c1r.bash

#! /bin/bash
 
# C1R compiler
 
# remove HTML tags, but replace "<br />" by newlines
function removeHTMLtags() {
 sed -e 's~<br */>~\
~g' | sed -e :a -e 's/<[^>]*>//g;/</N;//ba' 
}
 
# unescape HTML codes: replace "&lt;" by "<" etc
function unescapeHTML() {
 sed -e 's/&lt\;/</g;s/&gt\;/>/g;s/&nbsp\;/ /g;s/&#160\;/ /g;s/&#xA0\;/ /g;s/&quot\;/"/g;s/&#40\;/(/g;s/&#41\;/)/g;s/&#91\;/[/g;s/&#93\;/]/g;s/&#123\;/{/g;s/&#125\;/}/g'
}
 
FILENAME=$1
 
if [ -z ${FILENAME} ] 
then
  echo "Usage: $0 <fileName>"
  exit 1
fi
 
FILENAME1=${FILENAME}.c
ROSETTAURL=rosettacode.org/wiki
WORDCOUNT=`cat $FILENAME|wc -l`
 
cp ${FILENAME} ${FILENAME1}
 
if [ $WORDCOUNT -eq 1 ]
then
  # Note: the self-printing Quine program requires special treatment
  if [ `cat $FILENAME` = "Quine" ]
  then 
    cat << EOF > $FILENAME1
#include <stdio.h>
int main(char args[]) {printf("Quine\\n");}
EOF
  else 
    PAGEURL=$ROSETTAURL/`cat $FILENAME`
    curl $PAGEURL 2>/dev/null | grep -m 1 "<pre class=\"c highlighted_source\">" | removeHTMLtags | unescapeHTML >${FILENAME1}
  fi
fi
 
cc $FILENAME1