# Assignment 1
# URL
# by Steven Lehrburger
# lehrburger (at) gmail (dot) com
# NYU Interactive Telecommunications Program, Spring 2009
# Programming A to Z with Adam Parrish
 
rm output.txt                                           # this will be our output file, so get rid of it first           
grep -E "  "  <sonnets.txt | cut -c 3- >couplets.txt    # get only the lines with two spaces at the beginning,
                                                        # i.e. the couplets, remove the spaces, and put them in
                                                        # a new file

linecount=$(wc -l couplets.txt | tr -s " " | cut -d " " -f 2)
                                                        # count how many lines of couplets there are, remove extra
                                                        # spaces, and then get only the number and not the filename
while [ $linecount -gt 0 ]                              # loop while linecount is greater than zero
do
  head -n 1 <couplets.txt >>firstlines.txt              # get the first line of text, and append it to one file
  head -n 2 <couplets.txt | tail -n 1 >>secondlines.txt # get the other line of text, and append it to another file
  tail -n +3 <couplets.txt > couplets_temp.txt          # store all of the original file but those two lines
                                                        # in a temporary file (I don't know why this is needed?)
  mv couplets_temp.txt couplets.txt                     # and then overwrite the original with the temp file
  linecount=$(( $linecount - 2 ))                       # decrement the line count by two
done
rm couplets.txt                                         # and delete the source file - we don't need it any more

coupletcount=$(wc -l firstlines.txt | tr -s " " | cut -d " " -f 2)
                                                        # do the same thing for the number of lines in one file
                                                        # (there will be the same number in the second file)
while [ $coupletcount -gt 1 ]                           # loop, but only go to 1 and not zero - if there is an odd
                                                        # number then there is an unmatched couplet, just trash it
do
  head -n 1 <firstlines.txt >>output.txt                # put the first line from one file in the ouput
  head -n 2 <firstlines.txt | tail -n 1 >>output.txt    # and the second line from that file
  tail -n +3 <firstlines.txt >firstlines_temp.txt       # update the original file to remove those lines, as above
  mv firstlines_temp.txt firstlines.txt

  head -n 1 <secondlines.txt >>output.txt               # and then do the same thing with their pairs from the other
  head -n 2 <secondlines.txt | tail -n 1 >>output.txt   # file
  tail -n +3 <secondlines.txt >secondlines_temp.txt
  mv secondlines_temp.txt secondlines.txt

  coupletcount=$(( $coupletcount - 1 ))                 # decrement the counter
done
rm firstlines.txt                                       # and get rid of those two intermediate files
rm secondlines.txt