#!/bin/sh
# A very simple shell script to extract essential data from Google SoC
#  Application list pages
#  M. Connor Fri May 12 03:05:08 GMT 2006
#  Outputs name, email, project_title, score, has_mentor?

# First we Delete blank lines
sed -e "/^$/d" |

while read xx
do
   lineclass=`echo $xx | sed -e "s@^.*class=@@" -e "s@\"@@g" -e "s@\>.*@@"`
#  echo "lineclass: $lineclass"
   case "$lineclass" in
        "listrequestapp" )
            read xx
            read xx
            email=`echo $xx | sed -e "s@^.*csaid=@@" -e "s@\:.*@@"`
            read xx
            name=`echo $xx | sed -e "s@\:.*@@"`
            read xx
            projtitle=`echo $xx | sed -e "s@^\<b\>\<i\>@@" -e "s@</i></b>@@"`
            read xx
            read xx
            read xx
            read xx
            read score
            printf "%s\t%s\t%s\t%s\t1\n" "$name" "$email" "$projtitle" "$score"
            ;;
        "listapp" )
            read xx
            read xx
            email=`echo $xx | sed -e "s@^.*csaid=@@" -e "s@\:.*@@"`
            read xx
            name=`echo $xx | sed -e "s@\:.*@@"`
            read xx
            projtitle=`echo $xx | sed -e "s@^\<b\>\<i\>@@" -e "s@</i></b>@@"`
            read xx
            read xx
            read xx
            read score
            printf "%s\t%s\t%s\t%s\t0\n" "$name" "$email" "$projtitle" "$score"
            ;;
        esac
done
exit 0