HTTP stats - code, download time, number of assets and pageweight

http_stats.sh

#!/bin/bash
#A little shell script which uses wget to download
#all the requisites for a page, and returns the amount
#of time, the 'weight' of the page in K and the number
#of files, along with the HTTP code of the original request

#requires curl, bc, wget (version 1.13), awk and bash.

#expects the url (including http://) to be passed from STDIN

#Disclaimer - this was hacked together to scratch our itch
#there are plenty of better ways of achieving this!

#some config options
url=$1
dumpfolder="/tmp/http-stats"
mkdir -p $dumpfolder
tidy_url=`echo $url | cut -d "/" -f 3`
stats_file=$dumpfolder/"$tidy_url-stats"

#get the HTTP response code for the page
http_code=`curl -sL -w "%{http_code}" $url -o /dev/null`
if [ $http_code -ne 200 ]
then
	#if it's not 200, don't do the rest
	echo "code:$http_code files: size: time:"
	exit 1
else

	#get the other stats from wget
	/usr/local/bin/wget -P $dumpfolder -t 1 -p -nv $url &> $stats_file
	time_line=`grep "Total wall clock time" $stats_file `
	stats_line=`tail -1 $stats_file `
	download_time=`echo "$time_line" | cut -d ":" -f 2 | sed 's/ //g' | sed 's/s//g'`
	num_files=`echo $stats_line | awk '{print $2}'`
	size=`echo $stats_line | awk '{print $4}'`
	size_unit=${size: -1}
	size_value=${size:0:${#size}-1}
	case "$size_unit" in
		"B" )
			file_size=`echo "$size_value / 1024" | bc`
			;;
		"K" )
			file_size=$size_value
			;;
		"M" )
			file_size=`echo "$size_value * 1024" | bc`
			;;
		"G" )
			file_size=`echo "$size_value * 1024 * 1024" | bc`
			;;
	esac
	echo "code:$http_code files:$num_files size:$file_size time:$download_time"
	rm -rf $dumpfolder
fi