X-Git-Url: http://git.hungrycats.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=driver%2Fxscreensaver-text;h=e4c588c50cf41cd929f032ea11b34343c8cdae83;hb=ff35d056d723c9a5ffe728dbba5f1c25e141be04;hp=1f411851070f56173c581398397d038482f9293e;hpb=ec8d2b32b63649e6d32bdfb306eda062769af823;p=xscreensaver diff --git a/driver/xscreensaver-text b/driver/xscreensaver-text index 1f411851..e4c588c5 100755 --- a/driver/xscreensaver-text +++ b/driver/xscreensaver-text @@ -1,5 +1,5 @@ #!/usr/bin/perl -w -# Copyright © 2005-2011 Jamie Zawinski +# Copyright © 2005-2013 Jamie Zawinski # # Permission to use, copy, modify, distribute, and sell this software and its # documentation for any purpose is hereby granted without fee, provided that @@ -24,13 +24,17 @@ require 5; #use diagnostics; # Fails on some MacOS 10.5 systems use strict; +# Some Linux systems don't install LWP by default! +# Only error out if we're actually loading a URL instead of local data. +BEGIN { eval 'use LWP::UserAgent;' } + use Socket; use POSIX qw(strftime); use Text::Wrap qw(wrap); use bytes; my $progname = $0; $progname =~ s@.*/@@g; -my $version = q{ $Revision: 1.24 $ }; $version =~ s/^[^0-9]+([0-9.]+).*$/$1/; +my $version = q{ $Revision: 1.29 $ }; $version =~ s/^[^0-9]+([0-9.]+).*$/$1/; my $verbose = 0; my $http_proxy = undef; @@ -40,7 +44,8 @@ my $text_mode = 'date'; my $text_literal = ''; my $text_file = ''; my $text_program = ''; -my $text_url = ''; +my $text_url = 'http://twitter.com/statuses/public_timeline.atom'; +# Default URL needs to be set and match what's in OSX/XScreenSaverView.m my $wrap_columns = undef; my $nyarlathotep_p = 0; @@ -192,15 +197,16 @@ sub get_x11_prefs() { } if ($verbose > 1) { - printf STDERR "$progname: mode: $text_mode\n"; - printf STDERR "$progname: literal: $text_literal\n"; - printf STDERR "$progname: file: $text_file\n"; - printf STDERR "$progname: program: $text_program\n"; - printf STDERR "$progname: url: $text_url\n"; + print STDERR "$progname: mode: $text_mode\n"; + print STDERR "$progname: literal: $text_literal\n"; + print STDERR "$progname: file: $text_file\n"; + print STDERR "$progname: program: $text_program\n"; + print STDERR "$progname: url: $text_url\n"; } $text_mode =~ tr/A-Z/a-z/; $text_literal =~ s@\\n@\n@gs; + $text_literal =~ s@\\\n@\n@gs; } @@ -251,6 +257,8 @@ sub get_cocoa_prefs($) { $v = get_cocoa_pref_1 ($id, "textLiteral"); $text_literal = $v if defined ($v); + $text_literal =~ s@\\n@\n@gs; + $text_literal =~ s@\\\n@\n@gs; $v = get_cocoa_pref_1 ($id, "textFile"); $text_file = $v if defined ($v); @@ -426,167 +434,6 @@ sub output() { } -# Loads the given URL, returns: $http, $head, $body. -# -sub get_url_1($;$) { - my ($url, $referer) = @_; - - $url =~ s@^feed:@http:@si; - if (! ($url =~ m@^http://@i)) { - error ("not an HTTP URL: $url"); - } - - my ($url_proto, $dummy, $serverstring, $path) = split(/\//, $url, 4); - $path = "" unless $path; - - my ($them,$port) = split(/:/, $serverstring); - $port = 80 unless $port; - - 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 $port2; - } - - my ($remote, $iaddr, $paddr, $proto, $line); - $remote = $them2; - if ($port2 =~ /\D/) { $port2 = getservbyname($port2, 'tcp') } - if (!$port2) { - error ("unrecognised port in $url"); - } - - $iaddr = inet_aton($remote); - error ("host not found: $remote") unless ($iaddr); - - $paddr = sockaddr_in($port2, $iaddr); - - - my $head = ""; - my $body = ""; - - $proto = getprotobyname('tcp'); - if (!socket(S, PF_INET, SOCK_STREAM, $proto)) { - error ("socket: $!"); - } - if (!connect(S, $paddr)) { - error ("connect($serverstring): $!"); - } - - select(S); $| = 1; select(STDOUT); - - my $user_agent = "$progname/$version"; - - 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"; - } - $hdrs .= "\r\n"; - - if ($verbose > 3) { - foreach (split('\r?\n', $hdrs)) { - print STDERR " ==> $_\n"; - } - } - print S $hdrs; - my $http = || ""; - - $_ = $http; - s/[\r\n]+$//s; - print STDERR " <== $_\n" if ($verbose > 3); - - while () { - $head .= $_; - s/[\r\n]+$//s; - last if m@^$@; - print STDERR " <== $_\n" if ($verbose > 3); - } - - print STDERR " <== \n" if ($verbose > 4); - my $lines = 0; - while () { - s/\r\n/\n/gs; - print STDERR " <== $_" if ($verbose > 4); - $body .= $_; - $lines++; - } - - print STDERR " <== [ body ]: $lines lines, " . length($body) . " bytes\n" - if ($verbose == 4); - - close S; - - if (!$http) { - error ("null response: $url"); - } - - return ( $http, $head, $body ); -} - - -# Loads the given URL, processes redirects, returns (content-type, body). -# -sub get_url($;$) { - my ($url, $referer) = @_; - - print STDERR "$progname: loading $url\n" if ($verbose > 2); - - my $orig_url = $url; - my $loop_count = 0; - my $max_loop_count = 10; - - do { - my ( $http, $head, $body ) = get_url_1 ($url, $referer); - - $http =~ s/[\r\n]+$//s; - - if ( $http =~ m@^HTTP/[0-9.]+ 30[123]@ ) { - $_ = $head; - - my ( $location ) = m@^location:[ \t]*(.*)$@im; - if ( $location ) { - $location =~ s/[\r\n]$//; - - print STDERR "$progname: redirect from $url to $location\n" - if ($verbose > 3); - - $referer = $url; - $url = $location; - - if ($url =~ m@^/@) { - $referer =~ m@^(http://[^/]+)@i; - $url = $1 . $url; - } elsif (! ($url =~ m@^[a-z]+:@i)) { - $_ = $referer; - s@[^/]+$@@g if m@^http://[^/]+/@i; - $_ .= "/" if m@^http://[^/]+$@i; - $url = $_ . $url; - } - - } else { - error ("no Location with \"$http\""); - } - - if ($loop_count++ > $max_loop_count) { - error ("too many redirects ($max_loop_count) from $orig_url"); - } - - } elsif ( $http =~ m@^HTTP/[0-9.]+ ([4-9][0-9][0-9].*)$@ ) { - error ("failed: $1 ($url)"); - - } else { - my $ct = 'text/plain'; - $ct = $1 if ($head =~ m/^content-type:\s*([^\s]+)/mi); - return ($ct, $body); - } - } while (1); -} - - # Make an educated guess as to what's in this document. # We don't necessarily take the Content-Type header at face value. # Returns 'html', 'rss', or 'text'; @@ -614,6 +461,7 @@ sub guess_content_type($$) { return 'text'; } + sub reformat_html($$) { my ($body, $rss_p) = @_; $_ = $body; @@ -647,6 +495,19 @@ sub reformat_html($$) { s@<[^<>]*>?@@gs; # lose all other HTML tags $_ = de_entify ($_); # convert HTML entities + # For Wikipedia: delete anything inside {{ }} and unwrap [[tags]] + # + if ($rss_p eq 'wiki') { + s@/\*.*?\*/@@si; # /* ... */ + 1 while (s/{{[^{}]*}}//gs); # {{ ... }} + s/\[\[([^:\[\]\|]+)\|([^\[\]]+)\]\]/$2/gs; # [[link|anchor]] + s/\[\[([^:\[\]\|]+)\]\]/$1/gs; # [[anchor]] + s/\[http:[^\[\]\s]+\s+([^\[\]]+)\]/$1/gs; # [url anchor] +# s@\s*.*?@*@gs; # url -> "*" + s/<[^<>]*>//gs; # -- omit. + } + + # elide any remaining non-Latin1 binary data... s/([\177-\377]+(\s*[\177-\377]+)[^a-z\d]*)/«...» /g; #s/([\177-\377]+(\s*[\177-\377]+)[^a-z\d]*)/«$1» /g; @@ -658,6 +519,7 @@ sub reformat_html($$) { if (!defined($wrap_columns) || $wrap_columns > 0) { $Text::Wrap::columns = ($wrap_columns || 72); + $Text::Wrap::break = '[\s/]'; # wrap on slashes for URLs $_ = wrap ("", " ", $_); # wrap the lines as a paragraph s/[ \t]+$//gm; # lose whitespace at end of line again } @@ -672,6 +534,8 @@ sub reformat_html($$) { sub reformat_rss($) { my ($body) = @_; + my $wiki_p = ($body =~ m@[^<>]*Wiki@si); + $body =~ s/(<(ITEM|ENTRY)\b)/\001\001$1/gsi; my @items = split (/\001\001/, $body); @@ -730,7 +594,7 @@ sub reformat_rss($) { $title = '' if ($body1 eq $title); # Identical in Twitter's atom feed. - reformat_html ("$title

$body1", 1); + reformat_html ("$title

$body1", $wiki_p ? 'wiki' : 'rss'); print "\n"; } } @@ -760,6 +624,7 @@ sub reformat_text($) { if ($wrap_columns && $wrap_columns > 0) { print STDERR "$progname: wrapping at $wrap_columns...\n" if ($verbose > 2); $Text::Wrap::columns = $wrap_columns; + $Text::Wrap::break = '[\s/]'; # wrap on slashes for URLs $body = wrap ("", "", $body); $body =~ s/[ \t]+$//gm; } @@ -773,39 +638,54 @@ sub reformat_text($) { # variables or by parsing the output of the (MacOS) program "scutil", # which tells us what the system-wide proxy settings are. # -sub set_proxy() { - - # historical suckage: the environment variable name is lower case. - $http_proxy = $ENV{http_proxy} || $ENV{HTTP_PROXY}; - - if (defined ($http_proxy)) { - if ($http_proxy && $http_proxy =~ m@^http://([^/]*)/?$@ ) { - # historical suckage: allow "http://host:port" as well as "host:port". - $http_proxy = $1; - } +sub set_proxy($) { + my ($ua) = @_; - } else { + if (!defined($ENV{http_proxy}) && !defined($ENV{HTTP_PROXY})) { my $proxy_data = `scutil --proxy 2>/dev/null`; my ($server) = ($proxy_data =~ m/\bHTTPProxy\s*:\s*([^\s]+)/s); my ($port) = ($proxy_data =~ m/\bHTTPPort\s*:\s*([^\s]+)/s); - # Note: this ignores the "ExceptionsList". if ($server) { - $http_proxy = $server; - $http_proxy .= ":$port" if $port; - } + # Note: this ignores the "ExceptionsList". + $ENV{http_proxy} = "http://" . $server . ($port ? ":$port" : "") . "/"; + print STDERR "$progname: MacOS proxy: $ENV{http_proxy}\n" + if ($verbose > 2) + } } - print STDERR "$progname: proxy server: $http_proxy\n" - if ($verbose > 2 && $http_proxy); + $ua->env_proxy(); } sub get_url_text($) { my ($url) = @_; - set_proxy(); + my $ua = eval 'LWP::UserAgent->new'; + + if (! $ua) { + print STDOUT ("\n\tPerl is broken. Do this to repair it:\n" . + "\n\tsudo cpan LWP::UserAgent\n\n"); + return; + } - my ($ct, $body) = get_url ($url); + set_proxy ($ua); + $ua->agent ("$progname/$version"); + my $res = $ua->get ($url); + my $body; + my $ct; + + if ($res && $res->is_success) { + $body = $res->decoded_content || ''; + $ct = $res->header ('Content-Type') || 'text/plain'; + + } else { + my $err = ($res ? $res->status_line : '') || ''; + $err = 'unknown error' unless $err; + $err = "$url: $err"; + # error ($err); + $body = "Error loading URL $err\n\n"; + $ct = 'text/plain'; + } $ct = guess_content_type ($ct, $body); if ($ct eq 'html') {