xkcd Ripper

#!/bin/bash

# xkcd Ripper
# <http://xkcd.com/>
#
# Arkanon <arkanon@lsd.org.br>
# 2011/02/12 (Sáb) 01:25:50 (BRD)
# 2010/10/13 (Qua) 00:08:56 (BRS)
# 2010/10/12 (Ter) 23:43:14 (BRS)
#
# The ace in the hole :-p
# <http://recantodasletras.uol.com.br/artigos/69424>
# <http://denilsodelima.blogspot.com/2010/11/como-e-que-se-diz-o-pulo-do-gato-em.html>
#
# echo -e \\x$(printf %x 39)
# output|'
#
# eval "echo -e \"$(echo "&#39;Petit&#39;\nbeing a reference" | sed -r 's:&#([0-9]+);:$(echo -e \\\\x$(printf %x \1)):g')\""
# output|'Petit'
#       |being a reference
#
# lynx -dump -force_html <(echo '&quot;')
# output|   "
#       |

if [ $# != 1 ]
then
  echo "Usage: $(basename $0) <issue-number>"
  exit
fi

 out=$(wget -qO- http://xkcd.com/$1) || { echo "not found"; exit; }

data=$(eval "echo -e \"$(
  echo "$out" \
   | grep -e "<h1>" -e "/comics/.*title" \
   | sed -r '
              s:\\:_BS_:g
              s: *<h1>([^<]*)</h1>.*:\1:g
              s:.*/([^/]*)":\1:
              s:" title=":\n:
              s:".*::
            # here goes the line below
              s:_BS_:\\\\\\\\:g
            '
)\"")

t1=$(echo "$data" | sed '1!d')
i1=$(echo "$data" | sed '2!d')
l1=$(echo "$data" | sed '3!d')

# This command (l2) could be [merged with the sed above] OR [deleted if all we wanted was the next label (l3)]
l2=$(eval "echo -e \"$(
  echo "$l1" \
  | sed -r '
            s:\\:_BS_:g
            s:&#([0-9]+);:$(echo -e \\\\x$(printf %x \1)):g # put just this line in sed above
            s:_BS_:\\\\\\\\:g
          '
)\"")

l3=$(lynx -dump -force_html -width=1024 <(echo $l2) | sed -r 's/(^ +| +$)//')

echo -e "\n\e[1;33mTitle\e[0m"
echo    "$t1"

echo -e "\n\e[1;33mImage filename\e[0m"
echo    "$i1"

echo -e "\n\e[1;33mInside the img title attribute\e[0m"
echo    "$l2"

echo -e "\n\e[1;33mOutput\e[0m"
echo    "$l3"


if [ $1 = 859 ]
then

  echo -e "\n\e[1;32mFor reference, that's the correct answer\e[0m"

  echo -e "\n\e[1;33mTitle\e[0m"
  echo    "("

  echo -e "\n\e[1;33mImage filename\e[0m"
  echo    "(.png"

  echo -e "\n\e[1;33mInside the img title attribute\e[0m"
  echo    "Brains aside, I wonder how many poorly-written xkcd.com-parsing scripts will break on this title (or \\\\;;&quot;\''{\&lt;&lt;[' this mouseover text.&quot;"

  echo -e "\n\e[1;33mOutput\e[0m"
  echo    "Brains aside, I wonder how many poorly-written xkcd.com-parsing scripts will break on this title (or \\\\;;\"\''{\<<[' this mouseover text.\""

fi


echo

# EOF