splitcsv

gistfile1.sh

#!/bin/bash

if [ $# -lt 1 ]; then
  echo -e "Usage:"
  echo -e "$0 <csvfile>"
  exit
fi

total=$(wc -l $1|awk '{ print $1 }')
inc=4999
index=1
ctr=1
filename=$(echo $1|cut -d. -f1)
ext=$(echo $1|cut -d. -f2)
header=$(head -2 $1)
fio=">"

while [ $index -lt $total ]; do
  trange=$((index + inc))
  output="${filename}_${ctr}.${ext}"

  if [ $ctr -gt 1 ]; then
    #we need the headers for each csv
    echo "$header" > $output
    fio=">>"
  fi

  if [ $trange -lt $total ]; then
    cmd="sed -n '$index,${trange}p' $1 $fio $output"
  else
    remainder=$((inc - (trange - total)))
    cmd="sed -n '$index,$((index + remainder))p' $1 $fio $output"
  fi
  echo $cmd
  eval $cmd
  index=$((index + inc + 1))
  ctr=$((ctr+1))
done