vjt/placeholder.html

## placeholder.html
<html>
  <head>
    <title>The mirror is currently updating. Please wait.</title>
  </head>
  <body>
    <h2>The mirror is currently resyncing to the master site.</h2>
    <h3>The process will be completed in a matter of minutes.</h3>
    <p>In the meantime, try to visit the <a href="http://opensource.org">original site</a> or, if it's unreachable, try one of the following mirrors:</p>
    <p><a href="http://opensource.gds.tuwien.ac.at/">Austria</a> | <a href="http://os.fsfmirror.com/">Belgium</a> | <a href="http://opensource.usrbinruby.net/">Canada</a> (<a href="http://open2.mirrors-r-us.net/">2</a>) | <a href="http://opensource.mirrors.typhon.net/">France</a> (<a href="http://os3.fsfmirror.com/">2</a>) | <a href="http://opensource.mirroring.de/">Germany</a> (<a href="http://opensource.linux-mirror.org/">2</a>,<a href="http://os2.fsfmirror.com/">3</a>,<a href="http://opensource.erde3.net/">4</a>) | <a href="http://open3.mirrors-r-us.net">Japan</a> | (<a href="http://open1.mirrors-r-us.net/">2</a>) | <a href="http://os3.osmirror.com/">Singapore</a> | <a href="http://opensource.openmirrors.org">UK</a> (<a href="http://2opensource.openmirrors.org">2</a>) | USA: <a href="http://www.free-soft.org/mirrors/www.opensource.org/">LA</a>, <a href="http://osmirror.com/">Montana</a> (<a href="http://os2.osmirror.com/">2</a>), (<a href="http://opensource2.usrbinruby.net/">3</a>)</p>

    <p style="margin-top: 50px; font-style:italic">Thanks for visiting opensource.antifork.org!</p>
  </body>
</html>

## update_opensource_mirror.sh
#!/bin/bash
# Opensource.ORG quick&dirty mirror script.
# (C) 2007-2010 Marcello Barnaba <vjt@openssl.it>
#
# Released under the terms of the Beerware License: if we
# meet some day and you think this stuff is worth it, you
# can buy me a beer in return.
#

# The absolute parent path of the public mirror directory
MIRROR_BASE="/home/httpd/antifork.org/htdocs"

# The relative to $MIRROR_BASE path that contains the mirrored HTML
MIRROR_DIR="opensource.antifork.org"

# The URI to mirror - why is this a variable?!
MIRROR_URI="http://opensource.org"

# A directory containing a placeholder page to show to clients
# while the script is running - a fail whale, in twitter terms
WIP_DIR="${MIRROR_DIR}-updating"

# A Temporary working directory. Everyone needs one!
WORK_DIR="${MIRROR_DIR}-$RANDOM"

# The absolute path to the wget(1) binary
WGET="/usr/bin/wget --quiet"

# Abort on errors .. yuck!
set -e

# .................................................
# NO USER SERVICEABLE PARTS BELOW THE DOTTED LINE .
# .................................................

pushd "$MIRROR_BASE" >/dev/null

# wget madness downloads robots.txt continuously,
# so erase and re-download every time. for this kind
# of site, it's ok -- but should be fixed.
#
rm -rf $MIRROR_DIR
ln -s $WIP_DIR $MIRROR_DIR
mkdir $WORK_DIR

# Download HTML and images
#
$WGET --domains=opensource.org --convert-links --level=0  \
    --mirror --page-requisites --no-host-directories  \
    $MIRROR_URI --exclude-directories '/user,/event'  \
    --directory-prefix="$WORK_DIR" --html-extension

# Get all the stylesheets from the index page, the one that
# will never, hopefully, disappear.
# It uses Ruby and Nokogiri: you'd best have them installed :-p
#
pushd $WORK_DIR >/dev/null
PARSER="require 'nokogiri'; puts Nokogiri::HTML(File.read('index.html')).search('link[rel=stylesheet]').map {|f| CGI.unescape(f.attr('href'))}.select {|f| File.exists?(f)}.join(10.chr)"
STYLESHEETS=$(ruby -rubygems -rcgi -e "$PARSER")
popd >/dev/null

# Download all the images referenced in the CSS style sheets
#
for stylesheet in $STYLESHEETS; do
    base=`dirname $stylesheet`
    sed -n 's#.*url(\(.*\)).*#\1#p' < "$WORK_DIR/$stylesheet" | sort | uniq | while read image; do
      mkdir -p "`dirname $WORK_DIR/$base/$image`"
      $WGET -O $WORK_DIR/$base/$image $MIRROR_URI/$base/$image
    done
done

# Remove the placeholder page
rm -f $MIRROR_DIR

# Put the mirror online - yay!
mv $WORK_DIR $MIRROR_DIR

popd >/dev/null

# EOF
	<html>
	<head>
	<title>The mirror is currently updating. Please wait.</title>
	</head>
	<body>
	<h2>The mirror is currently resyncing to the master site.</h2>
	<h3>The process will be completed in a matter of minutes.</h3>
	<p>In the meantime, try to visit the <a href="http://opensource.org">original site</a> or, if it's unreachable, try one of the following mirrors:</p>
	<p><a href="http://opensource.gds.tuwien.ac.at/">Austria</a> \| <a href="http://os.fsfmirror.com/">Belgium</a> \| <a href="http://opensource.usrbinruby.net/">Canada</a> (<a href="http://open2.mirrors-r-us.net/">2</a>) \| <a href="http://opensource.mirrors.typhon.net/">France</a> (<a href="http://os3.fsfmirror.com/">2</a>) \| <a href="http://opensource.mirroring.de/">Germany</a> (<a href="http://opensource.linux-mirror.org/">2</a>,<a href="http://os2.fsfmirror.com/">3</a>,<a href="http://opensource.erde3.net/">4</a>) \| <a href="http://open3.mirrors-r-us.net">Japan</a> \| (<a href="http://open1.mirrors-r-us.net/">2</a>) \| <a href="http://os3.osmirror.com/">Singapore</a> \| <a href="http://opensource.openmirrors.org">UK</a> (<a href="http://2opensource.openmirrors.org">2</a>) \| USA: <a href="http://www.free-soft.org/mirrors/www.opensource.org/">LA</a>, <a href="http://osmirror.com/">Montana</a> (<a href="http://os2.osmirror.com/">2</a>), (<a href="http://opensource2.usrbinruby.net/">3</a>)</p>

	<p style="margin-top: 50px; font-style:italic">Thanks for visiting opensource.antifork.org!</p>
	</body>
	</html>
	#!/bin/bash
	# Opensource.ORG quick&dirty mirror script.
	# (C) 2007-2010 Marcello Barnaba <vjt@openssl.it>
	#
	# Released under the terms of the Beerware License: if we
	# meet some day and you think this stuff is worth it, you
	# can buy me a beer in return.
	#

	# The absolute parent path of the public mirror directory
	MIRROR_BASE="/home/httpd/antifork.org/htdocs"

	# The relative to $MIRROR_BASE path that contains the mirrored HTML
	MIRROR_DIR="opensource.antifork.org"

	# The URI to mirror - why is this a variable?!
	MIRROR_URI="http://opensource.org"

	# A directory containing a placeholder page to show to clients
	# while the script is running - a fail whale, in twitter terms
	WIP_DIR="${MIRROR_DIR}-updating"

	# A Temporary working directory. Everyone needs one!
	WORK_DIR="${MIRROR_DIR}-$RANDOM"

	# The absolute path to the wget(1) binary
	WGET="/usr/bin/wget --quiet"

	# Abort on errors .. yuck!
	set -e

	# .................................................
	# NO USER SERVICEABLE PARTS BELOW THE DOTTED LINE .
	# .................................................

	pushd "$MIRROR_BASE" >/dev/null

	# wget madness downloads robots.txt continuously,
	# so erase and re-download every time. for this kind
	# of site, it's ok -- but should be fixed.
	#
	rm -rf $MIRROR_DIR
	ln -s $WIP_DIR $MIRROR_DIR
	mkdir $WORK_DIR

	# Download HTML and images
	#
	$WGET --domains=opensource.org --convert-links --level=0 \
	--mirror --page-requisites --no-host-directories \
	$MIRROR_URI --exclude-directories '/user,/event' \
	--directory-prefix="$WORK_DIR" --html-extension

	# Get all the stylesheets from the index page, the one that
	# will never, hopefully, disappear.
	# It uses Ruby and Nokogiri: you'd best have them installed :-p
	#
	pushd $WORK_DIR >/dev/null
	PARSER="require 'nokogiri'; puts Nokogiri::HTML(File.read('index.html')).search('link[rel=stylesheet]').map {\|f\| CGI.unescape(f.attr('href'))}.select {\|f\| File.exists?(f)}.join(10.chr)"
	STYLESHEETS=$(ruby -rubygems -rcgi -e "$PARSER")
	popd >/dev/null

	# Download all the images referenced in the CSS style sheets
	#
	for stylesheet in $STYLESHEETS; do
	base=`dirname $stylesheet`
	sed -n 's#.url(\(.\)).*#\1#p' < "$WORK_DIR/$stylesheet" \| sort \| uniq \| while read image; do
	mkdir -p "`dirname $WORK_DIR/$base/$image`"
	$WGET -O $WORK_DIR/$base/$image $MIRROR_URI/$base/$image
	done
	done

	# Remove the placeholder page
	rm -f $MIRROR_DIR

	# Put the mirror online - yay!
	mv $WORK_DIR $MIRROR_DIR

	popd >/dev/null

	# EOF