#!/bin/sh

# Born Nov 2 1999

# Wed Nov  3 13:45:33 EDT 1999
# Fixed	lack of Host: hostname to make virtual hosts based on hostname work
#	(as in www.mozillazine.org)

# Thu Feb 17 00
# Can't use content-length because dynamic pages wouldn't allow for it
# We've got some IMG SRC parsing - alpha quality but ok
# Getting better (at 5am) and seems solid enough
# Damn yahoo uses SRC=filename with no quotes!
# And now, LINK processing - has to be redone into an awk function, but
# it's looking okay for now

# Wed May 17 00 - released v0.1
# Chucked HTTP 1.1 and added background processing

# Fri Jul 21 00 - released v0.2
# Integrated the awk script into this file (damned sprintfs for quotes added
# too) and added -v and -k processing. Preparing for the holy rewrite.

# Tue Aug 15 00 - released v0.31
# Fixed stupid bug with relative paths; possibly still broken (still
# checking) - looks okay to me

# Lacks CSS background-image processing
# Lacks FRAME processing (ugh)
# Will always lack: 
#	Javascript - die, monster.
# 	META REFRESH processing
# 	Location processing

httphead() {
	/bin/echo -n -e "GET /$2$1 HTTP/1.0\r\nHost: $hostname\r\nAccept: text/html, text/plain, audio/*, image/*, */*;q=0.01\r\nAccept-Encoding: gzip, compress\r\nAccept-Language: en\r\nUser-Agent: Ionax\r\n\r\n" ;
}

urlhost() {
	echo $1 | grep -qs //						# if we get SRC="aaa/aaa"
	if [ $? -eq 0 ] ; then						# then no host
		echo $1 | cut -d/ -f 3 | cut -d: -f1	# was specified
	else
		echo $hostname							# return default host
	fi
}

urlport() {
	echo $1 | grep -qs //
	if [ $? -eq 0 ] ; then 						# SRC="http://"
		echo $1 | cut -d/ -f 3 | grep -qs :
		if [ $? -eq 0 ] ; then					# SRC="http://xxx:yy/"
			echo $1 | cut -d/ -f 3 | cut -d: -f2
		else
			echo 80
		fi
	else
		echo 80
	fi
}

urldir() { # outputs "dir1/dir2/" from url
	tmp=$1
	echo $tmp | grep -qs //
	if [ $? -eq 0 ] ; then						# SRC="http://"
		tmp=`echo $tmp | cut -d/ -f4-`
	fi
	echo $tmp | awk -vFS="/" '{for (i=1;i<NF;i++) { printf $i"/" } }'
}

urlfile() {	# outputs "dir1/dir2/file.html" from url
	echo $1 | grep -qs //
	if [ $? -eq 0 ] ; then						# SRC="http://"
		echo $1 | cut -d/ -f4-
	else
		if [ x`echo $1 | cut -c1` == x/ ] ; then	# SRC="/blah/xx"
			echo $1 | cut -c2-
		else										# SRC="blah/xx"
			echo $1  
		fi
	fi
}

lochost() {
	echo $1 | grep -qs //			# got a http://
	if [ $? -eq 0 ] ; then
		str=`echo $1 | cut -d/ -f3`
	else
		str=`echo $1 | cut -d/ -f1`
	fi
	echo $str | cut -d: -f1
}

locport() {
	echo $1 | grep -qs //			# got a http://
	if [ $? -eq 0 ] ; then
		str=`echo $1 | cut -d/ -f3`
	else
		str=`echo $1 | cut -d/ -f1`
	fi
	echo $str | grep -qs :
	if [ $? -eq 0 ] ; then					# SRC="http://xxx:yy/"
		echo $str | cut -d: -f2	 
	else
		echo 80
	fi
}

locfile() {
	echo $1 | grep -qs //			# got a http://
	if [ $? -eq 0 ] ; then
		echo $1 | cut -d/ -f4-
	else 
		echo $1 | grep -qs /
		if [ $? -eq 0 ] ; then		# got a host/file
			echo $1 | cut -d/ -f2-
		fi
	fi
}

verbose=0
if [ "x$1" == "x-v" ] ; then
	verbose=1
	shift
fi

nokeep=1
if [ "x$1" == "x-k" ] ; then
	nokeep=0
	shift
fi

if [ "x$1" == "x" ] ; then
	echo "Usage: ionax [-v] [-k] <URL>"
	echo "ionax: where's da input?"
	echo "ionax: full URL please, and http only."
	echo "ionax: -v for verbose, -k to keep files pulled"
	echo ""
	exit
fi

hostname=`lochost $1`
port=`locport $1`
file=`locfile $1`
dir=`urldir $1`

tmpid=$hostname

if [ $verbose -gt 0 ]; then
	echo "Host: "$hostname
	echo "Port: $port"
	echo "File: /$file"
	echo "Dir: $dir"
fi

mkdir /tmp/ionax.$tmpid
size=`httphead $file | nc $hostname $port | tee /tmp/ionax.$tmpid/html | wc --bytes `
n=0
total=$size

if [ $verbose -gt 0 ]; then
	echo "HTML Size: $size"
	echo "--" ;
fi

files=`awk -- '

BEGIN	{ RS="<" }

toupper($0) ~ /^LINK/ { 
	p = index(toupper($0),"HREF=");
	s = substr ( $0, p+5, length($0));
	if ( ( substr ( s , 1 , 1 ) == "\"" ) || ( substr ( s, 1, 1) == sprintf("%c", 39) ) ) {
		split ( s, str, /"/ );
		print str[2];
	} else {								# if it evil src=aaa
		split ( s , str, / / );
		if ( substr ( str[1] , length ( str[1] ) , 1 ) == ">" ) { 
											# if it ends with a \> attached
			str[1] = substr ( str[1] , 1 , length ( str[1] ) - 1 );
		}
		print str[1];
	}
}

toupper($0) ~ /^IMG/ { 
	p = index(toupper($0),"SRC=");
	s = substr ( $0, p+4, length($0));
	if ( ( substr ( s , 1 , 1 ) == "\"" ) || ( substr ( s, 1, 1) ==  sprintf("%c", 39) ) ) {
		split ( s, str, /"/ );
		print str[2];
	} else {								# if it evil src=aaa
		split ( s , str, / / );
		if ( substr ( str[1] , length ( str[1] ) , 1 ) == ">" ) { 
											# if it ends with a \> attached
			str[1] = substr ( str[1] , 1 , length ( str[1] ) - 1 );
		}
		print str[1];
	}
}

toupper($0) ~ /BACKGROUND=/ {
	p = index(toupper($0),"BACKGROUND=");
	s = substr($0, p, length($0)-p);
	split ( s, str, /"/ );
	print str[2];
}

' /tmp/ionax.$tmpid/html | sort | uniq`

echo $files > /tmp/ionax.$tmpid/depend

for f in $files ; do
	hl=`urlhost $f`
	pl=`urlport $f`
	fl=`urlfile $f`
	size=`httphead $fl $dir | nc $hl $pl | tee /tmp/ionax.$tmpid/inline.$n | wc --bytes`
	total=`expr $total + $size` 
	n=`expr $n + 1`
	if [ $verbose -gt 0 ]; then
		echo "File $fl @ $size bytes"
		echo "Total $total bytes ($n inline files)"
		echo "--" ;
	fi
done
echo "Total $total bytes ($n inline files)"

if [ $nokeep -gt 0 ]; then
	rm -rf /tmp/ionax.$tmpid
fi