split.sh

## $FILES is names of all files to be split
## splits files into two


# find largest file
largest=`ls $FILES -l | sort -k5 -nr | head -n1 | awk '{ print $9 }'`

# count how many lines in it
lines=`wc -l $largest`


# find entity at middle
split_at=$((lines / 2))
# get nth line and find its entity (first field)
split_ent=`cat $largest | tail -n $split_at | head -n1 | awk -F '|' '{ print $1 }'`

# go through all the files (including the largest one..)
for file in $FILES
do
  # find total number of lines in the file
  total_lines=`wc -l $file`
  # and the line to split at, which is the first line of split entity
  split_line=`grep -n "^$split_ent|" $file | head -n1 | awk -F : '{ print $1 }'`
  
  # find all lines before split entity
  head -n $((split_line)) $file > $file.1
  # all lines after split entity
  tail -n $((total_lines-split_line)) $file > $file.2
done