google-rank

#!/bin/bash
# Simple munin plugin to find the google rank for a URL/WORD combination
#
# THIS SCRIPT BREAKS THE TOS OF GOOGLE SO USE WITH CARE AND DON'T BLAME ME IF THINGS GO WRONG
#
# (c) 2009 i.dobson@planet-ian.com
#
# For each url/words that you want to watch you need to create a variable/word pair in your
# munin-node configuration file for example
#
#[google_rank]
#user root
#timeout 60
#env.URL1 http://www.plant-ian.com
#env.WORD1 avr webcam
#env.URL2 http://www.plant-ian.com
#env.WORD2 bascom
#
# Version 0.5 24.1.2009
# Added loop to check the first 500 pages. Note the script sleeps 5 seconds between each page grab so
# If the word/url your looking for is in the higher positions then you need to increase the timeout
#
# Version 0.5 21.1.2009
# Dump each page grabbed from google into separate files (helps with debugging)
#
# Version 0.4 19.1.2009
# Fixed corrupt then empty cache file bug
#
# Version 0.3 19.1.2009
#  The script now grabs the google page based on the LASTHIT counter.
#  The script grabs the google page for URL1, then the next time it's called URL2 etc. If the url/word pair doesn't exist for LASTHIT then the script just dumps the cached data
#
# Version 0.2 18.01.2009
#  Cache added, the script only grabs the pages from google every 10 calls
#  The script still only checks to first 100 pages returned by google
#
# Version 0.1 17.01.2009 Initial release
#  The script only checks to first 100 pages returned by google
#

# Auto Configure, Check it word 1 is defined
if [ "$1" = "autoconf" ]; then
   if [ -n "$URL1" ] && [ -n "$WORD1" ]; then
      echo yes
   else
      echo no
   fi
   exit 0
fi

#Configure, loop through each variable defined WORDx URLx dumping it to munin
if [ "$1" = "config" ]; then
   iLoop=1
   echo 'graph_title Google page rank'
   echo 'graph_args --upper-limit 100 -l 0'
   echo 'graph_category search'
   echo 'graph_scale no'
   echo 'graph_info Google page rank for URLs & Words'

   URL="xxx"
   until [  "$URL" = "" ]; do
     TMPURL=URL$iLoop
     URL="${!TMPURL}"
     TMPWORD=WORD$iLoop
     WORD="${!TMPWORD}"
     if [ "$URL" = "" ]; then
       exit 0
     fi
     if [ "$WORD" = "" ]; then
       exit 0
     fi
     VAR=`echo $URL.$WORD | sed -e "s/http:\/\///g"| sed -e "s/ /_/g"| sed -e "s/\./_/g"| sed -e "s/\-/_/g"`
     URL=`echo $URL| sed -e "s/http:\/\///g"`
     echo $VAR.label Pagerank $URL - $WORD
     let iLoop="$iLoop +1"
   done
   exit 0
fi

#Meat of the program, grabs data from google for one word/url pair using LASTHIT as the pointer to which url/word pair to read

#Read update & save counter
LASTHIT=0
if [ -f /tmp/google_rank.status ]; then
  LASTHIT=`cat /tmp/google_rank.status | awk '{print $1}'`
fi

let LASTHIT="$LASTHIT + 1"
echo $LASTHIT > /tmp/google_rank.status

#Find URL/WORD PAIR for loop counter
TMPURL=URL$LASTHIT
URL="${!TMPURL}"
TMPWORD=WORD$LASTHIT
WORD="${!TMPWORD}"

if [ "$URL" != "" ]; then

#Setup defaults
  base=0
  num=1
  start=0
  FOUND=0
#Clean up URL/WORD pair, removing http:// replacing " " with "_", "." with "_", "-" with "-"
  VAR=`echo $URL.$WORD | sed -e "s/http:\/\///g"| sed -e "s/ /_/g"| sed -e "s/\./_/g"| sed -e "s/\-/_/g"`
  SEARCHWORD=`echo $WORD| sed -e "s/ /%20/g"`

until [ "$FOUND" -ne "0" ]; do
#Grab page from google for the WORD/PAGE combination.Pipe it into awk to pull out the url's only, one per line. Then dump only the lines containing the URL defined
    wget -q --user-agent=Firefox -O - http://www.google.com/search?q=$SEARCHWORD\&num=100\&hl=en\&safe=off\&pwst=1\&start=$start\&sa=N > /tmp/google_rank.$LASTHIT.data
    VALUE=`cat /tmp/google_rank.$LASTHIT.data|sed 's/<a href=\"\([^\"]*\)\" class=l>/\n\1\n/g'|awk -v num=$num -v base=$base '{ if ( $1 ~ /^http/ ) print base,num++,$NF }'|awk '{ print $2 "  " $3}'|grep -i $URL| awk '{ print $1}'`
    VALUE=`echo $VALUE| awk '{ print $1}'`
    if [ "$VALUE" = "" ]; then
      VALUE=-1
      let start="start + 100"
      sleep 5
    else
      FOUND=1
      let VALUE="$VALUE + $start"
    fi
###    echo Start=$start Value=$VALUE Found=$FOUND
    if [ "$start" -gt 500 ];then
      FOUND=-1
      VALUE=-1
    fi
done

#Read through cache file saving to array
  iLoop=1
  while read line ;do
    Data[$iLoop]=$line
    let iLoop="$iLoop +1"
  done < /tmp/google_rank.cache

#replace one line with the new value grabbed from google
  Data[$LASTHIT]="$VAR.value $VALUE"

#write data back
  rm /tmp/google_rank.cache
  for iLoop in `seq 1 10`; do
    echo ${Data[$iLoop]} >> /tmp/google_rank.cache
  done
fi

#Reset counter to start
  if [ "$LASTHIT" -gt 30 ]; then
     echo 0 > /tmp/google_rank.status
  fi

#Dump data to munin
  while read line ;do
    if [ "$line" != "" ]; then
       echo $line
    fi
  done < /tmp/google_rank.cache
exit 0