34545
34545

to separate into

#head
23434
23434
34324

#head
345345

and

#head
34543
34545
34545
34545

respectively.

Number of rows betwwen headers are not equal. Headers have same name

thanks
oliver

 

[feherke]

Hi

awk '$1=="#head"{if(NR>1)close(f);f=$2".txt"}{print$0>f}' /input/file

Feherke.

http://rootshell.be/~feherke/

 

[p5wizard]

csplit can do that as well, but first you need to find out how many "#head"s and how many digits to use for the individual file numbers...

heads=$(grep -c "^#head" /path/to/file)

((heads=heads-1))

digits=$(expr length $heads)

csplit -n $digits -f head /path/to/file "/^#head/" "{$heads}"

HTH,

p5wizard

 

[ovince]

thank you all for reply. Feherke has a very elegant solution with awk and it works. Thanks Feherke (koszonom)