#!/bin/bash

# Variáveis Globais

ERR_LOGFILE_NOT_FOUND=128

# Funções
# 
# filter -> recebe como argumento 
#

filter()
{
	LOGTOBEFILTERED=$1
	LOGALL="$TMPDIR/logall.log"
	NUMBEREDLOG="$TMPDIR/peerlogfullnumbered.log"
	RESULTNOT="$TMPDIR/resultnot.log"
	RESULTUP="$TMPDIR/resultup.log"
	RESULTDOWN="$TMPDIR/resultdown.log"
	RESULTNEED="$TMPDIR/resultneed.log"
	RESULTPEERUP="$TMPDIR/resultpeerup.log"
	RESULTPEERDOWN="$TMPDIR/resultpeerdown.log"
	RESULTCOMMGUMIN="$TMPDIR/resultcommunitygum_in.log"
	RESULTCOMMGUMOUT="$TMPDIR/resultcommunitygum_out.log"
	RESULT="$TMPDIR/result.log"
	FILTEREDLOG="$TMPDIR/filtered.log"

	# Adicionando número de linhas ao arquivo
	cat -n $LOGTOBEFILTERED > $NUMBEREDLOG

	# Extraindo eventos relevantes do arquivo 

	cat $NUMBEREDLOG | grep not\ connect | grep UAServerConnector.getUserAgentServer | awk '{print $1" "$2" "$3" "$13" not connected" }' > $RESULTNOT
  
	cat $NUMBEREDLOG | grep set\ up | awk '{print $1" "$2" "$3" "$8" up"}' > $RESULTUP

	cat $NUMBEREDLOG | grep set\ down | awk '{print $1" "$2" "$3" "$8" down"}' > $RESULTDOWN

   	cat $NUMBEREDLOG | grep CommunityObtainer.wannaGuMs | grep Local\ Peer\ needs | awk '{print $1" "$2" "$3" Peer needs "$10}' > $RESULTNEED

 	cat $NUMBEREDLOG | grep ourgrid.peer.main.Main.main | grep Starting\ doctor\ engine | awk '{print $1" "$2" "$3" Peer started!"}' >> $RESULTPEERUP

   	cat $NUMBEREDLOG | grep ourgrid.peer.main.Main.main | grep Starting\ doctor\ engine | awk '{print $1}' >> $RESULTPEERDOWN
	
	cat $NUMBEREDLOG | grep org.ourgrid.peer.communityobtainer.CommunityObtainer.hereIsOurGuM | awk '{print $1" "$2" "$3" "$11" Community GuM IN"}' >> $RESULTCOMMGUMIN
	
	cat $NUMBEREDLOG | grep org.ourgrid.peer.communityobtainer.CommunityObtainerFacade.disposeOf | awk '{print $1" "$2" "$3" "$9" Community GuM OUT"}' >> $RESULTCOMMGUMOUT

	for i in $( cat $RESULTPEERDOWN ); do
		POS=`expr $i - 1`
		if [ "$POS" != "0" ]; then
			cat $NUMBEREDLOG | sed -n ''$POS'p' | awk '{print $1" "$2" "$3" Peer down!"}' >> $LOGALL
		fi
	done
													  
	sort -m $RESULTDOWN $RESULTUP $RESULTNOT $RESULTNEED $RESULTPEERUP $RESULTCOMMGUMIN $RESULTCOMMGUMOUT >> $LOGALL

	sort -n $LOGALL > $RESULT

	cat $RESULT | awk '{print $2" "$3" "$4" "$5" "$6" "$7}' > $FILTEREDLOG

	tail -n 1 $NUMBEREDLOG | awk '{print $2" "$3" Peer Down!"}' >> $FILTEREDLOG

}



find_log_files()
{

	SRCDIR=$1
	PEERFILENAME="Peer*.log"
	LOGFILES="$TMPDIR/peerlogs.log"
	
	# Localiza todos os arquivos - a partir de SRCDIR - cujo nome
	# começa com "Peer" e terminam com ".log"

	find $SRCDIR -name $PEERFILENAME > $LOGFILES
	
	# testa se o arquivo existe e é vazio
	if [ ! -s $LOGFILES ]; then
		echo "$(basename $0): Peer log files don't exist in '$SRCDIR' directory"
		exit $ERR_LOGFILES_NOT_FOUND
	else
		echo "Using the following files..."
		cat $LOGFILES
	fi

}


create_full_log()
{
	LOGFILES=$1
	FULLLOG="$TMPDIR/peerfulllog.log"
	
	for i in $( cat $LOGFILES ); do
		cat $i >> $FULLLOG
	done

}



generate_events_by_entities()
{
	FILTEREDLOG=$1
	WORKLOADDIR="$TMPDIR/workload"


	mkdir -p $WORKLOADDIR

	while read line
	do
		id=$(echo $line | awk '{print $3}')
		echo $line >> $WORKLOADDIR/$id.log
	done < $FILTEREDLOG
 
}


workload()
{
	STATISTICSFILE="$TMPDIR/peerstatistics.log"
	PEERLOGFILE=Peer.log
	WORKLOADDIR="$TMPDIR/workload"

	mkdir -p $WORKLOADDIR

	echo "Starting workload"

	echo -n "Generating Peer statistics..."
	cat $WORKLOADDIR/$PEERLOGFILE > $STATISTICSFILE
	echo "done"
	
	echo "Generating GuMs statistics..."
	for i in $(ls $WORKLOADDIR | grep -v $PEERLOGFILE)
	do
		echo -n "Using $i..."
		cat $WORKLOADDIR/$i >> $STATISTICSFILE
		echo "done"
	done
	echo
	echo "done"
}

remove_not_connected_events()
{
	tmpfile=$1

	cat $1 | grep -v "not connected"
}	

create_event_files()
{

	FILE=$1

	INSTANTDIR="$TMPDIR/instant"
	mkdir -p $INSTANTDIR

        while read line
	do
		id=$(echo $line  | awk -F '[' '{print $2}' | awk -F':' '{print $1}')
		id=${id//\//-}
		echo $line >> "$INSTANTDIR/$id.log"
	done < $FILE

}

# Main 
#
# Este script é invocado com 1 argumento opcional
#

clear

red='\E[31;47m'
green='\E[32;47m'

nargs=$#
TMPDIR="/tmp/peer-$$"

# Cria um diretório temporário onde os logs serão armazenados
mkdir -p $TMPDIR


# Testa o número de argumentos e descobre a partir de onde vai
# procurar os arquivos de log. 

if [ ${nargs} -eq 0 ]
then
	SRCDIR=.
else
	if [ ${nargs} -eq 1 ]
	then
		SRCDIR=$1
	else
		usage
	fi
fi

FINALFILE="finalstatistics.log"
THISLOGFILE="workload.log"

echo "[`date +%x" "%T`] Step 1 -> Initializing Workload... $SRCDIR"
echo "[`date +%x" "%T`] Step 1 -> Initializing Workload... $SRCDIR" >> $THISLOGFILE
# Seta o valor da variavel $LOGFILES
find_log_files $SRCDIR

echo "[`date +%x" "%T`] Step 2 -> Creating a single log file to be processed... $SRCDIR"
echo "[`date +%x" "%T`] Step 2 -> Creating a single log file to be processed... $SRCDIR"  >> $THISLOGFILE
# Cria um unico arquivo de log concatenado todos os logs listados
# em LOGFILES
create_full_log $LOGFILES

echo "[`date +%x" "%T`] Step 3 -> Filtering the log file... $SRCDIR"
echo "[`date +%x" "%T`] Step 3 -> Filtering the log file... $SRCDIR"  >> $THISLOGFILE
# Filtra o arquivo gerado acima ($FULLLOG), gerando um arquivo
# menor ($FILTEREDLOG)
filter $FULLLOG


# Gera um arquivo por maquina em '$WORKLOAD' dir contendo todos os eventos de cada
# uma.
echo "[`date +%x" "%T`] Step 4 -> Generating Events by Entities... $SRCDIR"
echo "[`date +%x" "%T`] Step 4 -> Generating Events by Entities... $SRCDIR"  >> $THISLOGFILE
generate_events_by_entities $FILTEREDLOG


# Gera o workload do peer e mostra na tela

echo "[`date +%x" "%T`] Step 5 -> Generating peer workload... $SRCDIR"
echo "[`date +%x" "%T`] Step 5 -> Generating peer workload... $SRCDIR"  >> $THISLOGFILE
workload 

echo "[`date +%x" "%T`] Step 6 -> Removing not connected events $SRCDIR"
echo "[`date +%x" "%T`] Step 6 -> Removing not connected events $SRCDIR"  >> $THISLOGFILE
remove_not_connected_events $STATISTICSFILE > $FINALFILE


echo "[`date +%x" "%T`] Step 7 -> Create snapshot files $SRCDIR"
echo "[`date +%x" "%T`] Step 7 -> Create snapshot files $SRCDIR" >> $THISLOGFILE
create_event_files $FINALFILE

# remove todos os arquivos temporários gerados para o
# processamento dos logs.
#rm -rf $TMPDIR

echo "[`date +%x" "%T`] Finished $SRCDIR"
echo "[`date +%x" "%T`] Finished $SRCDIR"  >> $THISLOGFILE