--- get_book.sh	2004/02/15 11:40:43	1.7
+++ get_book.sh	2004/07/21 13:58:35	1.12
@@ -1,13 +1,20 @@
 #!/bin/sh
 
 # proxy settings (same as in firebird)
-fping proxy && export http_proxy=http://proxy:8080
+fping -q proxy && export http_proxy=http://proxy:8080
+if [ -z "$http_proxy" ] ; then
+	fping -q proxy.lan && export http_proxy=http://proxy.lan:8080
+fi
 # user agent (same as in firebird)
-ua="Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.5) Gecko/20031225 Firebird/0.7"
+ua="Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.6) Gecko/20040506 Firefox/0.8"
+cookie_file="cookie.txt"
 
 # wait between pages
 export wait=120
 
+if [ ! -z "$http_proxy" ] ; then
+	echo "Using proxy $http_proxy"
+fi
 
 if [ -z "$1" ] ; then
 	echo "Usage: $0 ISBN"
@@ -27,9 +34,7 @@
 	grep -l 'This is only a preview of the full book' * | xargs -i rm {}
 fi
 
-isbn=$1;
-
-isbn2=`echo $isbn | sed 's/-//g'`
+isbn=`echo $1 | sed 's/-//g' | tr '[a-z]' '[A-Z]'`
 
 function mirror() {
 
@@ -38,29 +43,38 @@
 	file=`echo $url | sed -e s,http://[^?]*?,index.html?, -e s,#.*$,, -e s,/,%2F,g`
 	if [ -e "$file" ] ; then
 #		echo "skip $url"
+		echo -n "."
 		return
 	fi
 
+	cookie=""
 	if echo $url | grep '/index' >/dev/null ; then
-		cookies=""
 		echo -n "no login (index) "
 	elif echo $url | grep 'mode=toc' >/dev/null ; then
-		cookies=""
 		echo -n "no login (toc) "
 	else
-		cookies="--load-cookies=/home/dpavlin/.phoenix/default/g6b45nt6.slt/cookies.txt"
-		echo -n "with login "
+		if [ ! -e $cookie_file ] ; then
+			echo "cookies file $cookie_file doesn't exits! Please create it."
+			echo "It should be in format:"
+			echo "Cookie: Site=UICode=&Portal=oreilly&GUID=..."
+			exit 1
+		fi
+		read cookie < $cookie_file
+		if [ -z "$cookie" ] ; then
+			echo "Empty cookie file $cookie_file !"
+			exit 1
+		fi
 	fi
-	echo $url
 
-	wget -p -nH -nc -k \
-		-t 1 -U "$ua" \
-		$cookies $url
+	if [ -z "$cookie" ] ; then
+		echo "$url [no cookie]"
+	else
+		echo "$url [with cookie]"
+	fi
 
-#	-D safari.oreilly.com \
-#	-A 0-201-41975-0 \
+	wget -q -p -nH -nc -k -t 1 -U "$ua" --cookies=off --header="$cookie" $url
+	perl -e '$t=rand($ENV{wait} || 120);print "sleep for $t sec.\n"; sleep($t);'
 
-	perl -e '$t=rand($ENV{wait} || 30);print "sleep for $t sec.\n"; sleep($t);'
 }
 
 function geturl() {
@@ -68,19 +82,21 @@
 	sed -e 's/^.*<a.*href="//i' \
 		-e 's/".*//' -e 's/amp;//g' \
 		-e 's,^[^\?]*\?,http://safari.oreilly.com/,' \
-		-e 's/#$//' \
+		-e 's/#.*$//' \
 		-e 's/\&srchText=//' \
 		-e 's/open=false/open=true/' | \
-		grep '&s=1&b=1&f=1&t=1&c=1&u=1&r=&o=1' | \
-		grep $2 | \
-		grep -v "$2/[0-9][0-9][0-9][0-9][0-9][0-9][0-9]" | \
+		grep '&s=1&b=1&f=1&t=1&c=1&u=1&r=&o=1&n=1&d=1&p=1&a=0' | \
+		grep -i "xmlid=[0-9A-Za-z\-][0-9A-Za-z\-]*" | \
+		grep -vi "xmlid=[0-9A-Za-z\-]*/[0-9][0-9][0-9][0-9][0-9][0-9][0-9]" | \
 		sort -u >> in
 }
 
 function uniqurl() {
 	mv in in.tmp
-	grep -v 'view=[A-Z]' in.tmp | sort -u > in
-	grep 'view=[A-Z].*/index' in.tmp | sort -u >> in
+	grep -v 'view=[A-Z]' in.tmp | grep 'a=0$' | sort -u > in
+	grep 'view=[A-Z].*/index' in.tmp | grep 'a=0$' | sort -u >> in
+	links=`wc -l in | cut -d" " -f1`
+	echo "found $links unique links"
 }
 
 function mirror_in() {
@@ -108,13 +124,17 @@
 			exit 1
 		fi
 	done
+	echo
 }
 
 echo -n > in
-mirror "http://safari.oreilly.com/?XmlId=$isbn"
+mirror "http://safari.oreilly.com/$isbn"
+echo
 
-echo "extract URLs from first page..."
-geturl "index.html?XmlId=$isbn" $isbn
+file=`ls *index.html?XmlId=*`
+isbn=`echo $file | cut -d= -f2`
+echo "extract URLs from first page $file... [$isbn]"
+geturl $file $isbn
 uniqurl
 
 mirror_in