File handling in Shell Script

[ashok rajagopal]

In my application, i have to create a shell script which takes in a data file and gets some data from it. Basically, i need to read n number of characters from each record in the file into a variable, something like 'fread' command in C/C++... which is the unix command or construct that i can use ?

I have to write some values to another file, and i want the command for writing a variable to the file also. like 'fprintf' in C/C++ . Any pointers ???

Want immediate help on this... Please...

[redhead]

Something like this:

#!/bin/sh                                                                       

if [ -z $1 ]; then
    echo "Usage: $0 <file> <n>"
    exit
fi
if [ -z $2 ]; then
    echo "Usage: $0 <file> <n>"
    exit
fi
if [ ! -f $1 ]; then
    echo "File: $1 isn't a file"
    exit
fi
for n in `grep -n $ $1 | awk -F ':' '{print $1}'`; do
    LINE=`head -$n $1|tail -1`
    # fetch the first n chars from line                                         
    CHARS=`echo $LINE | cut -c 0-$2`
    # do whatever with CHARS                                                    
    echo $CHARS
done

Remember you can form your echo statement any which way you want it to be displayed, and you can use awk/cut to pick anything out of your $LINE

[Belisarius]

The head/tail way is really the only way to do it with a conventional shell script. You can do shell-scripting with Perl and PHP, you might want consider using one of those if you're going to be parsing a file. I combine both shell-scripts and PHP scripts for my system-admin work.