gistfile1.txt

#!/bin/bash

# Last modified: 2011 Dec 14
# Version 0.9

# Author: Florian CROUZAT <[email protected]>
# Feel free to do whatever you want with this file.
# Just make sure to credit what deserve credits.

# Arg(s): $1 - Arte+7 html URL of the page used to actually see the video.

# VARIABLES AND FUNCTIONS
BAD=$'\e[31;01m'
GOOD=$'\e[32;01m'
WARN=$'\e[33;01m'
HILITE=$'\e[36;01m'
NORMAL=$'\e[0m'
function die() {
   echo " ${BAD}*${NORMAL} $@" > /dev/stderr
   # Try to clean up mktemp
   rm $xmlinit $xmlfr $xmlde $xmls &>/dev/null
   exit 1
}
# END OF VARIABLES AND FUNCTIONS

# I need a single arg
[[ -n $1 ]] && echo || die "Usage: $0 http://videos.arte.tv/url-of-the-page-used-to-watch-the-video.html"

# Will I be able to download things ?
wget=false
curl=false
if type -p wget &>/dev/null ; then
   get="wget -q -O"
   wget=true
elif type -p curl &>/dev/null ; then
   get="curl -s -o"
   curl=true
else
   die "This script ${HILITE}require${NORMAL} wget or curl."
fi

# Creating tempfiles to temporary store XMLs
xmlinit=$(mktemp /tmp/xmlinitXXXXX) || die "Cannot create temporary file 1/4 using mktemp. Exiting."
xmls=$(mktemp /tmp/xmlsXXXXX) || die "Cannot create temporary file 2/4 using mktemp. Exiting."
xmlfr=$(mktemp /tmp/xmlfrXXXXX) || die "Cannot create temporary file 3/4 using mktemp. Exiting."
xmlde=$(mktemp /tmp/xmldeXXXXX) || die "Cannot create temporary file 4/4 using mktemp. Exiting."

# This is where the real stuff begins
# It's a three step parsing. It requires multiple hop to get to the rtmp URL.
# Sometimes, Arte delivers gzipped webpages, we have to take care of that too...

###########
# Step 1/3
# Parses the first page and search the link for the second page

echo -e "Parsing $1 ..."
if $wget ; then
   wget -q -O "$xmlinit" "$1"
elif $curl ; then
   curl -s -o "$xmlinit" "$1"
fi

# zgrep handles gzipped and plain text files the same so ...
step1=$(zgrep -E -o 'videorefFileUrl = "http://[^ ]*"' "$xmlinit" | awk '{print $3}' | tr -d '"')

# Have we found what we are looking for ?
[[ -z $step1 ]] && die "Parsing error: no \"videorefFileUrl\" found. Exiting"

# End of step 1
################

#########
# Step 2
# Parse the second page and search for two new pages: FR and DE

echo -e " $GOOD*$NORMAL Now parsing $step1 ..."
$get "$xmls" "$step1"

# zgrep handles gzipped and plain text files the same so ...
fr=$(zgrep -E "<video lang=\"fr\" ref=\"[^ ]*\" ?/>" $xmls | cut -d= -f3 | sed 's,/>,,' | tr -d '"')
de=$(zgrep -E "<video lang=\"de\" ref=\"[^ ]*\" ?/>" $xmls | cut -d= -f3 | sed 's,/>,,' | tr -d '"')

# Have we found what we are looking for ?
[[ -z $fr && -z $de ]] && die "Parsing error: no \"<video ref=\" found. Exiting"
echo -e " $GOOD*$NORMAL Found FR XML file: $fr ..."
echo -e " $GOOD*$NORMAL Found DE XML file: $de ..."

# End of step 2
################

#########
# Step 3
# Parse FR and DE page and search for RTMP links.

$get $xmlfr "$fr"
$get $xmlde "$de"

echo -e " $GOOD*$NORMAL Now parsing $fr ..."
# zgrep handles gzipped and plain text files the same so ...
rtmp[1]=$(zgrep -F '"sd"' $xmlfr | egrep -o "rtmp://[^ <]*")
rtmp[2]=$(zgrep -F '"hd"' $xmlfr | egrep -o "rtmp://[^ <]*")

echo -e " $GOOD*$NORMAL Now parsing $de ..."
# zgrep handles gzipped and plain text files the same so ...
rtmp[3]=$(zgrep -F '"sd"' $xmlde | egrep -o "rtmp://[^ <]*")
rtmp[4]=$(zgrep -F '"hd"' $xmlde | egrep -o "rtmp://[^ <]*")

echo
echo -e "Found 4 RTMP streams..."
echo -e " $GOOD*$NORMAL [1] French Standard-definition RTMP (${HILITE}FR_SD${NORMAL})"
echo -e "     -> ${rtmp[1]}"
echo -e " $GOOD*$NORMAL [2] French High-definition RTMP (${HILITE}FR_HD${NORMAL})"
echo -e "     -> ${rtmp[2]}"
echo -e " $GOOD*$NORMAL [3] German Standard-definition RTMP (${HILITE}DE_SD${NORMAL})"
echo -e "     -> ${rtmp[3]}"
echo -e " $GOOD*$NORMAL [4] German High-definition RTMP (${HILITE}DE_HD${NORMAL})"
echo -e "     -> ${rtmp[4]}"

rm $xmlinit $xmlfr $xmlde $xmls # Cleaning mktemp

# End of step 3
################

if type -p rtmpdump &>/dev/null ; then

   echo
   echo -n "Found rtmpdump, please paste the ${HILITE}number${NORMAL} of the above rtmp stream you want to transcode: "
   read -r whichrtmp
   [[ $whichrtmp =~ ^[1-4]$ ]] || die "Please enter a valid number (between 1 and 4). Exiting."

   echo -n "Choose a name to save the transcribed FLV: "
   read -r flv
   [[ -z $flv ]] && flv="rip$$"
   flv="${flv}.flv"

   echo
   echo " $WARN*$NORMAL I'll now enter in a 'while !completely_downloaded; do ... done' loop to transcode the rtmp stream into flv." > /dev/stderr
   echo " $WARN*$NORMAL It means that ^C won't be able to stop the script, if you really want to abort the processus," > /dev/stderr
   echo " $WARN*$NORMAL just use ^Z then 'kill %1' to kill the backgrounded job (or use ps(1), or another terminal)." > /dev/stderr
   echo
   echo -n "Press any key to start transcoding ... "
   read

   # Sometimes, the download isn't possible in one single try but the good thing
   # is that the return code when the download is incomplete is 2 so one can
   # loop until return code is 0
   ok=false
   rtmpdump -r "${rtmp[$whichrtmp]}" -o "$flv" && ok=true
   while ! $ok ; do
   	rtmpdump -r "${rtmp[$whichrtmp]}" -o "$flv" --resume && ok=true
   done
else
   echo
   echo "${BAD}*${NORMAL} Sorry, ${HILITE}rtmpdump${NORMAL} hasn't been found in your \$PATH." > /dev/stderr
   echo "${BAD}*${NORMAL} I cannot automagically transcode the rtmp ${HILITE}stream${NORMAL} into a flv ${HILITE}file${NORMAL} without this tool." > /dev/stderr
   echo "${BAD}*${NORMAL} Note that you still can save the rtmp link for future usage or to be used with another transcoding tool." > /dev/stderr
   exit 0
fi