root/misc/cronrss

#!/bin/bash
#
#                        Cronrss
#                   Old style RSS feed
#
# http://freaknet.org/alpt/src/utils/cronrss
# 
# It verifies if the web page pointed by a given URL has changed.
# If it has, cronrss will send a notice email to you.
# The URL list is kept in the ~/.cronrss/url text file (one URL per line).
#
# It is very useful to use URLs which point to rss feeds.
#
# *** Usage:
# 
# 1) Put all the URLs you want to check in the ~/.cronrss/url text file (one
#    URL per line).
# 
# 2) Run cronrss
#
#          cronrss [-d [seconds]]
#
# If the `-d' option is specified, it will check periodically the URLs. 
# You can specify its frequency in seconds (default is 1800s = 30min)
#
#          cronrss update
#
# Each time you modify the url_list, you have to run `cronrss update`
#
# *** Examples: 
#
#        echo http://www.hackaday.org >> ~/.cronrss/url
#        cronrss update
#        cronrss -d
#        # From this point on, whenever hackday.org changes you will get an email
#        # which shows you the text differences
#        # Since hackaday has a rss feed it is better to use this URL instead:
#        # http://www.hackaday.com/rss.xml
#
# *** Links:
#
#       This webservice maybe useful:
#               http://page2rss.com/
#
# PS: the idea of Cronrss was initially published on Idiki:
# http://idiki.dyne.org/wiki/Cronrss
#
#
# AlpT (@freaknet.org)
#


#
#
## Configure here
#

dir=`echo ~/.cronrss`   # cronrss stuff is saved in this directory
dump="$dir/dump"        # dumps of the web pages
url_list="$dir/url"     # url list file

mail_to="`whoami`"      # this is the email address where the notice is sent
#mail_to="foo@bar.org`"
mail_subject="Cronrss digest"  

#
##
#
#

check_it() {
        local notice

        if [ "$1" == "no_notice" ]
        then
                notice=0
        else
                notice=1
        fi

        if [ ! -d $dump ]
        then
                mkdir -p $dump/
        fi

        for i in `cat $url_list`
        do
                md5=`echo "$i" | md5sum | cut -d ' ' -f 1`

                rm -f $dump/old-$md5                    &> /dev/null
                mv -f $dump/new-$md5 $dump/old-$md5     &> /dev/null
                links -dump "$i" > $dump/new-$md5

                if [ "$?" != "0" ] || [ ! -f $dump/old-$md5 ] || [ ! -f $dump/new-$md5 ]
                then
                        continue
                fi

                diff -q $dump/old-$md5 $dump/new-$md5   &> /dev/null
                if [ "$?" == 1 ] && [ "$notice" == "1" ]
                then
                        local t
                        t=`tempfile`

                        echo "$i changed"                     > $t
                        echo ""                               >> $t
                        diff -u $dump/old-$md5 $dump/new-$md5 >> $t

                        cat $t | mail -s "$mail_subject" $mail_to

                        rm $t
                fi
        done
}

if test "$1" = "help" -o "$1" = "-h" -o "$1" = "--help"
then
        echo "Usage:"
        echo "        cronrss [-d [seconds]]"
        echo "        cronrss update"
        echo ""
        echo " If the \"-d\" option is specified, it will check periodically the URLs."
        echo " You can specify its frequency in seconds (default is 1800s = 30min)"
        echo ""
        echo " Each time you modify the url_list, you have to run \"cronrss update\""
        exit 1
fi

if [ "$1" == "-d" ]
then
        seconds=1800
        if [ ! -z "$2" ]
        then
                seconds=$2
        fi

        while true
        do
                check_it
                sleep $seconds
        done
elif [ "$1" == "update" ]
then
        
        check_it no_notice
else
        check_it
fi