asked on

Splitting a long field into multiple shorter fields.

On a Linux machine, I need to split product description fields that are up to 256 characters long, into multiple fields no longer than 76 characters. The catch is, I need to split it up at a space. So the multiple fields need to be as long as possible but split at a space and no longer than 76 characters. I am assuming awk or sed can do this but not sure where to start.

Any help would be appreciated.

Bill Prew

Can you provide a sample data file for testing / clarification? Is the product description the only thing on each line?

Do you want each "chunk" less than 76 characters output on a separate line?

»bp

Kent Olsen

Hi Tee,

What do you need to do with the items once they're split?

awk is probably the right tool. It uses a space for the default separator. The only question is once split, then what?

arnold

Perl is the way to go, identify all space positions.
Then cut the Sata accordingly.

Bill Prew

Could probably be optimized slightly, but here's a basic AWK script that seems to get the job done.

BEGIN {
    # set max length of output lines
    maxLen = 76

    # initialize work variables for output line
    outLine = ""
    outLen = 0
}

{
    # loop through all space delimited fields in this input line
    for (i=1; i<=NF; i++) {
        # get lenth of this chunk
        l = length($i)

        # will this exceen max output line length?
        if (outLen + l + 1 > maxLen) {
            # print accumulated output line, and clear it
            print outLine
            outLine = ""
        }

        # if first chunck added to output line, no space seperator added
        if (outLine == "") {
            outLine = $i
            outLen = l
        } else {
            outLine = outLine " " $i
            outLen = outLen + l + 1
        }
    }

    # print any pending output line that was built
    if (outLine != "") {
        print outLine
    }

    # initialize work variables for output line
    outLine = ""
    outLen = 0
}

Open in new window

EDITED: Added comments, and fixed output length calculation.

»bp

9thTee

ASKER

Hi Bill,
This does exactly what I asked. But I need a small change, for the output, I would like the newly created 76 maximum character fields to all be on one line and pipe "|" delimited. Is that possible?

abc….76 characters max...xyz|abc….76 characters max...xyz|abc….76 characters max...xyz

Thanks,
Mark

ASKER CERTIFIED SOLUTION

Bill Prew

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

9thTee

ASKER

Perfect, thanks for your help.

Bill Prew

Welcome.

»bp