Shell script to create trimmed versions of a CSV. Trims the file to the number of lines supplied in the second input.

shellTrim.sh

#!/bin/bash
#
#AUTHOR: Pete Williamson
#DATE: 2014-10-21
#
#USAGE:
# Creates smaller versions of supplied CSV file,
# based on the line number values as the second parameter
#
# The resulting CSVs are zipped up (for windows compatibility)
# as a zip file based on the original name.
#
# The generated zip files are then deleted.
#
# The following will create foo.zip, containing 3 files
# - foo.csv (original file)
# - foo.5.csv (first 5 lines)
# - foo.50.csv (first 50 lines
#
# ./trim.sh foo.csv 5,50
#

# localize script params
input=$1
sizes=$2

# check input file is supplied
if [[ -z "$input" ]]
then
        echo 'No input file supplied'
        exit
fi

# check there is a valid csv of sizes
if ! [[ "$sizes" =~ ^[0-9]+(,[0-9]+)*$ ]]
then
        echo 'No line counts specified'
        exit
fi

# convert sizes csv into array
IFS=,
read -r -a arr_sizes <<< $sizes
unset IFS

# trim the file to each size specified
# save as new file, X = LINE COUNT > ORIGINAL.X.CSV
for cnt in $arr_sizes
do
        new_file=$(echo $input | sed -e s/.csv/.$cnt.csv/)
        sed -n '1,'$cnt'p' $input > $new_file
done

# build a zipped archive based on original name
archive=${input%.csv}

zip $archive *.csv

# remove csvs
rm *.csv