#!/usr/bin/perl -w
#
-# webcollage, Copyright (c) 1999-2012 by Jamie Zawinski <jwz@jwz.org>
+# webcollage, Copyright © 1999-2014 by Jamie Zawinski <jwz@jwz.org>
# This program decorates the screen with random images from the web.
# One satisfied customer described it as "a nonstop pop culture brainbath."
#
#use diagnostics;
-use Socket;
require Time::Local;
require POSIX;
use Fcntl ':flock'; # import LOCK_* constants
use POSIX qw(strftime);
-
-use bytes; # Larry can take Unicode and shove it up his ass sideways.
- # Perl 5.8.0 causes us to start getting incomprehensible
- # errors about UTF-8 all over the place without this.
+use LWP::UserAgent;
+use bytes;
my $progname = $0; $progname =~ s@.*/@@g;
-my $version = q{ $Revision: 1.158 $ }; $version =~ s/^[^0-9]+([0-9.]+).*$/$1/;
-my $copyright = "WebCollage $version, Copyright (c) 1999-2011" .
+my ($version) = ('$Revision: 1.167 $' =~ m/\s(\d[.\d]+)\s/s);
+my $copyright = "WebCollage $version, Copyright (c) 1999-2014" .
" Jamie Zawinski <jwz\@jwz.org>\n" .
" http://www.jwz.org/webcollage/\n";
-my @search_methods = ( 24, "googlephotos", \&pick_from_google_image_photos,
- 13, "googleimgs", \&pick_from_google_images,
- 13, "googlenums", \&pick_from_google_image_numbers,
- 16, "flickr_recent", \&pick_from_flickr_recent,
- 13, "flickr_random", \&pick_from_flickr_random,
- 10, "twitpic", \&pick_from_twitpic_images,
- 8, "livejournal", \&pick_from_livejournal_images,
- 3, "yahoorand", \&pick_from_yahoo_random_link,
+my @search_methods = (
+ # Google is rate-limiting us now, so this works ok from
+ # a short-running screen saver, but not as a batch job.
+ # I haven't found a workaround.
+ #
+ 7, "googlephotos", \&pick_from_google_image_photos,
+ 5, "googleimgs", \&pick_from_google_images,
+ 5, "googlenums", \&pick_from_google_image_numbers,
+
+ # So let's try Bing instead. No rate limiting yet!
+ #
+ 7, "bingphotos", \&pick_from_bing_image_photos,
+ 6, "bingimgs", \&pick_from_bing_images,
+ 6, "bingnums", \&pick_from_bing_image_numbers,
+
+ 21, "flickr_recent", \&pick_from_flickr_recent,
+ 16, "flickr_random", \&pick_from_flickr_random,
+ 23, "instagram", \&pick_from_instagram,
+ 4, "livejournal", \&pick_from_livejournal_images,
+
+ # No longer exists, as of Apr 2014
+ # 4, "yahoorand", \&pick_from_yahoo_random_link,
- # This one doesn't work very well: too many non-img links.
- 0, "twitter", \&pick_from_twitter_images,
+ # Twitter destroyed their whole API in 2013.
+ # 0, "twitpic", \&pick_from_twitpic_images,
+ # 0, "twitter", \&pick_from_twitter_images,
# This is a cute way to search for a certain webcams.
# Not included in default methods, since these images
"pics.livejournal.com" => 1,
"tinypic.com" => 1,
"flickr.com" => 1,
+ "staticflickr.com" => 1,
"pbase.com" => 1,
"blogger.com" => 1,
"multiply.com" => 1,
"yandex.ru" => 1,
"imgur.com" => 1,
"yfrog.com" => 1,
+ "cdninstagram.com" => 1,
"yimg.com" => 1, # This is where dailynews.yahoo.com stores
"eimg.com" => 1, # its images, so pick_from_yahoo_news_text()
#
my %entity_table = (
"apos" => '\'',
- "quot" => '"', "amp" => '&', "lt" => '<', "gt" => '>',
- "nbsp" => ' ', "iexcl" => '¡', "cent" => '¢', "pound" => '£',
- "curren" => '¤', "yen" => '¥', "brvbar" => '¦', "sect" => '§',
- "uml" => '¨', "copy" => '©', "ordf" => 'ª', "laquo" => '«',
- "not" => '¬', "shy" => '', "reg" => '®', "macr" => '¯',
- "deg" => '°', "plusmn" => '±', "sup2" => '²', "sup3" => '³',
- "acute" => '´', "micro" => 'µ', "para" => '¶', "middot" => '·',
- "cedil" => '¸', "sup1" => '¹', "ordm" => 'º', "raquo" => '»',
- "frac14" => '¼', "frac12" => '½', "frac34" => '¾', "iquest" => '¿',
- "Agrave" => 'À', "Aacute" => 'Á', "Acirc" => 'Â', "Atilde" => 'Ã',
- "Auml" => 'Ä', "Aring" => 'Å', "AElig" => 'Æ', "Ccedil" => 'Ç',
- "Egrave" => 'È', "Eacute" => 'É', "Ecirc" => 'Ê', "Euml" => 'Ë',
- "Igrave" => 'Ì', "Iacute" => 'Í', "Icirc" => 'Î', "Iuml" => 'Ï',
- "ETH" => 'Ð', "Ntilde" => 'Ñ', "Ograve" => 'Ò', "Oacute" => 'Ó',
- "Ocirc" => 'Ô', "Otilde" => 'Õ', "Ouml" => 'Ö', "times" => '×',
- "Oslash" => 'Ø', "Ugrave" => 'Ù', "Uacute" => 'Ú', "Ucirc" => 'Û',
- "Uuml" => 'Ü', "Yacute" => 'Ý', "THORN" => 'Þ', "szlig" => 'ß',
- "agrave" => 'à', "aacute" => 'á', "acirc" => 'â', "atilde" => 'ã',
- "auml" => 'ä', "aring" => 'å', "aelig" => 'æ', "ccedil" => 'ç',
- "egrave" => 'è', "eacute" => 'é', "ecirc" => 'ê', "euml" => 'ë',
- "igrave" => 'ì', "iacute" => 'í', "icirc" => 'î', "iuml" => 'ï',
- "eth" => 'ð', "ntilde" => 'ñ', "ograve" => 'ò', "oacute" => 'ó',
- "ocirc" => 'ô', "otilde" => 'õ', "ouml" => 'ö', "divide" => '÷',
- "oslash" => 'ø', "ugrave" => 'ù', "uacute" => 'ú', "ucirc" => 'û',
- "uuml" => 'ü', "yacute" => 'ý', "thorn" => 'þ', "yuml" => 'ÿ',
+ "quot" => '"', "amp" => '&', "lt" => '<',
+ "gt" => '>', "nbsp" => ' ', "iexcl" => '',
+ "cent" => "\xA2", "pound" => "\xA3", "curren" => "\xA4",
+ "yen" => "\xA5", "brvbar" => "\xA6", "sect" => "\xA7",
+ "uml" => "\xA8", "copy" => "\xA9", "ordf" => "\xAA",
+ "laquo" => "\xAB", "not" => "\xAC", "shy" => "\xAD",
+ "reg" => "\xAE", "macr" => "\xAF", "deg" => "\xB0",
+ "plusmn" => "\xB1", "sup2" => "\xB2", "sup3" => "\xB3",
+ "acute" => "\xB4", "micro" => "\xB5", "para" => "\xB6",
+ "middot" => "\xB7", "cedil" => "\xB8", "sup1" => "\xB9",
+ "ordm" => "\xBA", "raquo" => "\xBB", "frac14" => "\xBC",
+ "frac12" => "\xBD", "frac34" => "\xBE", "iquest" => "\xBF",
+ "Agrave" => "\xC0", "Aacute" => "\xC1", "Acirc" => "\xC2",
+ "Atilde" => "\xC3", "Auml" => "\xC4", "Aring" => "\xC5",
+ "AElig" => "\xC6", "Ccedil" => "\xC7", "Egrave" => "\xC8",
+ "Eacute" => "\xC9", "Ecirc" => "\xCA", "Euml" => "\xCB",
+ "Igrave" => "\xCC", "Iacute" => "\xCD", "Icirc" => "\xCE",
+ "Iuml" => "\xCF", "ETH" => "\xD0", "Ntilde" => "\xD1",
+ "Ograve" => "\xD2", "Oacute" => "\xD3", "Ocirc" => "\xD4",
+ "Otilde" => "\xD5", "Ouml" => "\xD6", "times" => "\xD7",
+ "Oslash" => "\xD8", "Ugrave" => "\xD9", "Uacute" => "\xDA",
+ "Ucirc" => "\xDB", "Uuml" => "\xDC", "Yacute" => "\xDD",
+ "THORN" => "\xDE", "szlig" => "\xDF", "agrave" => "\xE0",
+ "aacute" => "\xE1", "acirc" => "\xE2", "atilde" => "\xE3",
+ "auml" => "\xE4", "aring" => "\xE5", "aelig" => "\xE6",
+ "ccedil" => "\xE7", "egrave" => "\xE8", "eacute" => "\xE9",
+ "ecirc" => "\xEA", "euml" => "\xEB", "igrave" => "\xEC",
+ "iacute" => "\xED", "icirc" => "\xEE", "iuml" => "\xEF",
+ "eth" => "\xF0", "ntilde" => "\xF1", "ograve" => "\xF2",
+ "oacute" => "\xF3", "ocirc" => "\xF4", "otilde" => "\xF5",
+ "ouml" => "\xF6", "divide" => "\xF7", "oslash" => "\xF8",
+ "ugrave" => "\xF9", "uacute" => "\xFA", "ucirc" => "\xFB",
+ "uuml" => "\xFC", "yacute" => "\xFD", "thorn" => "\xFE",
+ "yuml" => "\xFF",
# HTML 4 entities that do not have 1:1 Latin1 mappings.
- "bull" => "*", "hellip"=> "...", "prime" => "'", "Prime" => "\"",
- "frasl" => "/", "trade" => "[tm]", "larr" => "<-", "rarr" => "->",
- "harr" => "<->", "lArr" => "<=", "rArr" => "=>", "hArr" => "<=>",
- "empty" => "Ø", "minus" => "-", "lowast"=> "*", "sim" => "~",
- "cong" => "=~", "asymp" => "~", "ne" => "!=", "equiv" => "==",
- "le" => "<=", "ge" => ">=", "lang" => "<", "rang" => ">",
- "loz" => "<>", "OElig" => "OE", "oelig" => "oe", "Yuml" => "Y",
- "circ" => "^", "tilde" => "~", "ensp" => " ", "emsp" => " ",
- "thinsp"=> " ", "ndash" => "-", "mdash" => "--", "lsquo" => "`",
- "rsquo" => "'", "sbquo" => "'", "ldquo" => "\"", "rdquo" => "\"",
- "bdquo" => "\"", "lsaquo"=> "<", "rsaquo"=> ">",
+ "bull" => "*", "hellip"=> "...", "prime" => "'", "Prime" => "\"",
+ "frasl" => "/", "trade" => "[tm]", "larr" => "<-", "rarr" => "->",
+ "harr" => "<->", "lArr" => "<=", "rArr" => "=>", "hArr" => "<=>",
+ "empty" => "\xD8", "minus" => "-", "lowast"=> "*", "sim" => "~",
+ "cong" => "=~", "asymp" => "~", "ne" => "!=", "equiv" => "==",
+ "le" => "<=", "ge" => ">=", "lang" => "<", "rang" => ">",
+ "loz" => "<>", "OElig" => "OE", "oelig" => "oe", "Yuml" => "Y",
+ "circ" => "^", "tilde" => "~", "ensp" => " ", "emsp" => " ",
+ "thinsp"=> " ", "ndash" => "-", "mdash" => "--", "lsquo" => "`",
+ "rsquo" => "'", "sbquo" => "'", "ldquo" => "\"", "rdquo" => "\"",
+ "bdquo" => "\"", "lsaquo"=> "<", "rsaquo"=> ">",
);
if (!defined($timeout)) { $timeout = $http_timeout; }
if ($timeout > $http_timeout) { $timeout = $http_timeout; }
- if ($timeout <= 0) {
- LOG (($verbose_net || $verbose_load), "timed out for $url");
- return ();
- }
+ my $user_agent = "$progname/$version";
- LOG ($verbose_net, "get_document_1 $url " . ($referer ? $referer : ""));
+ if ($url =~ m@^http://www\.altavista\.com/@ ||
+ $url =~ m@^http://random\.yahoo\.com/@ ||
+ $url =~ m@^http://images\.google\.com/@ ||
+ $url =~ m@^http://www\.google\.com/@) {
+ # block this, you turkeys.
+ $user_agent = "Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.7)" .
+ " Gecko/20070914 Firefox/2.0.0.7";
- if (! ($url =~ m@^http://@i)) {
- LOG ($verbose_net, "not an HTTP URL: $url");
- return ();
+ # 28-Jun-2007: Google Images now emits the entire page in JS if
+ # you claim to be Gecko. They also still block "webcollage".
+ # They serve non-JS for unrecognised agents, so let's try this...
+ $user_agent = "NoJavascriptPlease/1.0"
+ if ($url =~ m@^http://[a-z]+\.google\.com/@);
}
- my ($url_proto, $dummy, $serverstring, $path) = split(/\//, $url, 4);
- $path = "" unless $path;
+ my $ua = LWP::UserAgent->new;
+ $ua->env_proxy();
+ $ua->agent ("$progname/$version");
+ $ua->default_header ('Referer' => $referer);
+ $ua->timeout($timeout) if $timeout;
- if (!$url_proto || !$serverstring) {
- LOG (($verbose_net || $verbose_load), "unparsable URL: $url");
- return ();
+ if ($verbose_http) {
+ LOG (1, " ==> GET $url");
+ LOG (1, " ==> User-Agent: $user_agent");
+ LOG (1, " ==> Referer: $referer") if $referer;
}
- my ($them,$port) = split(/:/, $serverstring);
- $port = 80 unless $port;
+ my $res = $ua->get ($url);
- my $them2 = $them;
- my $port2 = $port;
- if ($http_proxy) {
- $serverstring = $http_proxy if $http_proxy;
- $serverstring =~ s@^[a-z]+://@@;
- ($them2,$port2) = split(/:/, $serverstring);
- $port2 = 80 unless $port2;
- }
+ my $http = ($res ? $res->status_line : '') || '';
+ my $head = ($res ? $res->headers()->as_string : '') || '';
+ my $body = ($res && $res->is_success ? $res->decoded_content : '') || '';
- my ($remote, $iaddr, $paddr, $proto, $line);
- $remote = $them2;
- if ($port2 =~ /\D/) { $port2 = getservbyname($port2, 'tcp') }
- if (!$port2) {
- LOG (($verbose_net || $verbose_load), "unrecognised port in $url");
- return ();
- }
- $iaddr = inet_aton($remote);
- if (!$iaddr) {
- LOG (($verbose_net || $verbose_load), "host not found: $remote");
- return ();
- }
- $paddr = sockaddr_in($port2, $iaddr);
-
-
- my $head = "";
- my $body = "";
-
- @_ =
- eval {
- local $SIG{ALRM} = sub {
- LOG (($verbose_net || $verbose_load), "timed out ($timeout) for $url");
- die "alarm\n";
- };
- alarm $timeout;
-
- $proto = getprotobyname('tcp');
- if (!socket(S, PF_INET, SOCK_STREAM, $proto)) {
- LOG (($verbose_net || $verbose_load), "socket: $!");
- return ();
- }
- if (!connect(S, $paddr)) {
- LOG (($verbose_net || $verbose_load), "connect($serverstring): $!");
- return ();
- }
-
- select(S); $| = 1; select(STDOUT);
-
- my $cookie = $cookies{$them};
-
- my $user_agent = "$progname/$version";
-
- if ($url =~ m@^http://www\.altavista\.com/@ ||
- $url =~ m@^http://random\.yahoo\.com/@ ||
- $url =~ m@^http://images\.google\.com/@ ||
- $url =~ m@^http://www\.google\.com/@) {
- # block this, you turkeys.
- $user_agent = "Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.7)" .
- " Gecko/20070914 Firefox/2.0.0.7";
-
- # 28-Jun-2007: Google Images now emits the entire page in JS if
- # you claim to be Gecko. They also still block "webcollage".
- # They serve non-JS for unrecognised agents, so let's try this...
- $user_agent = "NoJavascriptPlease/1.0"
- if ($url =~ m@^http://[a-z]+\.google\.com/@);
- }
-
- my $hdrs = "GET " . ($http_proxy ? $url : "/$path") . " HTTP/1.0\r\n" .
- "Host: $them\r\n" .
- "User-Agent: $user_agent\r\n";
- if ($referer) {
- $hdrs .= "Referer: $referer\r\n";
- }
- if ($cookie) {
- my @cc = split(/\r?\n/, $cookie);
- $hdrs .= "Cookie: " . join('; ', @cc) . "\r\n";
- }
- $hdrs .= "\r\n";
-
- foreach (split('\r?\n', $hdrs)) {
- LOG ($verbose_http, " ==> $_");
- }
- print S $hdrs;
- my $http = <S> || "";
-
- # Kludge: the Yahoo Random Link is now returning as its first
- # line "Status: 301" instead of "HTTP/1.0 301 Found". Fix it...
- #
- $http =~ s@^Status:\s+(\d+)\b@HTTP/1.0 $1@i;
+ LOG ($verbose_net, "get_document_1 $url " . ($referer ? $referer : ""));
- $_ = $http;
- s/[\r\n]+$//s;
+ $head =~ s/\r\n/\n/gs;
+ $head =~ s/\r/\n/gs;
+ if ($verbose_http) {
+ foreach (split (/\n/, $head)) {
LOG ($verbose_http, " <== $_");
-
- while (<S>) {
- $head .= $_;
- s/[\r\n]+$//s;
- last if m@^$@;
- LOG ($verbose_http, " <== $_");
-
- if (m@^Set-cookie:\s*([^;\r\n]+)@i) {
- set_cookie($them, $1)
- }
- }
-
- my $lines = 0;
- while (<S>) {
- $body .= $_;
- $lines++;
- }
-
- LOG ($verbose_http,
- " <== [ body ]: $lines lines, " . length($body) . " bytes");
-
- close S;
-
- if (!$http) {
- LOG (($verbose_net || $verbose_load), "null response: $url");
- return ();
- }
-
- $SIG{ALRM} = 'DEFAULT'; # seem to be suffering a race?
- return ( $http, $head, $body );
- };
- die if ($@ && $@ ne "alarm\n"); # propagate errors
-
- if ($@ && $@ ne "alarm\n") {
- print STDERR blurb() . "DIE " . join(" ", $@) . "\n";
- die;
+ }
}
- if ($@) {
- # timed out
- $head = undef;
- $body = undef;
- $suppress_audit = 1;
+ my @L = split(/\r\n|\r|\n/, $body);
+ my $lines = @L;
+ LOG ($verbose_http,
+ " <== [ body ]: $lines lines, " . length($body) . " bytes");
+
+ if (!$http) {
+ LOG (($verbose_net || $verbose_load), "null response: $url");
return ();
- } else {
- # didn't
- alarm 0;
- return @_;
}
+
+ return ( $http, $head, $body );
}
my @subpages;
+ if ($body =~ m/^\{\"/s) { # Google AJAX JSON response.
+
+ my @chunks = split (/"GsearchResultClass"/, $body);
+ shift @chunks;
+ my $body2 = '';
+ my $n = 1;
+ foreach (@chunks) {
+ my ($img) = m/"unescapedUrl":"(.*?)"/si;
+ my ($url) = m/"originalContextUrl":"(.*?)"/si;
+ next unless ($img && $url);
+ $url = ("/imgres" .
+ "?imgurl=" . url_quote($img) .
+ "&imgrefurl=" . url_quote($url) .
+ "&...");
+ $body2 .= "<A HREF=\"" . html_quote($url) . "\">$n</A>\n";
+ $n++;
+ }
+ $body = $body2 if $body2;
+ }
+
my $search_count = "?";
if ($body =~ m@found (approximately |about )?(<B>)?(\d+)(</B>)? image@) {
$search_count = $3;
my ($u) = m@<A\s.*\bHREF\s*=\s*([^>]+)>@i;
next unless $u;
- if ($u =~ m/^\"([^\"]*)\"/) { $u = $1; } # quoted string
- elsif ($u =~ m/^([^\s]*)\s/) { $u = $1; } # or token
+ if (m/\bm="{(.*?)}"/s) { # Bing info is inside JSON crud
+ my $json = html_unquote($1);
+ my ($href) = ($json =~ m/\bsurl:"(.*?)"/s);
+ my ($img) = ($json =~ m/\bimgurl:"(.*?)"/s);
+ $u = "$img\t$href" if ($img && $href);
+
+ } elsif ($u =~ m/^\"([^\"]*)\"/) { $u = $1 # quoted string
+ } elsif ($u =~ m/^([^\s]*)\s/) { $u = $1; # or token
+ }
if ( $rejected_urls{$u} ) {
LOG ($verbose_filter, " pre-rejecting candidate: $u");
}
\f
-############################################################################
+#############################################################################
+##
+## Pick images from random pages returned by the Yahoo Random Link
+##
+#############################################################################
+#
+## yahoorand
+#my $yahoo_random_link = "http://random.yahoo.com/fast/ryl";
#
-# Pick images from random pages returned by the Yahoo Random Link
#
-############################################################################
-
-# yahoorand
-my $yahoo_random_link = "http://random.yahoo.com/fast/ryl";
-
-
# Picks a random page; picks a random image on that page;
# returns two URLs: the page containing the image, and the image.
# Returns () if nothing found this time.
#
-sub pick_from_yahoo_random_link($) {
- my ($timeout) = @_;
-
- print STDERR "\n\n" if ($verbose_load);
- LOG ($verbose_load, "URL: $yahoo_random_link");
-
- $last_search = $yahoo_random_link; # for warnings
-
- $suppress_audit = 1;
-
- my ( $base, $body ) = get_document ($yahoo_random_link, undef, $timeout);
- if (!$base || !$body) {
- $body = undef;
- return;
- }
-
- LOG ($verbose_load, "redirected to: $base");
-
- my $img = pick_image_from_body ($base, $body);
- $body = undef;
-
- if ($img) {
- return ($base, $img);
- } else {
- return ();
- }
-}
+#sub pick_from_yahoo_random_link($) {
+# my ($timeout) = @_;
+#
+# print STDERR "\n\n" if ($verbose_load);
+# LOG ($verbose_load, "URL: $yahoo_random_link");
+#
+# $last_search = $yahoo_random_link; # for warnings
+#
+# $suppress_audit = 1;
+#
+# my ( $base, $body ) = get_document ($yahoo_random_link, undef, $timeout);
+# if (!$base || !$body) {
+# $body = undef;
+# return;
+# }
+#
+# LOG ($verbose_load, "redirected to: $base");
+#
+# my $img = pick_image_from_body ($base, $body);
+# $body = undef;
+#
+# if ($img) {
+# return ($base, $img);
+# } else {
+# return ();
+# }
+#}
\f
############################################################################
############################################################################
-my $google_images_url = "http://images.google.com/images" .
- "?site=images" . # photos
- "&btnG=Search" . # graphics
- "&safe=off" . # no screening
- "&imgsafe=off" .
+my $google_images_url = "http://ajax.googleapis.com/ajax/services/" .
+ "search/images" .
+ "?v=1.0" .
+ "&rsz=large" .
"&q=";
# googleimgs
$words = random_word(); # only one word for Google
}
- my $page = (int(rand(9)) + 1);
- my $num = 20; # 20 images per page
- my $search_url = $google_images_url . $words;
-
- if ($page > 1) {
- $search_url .= "&start=" . $page*$num; # page number
- $search_url .= "&num=" . $num; #images per page
- }
+ my $off = int(rand(40));
+ my $search_url = $google_images_url . $words . "&start=" . $off;
my ($search_hit_count, @subpages) =
pick_from_search_engine ($timeout, $search_url, $words);
if ($u =~ m@^/imgres\?imgurl=(.*?)&imgrefurl=(.*?)\&@) {
my $ref = $2;
my $img = $1;
- $img = "http://$img" unless ($img =~ m/^http:/i);
-
$ref = url_decode($ref);
$img = url_decode($img);
+ $img = "http://$img" unless ($img =~ m/^https?:/i);
+
LOG ($verbose_filter, " candidate: $ref");
push @candidates, $img;
$referers{$img} = $ref;
my $i = int(rand($#photomakers + 1));
my $fn = $photomakers[$i];
my $file = &$fn;
- my $words .= $file . "%20filetype:jpg";
+ #$file .= "%20filetype:jpg";
- pick_from_google_images ($timeout, $words);
+ pick_from_google_images ($timeout, $file);
}
+\f
+############################################################################
+#
+# Pick images by feeding random words into Google Image Search.
+# By the way: fuck Microsoft.
+#
+############################################################################
+
+my $bing_images_url = "http://www.bing.com/images/async?q=";
+
+
+# bingimgs
+sub pick_from_bing_images($;$$) {
+ my ($timeout, $words, $max_page) = @_;
+
+ if (!defined($words)) {
+ $words = random_word(); # only one word for Bing
+ }
+
+ my $off = int(rand(300));
+ my $search_url = $bing_images_url . $words . "&first=" . $off;
+
+ my ($search_hit_count, @subpages) =
+ pick_from_search_engine ($timeout, $search_url, $words);
+
+ my @candidates = ();
+ my %referers;
+ foreach my $u (@subpages) {
+ my ($img, $ref) = ($u =~ m/^(.*?)\t(.*)$/s);
+ next unless $img;
+ LOG ($verbose_filter, " candidate: $ref");
+ push @candidates, $img;
+ $referers{$img} = $ref;
+ }
+
+ @candidates = depoison (@candidates);
+ return () if ($#candidates < 0);
+ my $i = int(rand($#candidates+1));
+ my $img = $candidates[$i];
+ my $ref = $referers{$img};
+
+ LOG ($verbose_load, "picked image " . ($i+1) . ": $img (on $ref)");
+ return ($ref, $img);
+}
+
+
+
+\f
+############################################################################
+#
+# Pick images by feeding random numbers into Bing Image Search.
+#
+############################################################################
+
+# bingnums
+sub pick_from_bing_image_numbers($) {
+ my ($timeout) = @_;
+
+ my $max = 9999;
+ my $number = int(rand($max));
+
+ $number = sprintf("%04d", $number)
+ if (rand() < 0.3);
+
+ pick_from_bing_images ($timeout, "$number");
+}
+
+\f
+############################################################################
+#
+# Pick images by feeding random numbers into Bing Image Search.
+#
+############################################################################
+
+# bingphotos
+sub pick_from_bing_image_photos($) {
+ my ($timeout) = @_;
+
+ my $i = int(rand($#photomakers + 1));
+ my $fn = $photomakers[$i];
+ my $file = &$fn;
+
+ pick_from_bing_images ($timeout, $file);
+}
\f
############################################################################
#
############################################################################
-my $flickr_img_url = "http://www.flickr.com/photos/";
+my $flickr_img_url = "http://www.flickr.com/explore/";
# Like LiveJournal, the Flickr page of images tends to update slowly,
# so remember the last N entries on it and randomly select from those.
my $count = 0;
my $count2 = 0;
foreach (split (/\n/, $body)) {
+
my ($page, $thumb) = m@<A \s [^<>]* \b HREF=\"([^<>\"]+)\" [^<>]* > \s*
- <IMG \s [^<>]* \b SRC=\"([^<>\"]+)\" @xsi;
+ <IMG \s [^<>]* \b
+ data-defer-src = \"([^<>\"]+)\" @xsi;
next unless defined ($thumb);
$page = html_unquote ($page);
$thumb = html_unquote ($thumb);
- next unless ($thumb =~ m@^http://farm\d*\.static\.?flickr\.com/@);
+ next unless ($thumb =~ m@^https?://[^/.]+\d*\.static\.?flickr\.com/@);
my $base = "http://www.flickr.com/";
$page =~ s@^/@$base@;
return ($base, $img);
}
+\f
+############################################################################
+#
+# Pick random images from Instagram.
+#
+############################################################################
+
+my $instagram_url_base = "https://api.instagram.com/v1/media/popular";
+
+# instagram_random
+sub pick_from_instagram($) {
+ my $timeout = shift;
+
+ # Liberated access tokens.
+ # jsdo.it search for: instagram client_id
+ # Google search for: instagram "&client_id=" site:jsfiddle.net
+ my @tokens = ('b59fbe4563944b6c88cced13495c0f49', # gramfeed.com
+ 'fa26679250df49c48a33fbcf30aae989', # instac.at
+ 'd9494686198d4dfeb954979a3e270e5e', # iconosquare.com
+ '793ef48bb18e4197b61afce2d799b81c', # jsdo.it
+ '67b8a3e0073449bba70600d0fc68e6cb', # jsdo.it
+ '26a098e0df4d4b9ea8b4ce6c505b7742', # jsdo.it
+ '2437cbcd906a4c10940f990d283d3cd5', # jsdo.it
+ '191c7d7d5312464cbd92134f36ffdab5', # jsdo.it
+ 'acfec809437b4340b2c38f66503af774', # jsdo.it
+ 'e9f77604a3a24beba949c12d18130988', # jsdo.it
+ '2cd7bcf68ae346529770073d311575b3', # jsdo.it
+ '830c600fe8d742e2ab3f3b94f9bb22b7', # jsdo.it
+ '55865a0397ad41e5997dd95ef4df8da1', # jsdo.it
+ '192a5742f3644ea8bed1d25e439286a8', # jsdo.it
+ '38ed1477e7a44595861b8842cdb8ba23', # jsdo.it
+ 'e52f79f645f54488ad0cc47f6f55ade6', # jsfiddle.net
+ );
+
+ my $tok = $tokens[int(rand($#tokens+1))];
+ $last_search = $instagram_url_base . "?client_id=" . $tok;
+
+ print STDERR "\n\n" if ($verbose_load);
+ LOG ($verbose_load, "URL: $last_search");
+
+ my ( $base, $body ) = get_document ($last_search, undef, $timeout);
+ if (!$base || !$body) {
+ $body = undef;
+ return;
+ }
+
+ $body =~ s/("link")/\001$1/gs;
+ my @chunks = split(/\001/, $body);
+ shift @chunks;
+ my @urls = ();
+ foreach (@chunks) {
+ s/\\//gs;
+ my ($url) = m/"link":\s*"(.*?)"/s;
+ my ($img) = m/"standard_resolution":{"url":\s*"(.*?)"/s;
+ ($img) = m/"url":\s*"(.*?)"/s unless $url;
+ next unless ($url && $img);
+ push @urls, [ $url, $img ];
+ }
+
+ if ($#urls < 0) {
+ LOG ($verbose_load, "no images on $last_search");
+ return ();
+ }
+
+ my $i = int(rand($#urls+1));
+ my ($url, $img) = @{$urls[$i]};
+
+ LOG ($verbose_load, "picked image " .($i+1) . "/" . ($#urls+1) . ": $url");
+ return ($url, $img);
+}
+
\f
############################################################################
#
sub which($) {
my ($prog) = @_;
foreach (split (/:/, $ENV{PATH})) {
- if (-x "$_/$prog") {
- return $prog;
+ my $path = "$_/$prog";
+ if (-x $path) {
+ return $path;
}
}
return undef;
#
##############################################################################
-my $image_ppm = sprintf ("%s/webcollage-%08x.ppm",
- ($ENV{TMPDIR} ? $ENV{TMPDIR} : "/tmp"),
- rand(0xFFFFFFFF));
-my $image_tmp1 = sprintf ("%s/webcollage-1-%08x.ppm",
- ($ENV{TMPDIR} ? $ENV{TMPDIR} : "/tmp"),
- rand(0xFFFFFFFF));
-my $image_tmp2 = sprintf ("%s/webcollage-2-%08x.ppm",
- ($ENV{TMPDIR} ? $ENV{TMPDIR} : "/tmp"),
- rand(0xFFFFFFFF));
+my ($image_ppm, $image_tmp1, $image_tmp2);
+{
+ my $seed = rand(0xFFFFFFFF);
+ $image_ppm = sprintf ("%s/webcollage-%08x",
+ ($ENV{TMPDIR} ? $ENV{TMPDIR} : "/tmp"),
+ $seed);
+ $image_tmp1 = $image_ppm . '-1.ppm';
+ $image_tmp2 = $image_ppm . '-2.ppm';
+ $image_ppm .= '.ppm';
+}
+
my $filter_cmd = undef;
my $post_filter_cmd = undef;
sub x_or_pbm_output($) {
my ($window_id) = @_;
+ # Adjust the PATH for OS X 10.10.
+ #
+ $_ = $0;
+ s:/[^/]*$::;
+ s/([^a-zA-Z0-9._\-+\/])/\\$1/g;
+ $ENV{PATH} = "$_:$ENV{PATH}";
+
# Check for our helper program, to see whether we need to use PPM pipelines.
#
$_ = "webcollage-helper";
- if (defined ($webcollage_helper) || which ($_)) {
- $webcollage_helper = $_ unless (defined($webcollage_helper));
+
+ if (! defined ($webcollage_helper)) {
+ $webcollage_helper = which ($_);
+ }
+
+ if (defined ($webcollage_helper)) {
LOG ($verbose_pbm, "found \"$webcollage_helper\"");
$webcollage_helper .= " -v";
} else {
which ($_) || error "$_ not found on \$PATH.";
}
+ # If we're using webcollage-helper and not a filter, then the tmp files
+ # are JPEGs, not PPMs.
+ #
+ if (defined ($webcollage_helper) && !defined ($filter_cmd)) {
+ foreach ($image_ppm, $image_tmp1, $image_tmp2) {
+ s/\.ppm$/.jpg/s;
+ }
+ }
+
+
# find a root-window displayer program.
#
if (!$no_output_p) {
return 0;
}
+ if ($iw <= 0 || $ih <= 0 || $iw > 9999 || $ih > 9999) {
+ LOG (($verbose_pbm || $verbose_load),
+ "ludicrous image dimensions: $iw x $ih (" . length($body) .
+ "): $img");
+ $body = undef;
+ return 0;
+ }
+
open (my $out, '>', $image_tmp1) || error ("writing $image_tmp1: $!");
(print $out $body) || error ("writing $image_tmp1: $!");
close ($out) || error ("writing $image_tmp1: $!");
my $imagemap_html = $imagemap_base . ".html";
my $imagemap_jpg = $imagemap_base . ".jpg";
+ my $imagemap_jpg2 = $imagemap_jpg;
+ $imagemap_jpg2 =~ s@^.*/@@gs;
if (!defined ($imagemap_html_tmp)) {
$imagemap_html_tmp = $imagemap_html . sprintf (".%08x", rand(0xffffffff));
if ($template_html =~ m/^\s*$/s) {
$template_html = ("<MAP NAME=\"$map_name\"></MAP>\n" .
- "<IMG SRC=\"$imagemap_base.jpg\"" .
+ "<IMG SRC=\"$imagemap_jpg2\"" .
" USEMAP=\"$map_name\">\n");
LOG ($verbose_pbm, "created dummy template");
}
my $body = $template_html;
my $areas = join ("\n\t", @imagemap_areas);
my $map = ("<MAP NAME=\"$map_name\">\n\t$areas\n</MAP>");
- my $img = ("<IMG SRC=\"$imagemap_base.jpg\" " .
+ my $img = ("<IMG SRC=\"$imagemap_jpg2\" " .
"BORDER=0 " .
"WIDTH=$image_width HEIGHT=$image_height " .
"USEMAP=\"#$map_name\">");