Shell Programming and Scripting

Code:

sed -n '1d; $!{x;s/\(.\)/\1\
/p512 ;}' file

Hi Danmero and soleil,

Thanks for you response. I am getting this. But i need header record and trailer record in the samefile. my detail records are splitting but my header and trailer records are not showing.
I need my file as

header
summary1
detail1
summary2
detail2
summary3
detail3
trailer1
trailer2

Can you please let me know how to get my header and trailer record

Code:

host1> head -1 file1 > outfile
host1> sed '1d;$d' file1|sed '$d'|awk '{print substr($0, 1, 8); print substr($0, 9, 7)}' >> outfile
host1> tail -2 file1 >> outfile
host1> cat outfile
header
summary1
detail1
summary2
detail2
summary3
detail3
trailer1
trailer2

Hi soleil
sorry for asking many questions.
I am getting this message
awk: record `RW587241376ELVIA ...' too long

actually my file is like xxxx.mf file. This contain one big record with 32MB. If I open through textpad, entire file shows in 1 line as
headersummary1detail1summary2detail2summary3detail3trailer1trailer2.

If i open with notepad, file shows like
header
summary1detail1
summary2detail2
summary3detail3
trailer1
trailer2

when I open in unix box, it says too long. so user needs to cut the detailrecords.
Can you please suggest me.
Sorry for asking many questions. I am new to shell scripts.
Plz help me.

Try using 'gawk' instead of 'awk' if you have it available on your system and see if that fixes the problem or not. I think 'awk' has some limitation with record lengths.

It saying
ksh: gawk: not found

I think there is no gawk command in my system.
I dont know why my file is showing 1 line when I open with textpad. shows multiple lines when I open with notepad. It says line too long when I open in UNIX.
Can you please suggest me how to resolve this issue.

To keep the first line and the last two lines unchanged, while splitting any other lines at the 512th byte:

Code:

sed '1n; 2{h;d;}; ${H;g;q;}; $!{x;s/\(.\)/\1\
/512 ;}' file