#!/usr/bin/perl -w
-# Copyright © 2005, 2006 Jamie Zawinski <jwz@jwz.org>
+# Copyright © 2005-2013 Jamie Zawinski <jwz@jwz.org>
#
# Permission to use, copy, modify, distribute, and sell this software and its
# documentation for any purpose is hereby granted without fee, provided that
# Created: 19-Mar-2005.
require 5;
-use diagnostics;
+#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: 1.13 $ }; $version =~ s/^[^0-9]+([0-9.]+).*$/$1/;
+my $version = q{ $Revision: 1.31 $ }; $version =~ s/^[^0-9]+([0-9.]+).*$/$1/;
my $verbose = 0;
my $http_proxy = undef;
my $text_literal = '';
my $text_file = '';
my $text_program = '';
-my $text_url = '';
+my $text_url = 'http://en.wikipedia.org/w/index.php?title=Special:NewPages&feed=rss';
+# Default URL needs to be set and match what's in OSX/XScreenSaverView.m
my $wrap_columns = undef;
my $nyarlathotep_p = 0;
# Maps HTML character entities to the corresponding Latin1 characters.
#
my %entity_table = (
- "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',
- "apos" => '\''
+ "quot" => '"', "amp" => '&', "lt" => '<',
+ "gt" => '>', "nbsp" => " ", "iexcl" => "\xA1",
+ "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", "apos" => "\'",
+
+ # 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" => "\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"=> ">",
);
# Maps certain UTF8 characters (2 or 3 bytes) to the corresponding
# Latin1 characters.
#
my %unicode_latin1_table = (
- "\xC2\xA1" => '¡', "\xC2\xA2" => '¢', "\xC2\xA3" => '£', "\xC2\xA4" => '¤',
- "\xC2\xA5" => '¥', "\xC2\xA6" => '¦', "\xC2\xA7" => '§', "\xC2\xA8" => '¨',
- "\xC2\xA9" => '©', "\xC2\xAA" => 'ª', "\xC2\xAB" => '«', "\xC2\xAC" => '¬',
- "\xC2\xAD" => '', "\xC2\xAE" => '®', "\xC2\xAF" => '¯', "\xC2\xB0" => '°',
- "\xC2\xB1" => '±', "\xC2\xB2" => '²', "\xC2\xB3" => '³', "\xC2\xB4" => '´',
- "\xC2\xB5" => 'µ', "\xC2\xB6" => '¶', "\xC2\xB7" => '·', "\xC2\xB8" => '¸',
- "\xC2\xB9" => '¹', "\xC2\xBA" => 'º', "\xC2\xBB" => '»', "\xC2\xBC" => '¼',
- "\xC2\xBD" => '½', "\xC2\xBE" => '¾', "\xC2\xBF" => '¿', "\xC3\x80" => 'À',
- "\xC3\x81" => 'Á', "\xC3\x82" => 'Â', "\xC3\x83" => 'Ã', "\xC3\x84" => 'Ä',
- "\xC3\x85" => 'Å', "\xC3\x86" => 'Æ', "\xC3\x87" => 'Ç', "\xC3\x88" => 'È',
- "\xC3\x89" => 'É', "\xC3\x8A" => 'Ê', "\xC3\x8B" => 'Ë', "\xC3\x8C" => 'Ì',
- "\xC3\x8D" => 'Í', "\xC3\x8E" => 'Î', "\xC3\x8F" => 'Ï', "\xC3\x90" => 'Ð',
- "\xC3\x91" => 'Ñ', "\xC3\x92" => 'Ò', "\xC3\x93" => 'Ó', "\xC3\x94" => 'Ô',
- "\xC3\x95" => 'Õ', "\xC3\x96" => 'Ö', "\xC3\x97" => '×', "\xC3\x98" => 'Ø',
- "\xC3\x99" => 'Ù', "\xC3\x9A" => 'Ú', "\xC3\x9B" => 'Û', "\xC3\x9C" => 'Ü',
- "\xC3\x9D" => 'Ý', "\xC3\x9E" => 'Þ', "\xC3\x9F" => 'ß', "\xC3\xA0" => 'à',
- "\xC3\xA1" => 'á', "\xC3\xA2" => 'â', "\xC3\xA3" => 'ã', "\xC3\xA4" => 'ä',
- "\xC3\xA5" => 'å', "\xC3\xA6" => 'æ', "\xC3\xA7" => 'ç', "\xC3\xA8" => 'è',
- "\xC3\xA9" => 'é', "\xC3\xAA" => 'ê', "\xC3\xAB" => 'ë', "\xC3\xAC" => 'ì',
- "\xC3\xAD" => 'í', "\xC3\xAE" => 'î', "\xC3\xAF" => 'ï', "\xC3\xB0" => 'ð',
- "\xC3\xB1" => 'ñ', "\xC3\xB2" => 'ò', "\xC3\xB3" => 'ó', "\xC3\xB4" => 'ô',
- "\xC3\xB5" => 'õ', "\xC3\xB6" => 'ö', "\xC3\xB7" => '÷', "\xC3\xB8" => 'ø',
- "\xC3\xB9" => 'ù', "\xC3\xBA" => 'ú', "\xC3\xBB" => 'û', "\xC3\xBC" => 'ü',
- "\xC3\xBD" => 'ý', "\xC3\xBE" => 'þ', "\xC3\xBF" => 'ÿ',
-
- "\xE2\x80\x93" => '--', "\xE2\x80\x94" => '--',
- "\xE2\x80\x98" => '`', "\xE2\x80\x99" => '\'',
- "\xE2\x80\x9C" => "``", "\xE2\x80\x9D" => "''",
- "\xE2\x80\xA6" => '...',
-);
-
+ "\xC2\xA1" => "\xA1", "\xC2\xA2" => "\xA2", "\xC2\xA3" => "\xA3",
+ "\xC2\xA4" => "\xA4", "\xC2\xA5" => "\xA5", "\xC2\xA6" => "\xA6",
+ "\xC2\xA7" => "\xA7", "\xC2\xA8" => "\xA8", "\xC2\xA9" => "\xA9",
+ "\xC2\xAA" => "\xAA", "\xC2\xAB" => "\xAB", "\xC2\xAC" => "\xAC",
+ "\xC2\xAD" => "\xAD", "\xC2\xAE" => "\xAE", "\xC2\xAF" => "\xAF",
+ "\xC2\xB0" => "\xB0", "\xC2\xB1" => "\xB1", "\xC2\xB2" => "\xB2",
+ "\xC2\xB3" => "\xB3", "\xC2\xB4" => "\xB4", "\xC2\xB5" => "\xB5",
+ "\xC2\xB6" => "\xB6", "\xC2\xB7" => "\xB7", "\xC2\xB8" => "\xB8",
+ "\xC2\xB9" => "\xB9", "\xC2\xBA" => "\xBA", "\xC2\xBB" => "\xBB",
+ "\xC2\xBC" => "\xBC", "\xC2\xBD" => "\xBD", "\xC2\xBE" => "\xBE",
+ "\xC2\xBF" => "\xBF", "\xC3\x80" => "\xC0", "\xC3\x81" => "\xC1",
+ "\xC3\x82" => "\xC2", "\xC3\x83" => "\xC3", "\xC3\x84" => "\xC4",
+ "\xC3\x85" => "\xC5", "\xC3\x86" => "\xC6", "\xC3\x87" => "\xC7",
+ "\xC3\x88" => "\xC8", "\xC3\x89" => "\xC9", "\xC3\x8A" => "\xCA",
+ "\xC3\x8B" => "\xCB", "\xC3\x8C" => "\xCC", "\xC3\x8D" => "\xCD",
+ "\xC3\x8E" => "\xCE", "\xC3\x8F" => "\xCF", "\xC3\x90" => "\xD0",
+ "\xC3\x91" => "\xD1", "\xC3\x92" => "\xD2", "\xC3\x93" => "\xD3",
+ "\xC3\x94" => "\xD4", "\xC3\x95" => "\xD5", "\xC3\x96" => "\xD6",
+ "\xC3\x97" => "\xD7", "\xC3\x98" => "\xD8", "\xC3\x99" => "\xD9",
+ "\xC3\x9A" => "\xDA", "\xC3\x9B" => "\xDB", "\xC3\x9C" => "\xDC",
+ "\xC3\x9D" => "\xDD", "\xC3\x9E" => "\xDE", "\xC3\x9F" => "\xDF",
+ "\xC3\xA0" => "\xE0", "\xC3\xA1" => "\xE1", "\xC3\xA2" => "\xE2",
+ "\xC3\xA3" => "\xE3", "\xC3\xA4" => "\xE4", "\xC3\xA5" => "\xE5",
+ "\xC3\xA6" => "\xE6", "\xC3\xA7" => "\xE7", "\xC3\xA8" => "\xE8",
+ "\xC3\xA9" => "\xE9", "\xC3\xAA" => "\xEA", "\xC3\xAB" => "\xEB",
+ "\xC3\xAC" => "\xEC", "\xC3\xAD" => "\xED", "\xC3\xAE" => "\xEE",
+ "\xC3\xAF" => "\xEF", "\xC3\xB0" => "\xF0", "\xC3\xB1" => "\xF1",
+ "\xC3\xB2" => "\xF2", "\xC3\xB3" => "\xF3", "\xC3\xB4" => "\xF4",
+ "\xC3\xB5" => "\xF5", "\xC3\xB6" => "\xF6", "\xC3\xB7" => "\xF7",
+ "\xC3\xB8" => "\xF8", "\xC3\xB9" => "\xF9", "\xC3\xBA" => "\xFA",
+ "\xC3\xBB" => "\xFB", "\xC3\xBC" => "\xFC", "\xC3\xBD" => "\xFD",
+ "\xC3\xBE" => "\xFE", "\xC3\xBF" => "\xFF",
+
+ "\xE2\x80\x93" => '--', "\xE2\x80\x94" => '--',
+ "\xE2\x80\x98" => '`', "\xE2\x80\x99" => '\'',
+ "\xE2\x80\x9C" => "``", "\xE2\x80\x9D" => "''",
+ "\xE2\x80\xB2" => "'", "\xE2\x80\xA6" => '...',
+ "\xE2\x86\x90" => ' ← ', "\xE2\x84\xA2" => '™',
+ "\xE2\x80\xA2" => '•', "\xC2\xA0" => ' ', #
+ );
# Convert any HTML entities to Latin1 characters.
#
my ($text) = @_;
$text =~ s/(&(\#)?([[:alpha:]\d]+);?)/
{
- my $c;
- if ($2) {
- $c = chr($3); # the &#number is always decimal, right?
+ my $c = $3;
+ if (! defined($2)) {
+ $c = $entity_table{$c}; # for Á
} else {
- $c = $entity_table{$3};
+ if ($c =~ m@^x([\dA-F]+)$@si) { # for A
+ $c = chr(hex($1));
+ } elsif ($c =~ m@^\d+$@si) { # for A
+ $c = chr($c);
+ } else {
+ $c = undef;
+ }
}
-# print STDERR "$progname: warning: unknown HTML character entity \"$1\"\n"
-# unless $c;
- ($c ? $c : "[$3]");
+ ($c || "[$3]"); # for &unknown; => "[unknown]"
}
/gexi;
return $text;
#
sub de_unicoddle($) {
my ($text) = @_;
- foreach my $key (keys (%unicode_latin1_table)) {
+ foreach my $key (sort { length($b) <=> length($a) }
+ keys (%unicode_latin1_table)) {
my $val = $unicode_latin1_table{$key};
$text =~ s/$key/$val/gs;
}
#
sub get_x11_prefs() {
my $got_any_p = 0;
- local *IN;
- if (open (IN, "<$config_file")) {
+ if (open (my $in, '<', $config_file)) {
print STDERR "$progname: reading $config_file\n" if ($verbose > 1);
- my $body = '';
- while (<IN>) { $body .= $_; }
- close IN;
+ local $/ = undef; # read entire file
+ my $body = <$in>;
+ close $in;
$got_any_p = get_x11_prefs_1 ($body);
} elsif ($verbose > 1) {
}
if ($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;
+ $text_literal =~ s@\\\n@\n@gs;
}
elsif ($text_mode eq '1') { $text_mode = 'literal'; }
elsif ($text_mode eq '2') { $text_mode = 'file'; }
elsif ($text_mode eq '3') { $text_mode = 'url'; }
+ elsif ($text_mode eq '4') { $text_mode = 'program'; }
$v = get_cocoa_pref_1 ($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);
$text_file =~ s@^~/@$ENV{HOME}/@s; # allow literal "~/"
- local *IN;
- if (open (IN, "<$text_file")) {
+ if (open (my $in, '<', $text_file)) {
print STDERR "$progname: reading $text_file\n" if ($verbose);
if ($wrap_columns && $wrap_columns > 0) {
# read it, then reformat it.
- my $body = '';
- while (<IN>) { $body .= $_; }
+ local $/ = undef; # read entire file
+ my $body = <$in>;
reformat_text ($body);
} else {
- # stream it
- while (<IN>) {
+ # stream it by lines
+ while (<$in>) {
y/A-Za-z/N-ZA-Mn-za-m/ if ($nyarlathotep_p);
- print $_;
+ print STDOUT $_;
}
}
- close IN;
+ close $in;
} else {
error ("$text_file: $!");
}
} else { # $text_mode eq 'date'
- safe_system ("uname", "-n");
+ my $n = `uname -n`;
+ $n =~ s/\.local\n/\n/s;
+ print $n;
my $unamep = 1;
if (-f "/etc/redhat-release") { # "Fedora Core release 4 (Stentz)"
- system ("cat", "/etc/redhat-release");
+ safe_system ("cat", "/etc/redhat-release");
+ }
+
+ if (-f "/etc/release") { # "Solaris 10 3/05 s10_74L2a X86"
+ safe_system ("head", "-1", "/etc/release");
}
if (-f "/usr/sbin/system_profiler") { # "Mac OS X 10.4.5 (8H14)"
my $sp = # "iMac G5"
`/usr/sbin/system_profiler SPSoftwareDataType SPHardwareDataType`;
my ($v) = ($sp =~ m/^\s*System Version:\s*(.*)$/mi);
- my ($s) = ($sp =~ m/^\s*CPU Speed:\s*(.*)$/mi);
- my ($t) = ($sp =~ m/^\s*Machine Name:\s*(.*)$/mi);
+ my ($s) = ($sp =~ m/^\s*(?:CPU|Processor) Speed:\s*(.*)$/mi);
+ my ($t) = ($sp =~ m/^\s*(?:Machine|Model) Name:\s*(.*)$/mi);
print "$v\n" if ($v);
print "$s $t\n" if ($s && $t);
$unamep = !defined ($v);
}
-# Loads the given URL, returns: $http, $head, $body.
-#
-sub get_url_1($;$) {
- my ($url, $referer) = @_;
-
- if (! ($url =~ m@^(http|feed)://@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 = <S> || "";
-
- $_ = $http;
- s/[\r\n]+$//s;
- print STDERR " <== $_\n" if ($verbose > 3);
-
- while (<S>) {
- $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>) {
- 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';
return 'text';
}
+
sub reformat_html($$) {
my ($body, $rss_p) = @_;
$_ = $body;
+ # In HTML, try to preserve newlines inside of PRE.
+ #
+ if (! $rss_p) {
+ s@(<PRE\b[^<>]*>\s*)(.*?)(</PRE)@{
+ my ($a, $b, $c) = ($1, $2, $3);
+ $b =~ s/[\r\n]/<BR>/gs;
+ $a . $b . $c;
+ }@gsexi;
+ }
+
if (! $rss_p) {
- # In HTML, unfold lines (this breaks PRE. Sue me.)
+ # In HTML, unfold lines.
# In RSS, assume \n means literal line break.
s@[\r\n]@ @gsi;
}
s@<[^<>]*>?@@gs; # lose all other HTML tags
$_ = de_entify ($_); # convert HTML entities
+ # For Wikipedia: delete anything inside {{ }} and unwrap [[tags]],
+ # among other things.
+ #
+ if ($rss_p eq 'wiki') {
+
+ # Creation line is often truncated: screws up parsing with unbalanced {{.
+ s@(: +<- +Created page) with [^\n]+@$1@s;
+
+ s@/\*.*?\*/@@si; # /* ... */
+
+ # Try to omit all tables, since they're impossible to read as text.
+ #
+ 1 while (s/{{[^{}]*}}/ /gs); # {{ ... }}
+ 1 while (s/{\|.*?\|}/\n\n/gs); # {| ... |}
+ 1 while (s/\|-.*?\|/ /gs); # |- ... | (table cell)
+
+ # Convert anchors to something more readable.
+ #
+ s/\[\[([^\[\]\|]+)\|([^\[\]]+)\]\]/$2/gs; # [[link|anchor]]
+ s/\[\[([^:\[\]\|]+)\]\]/$1/gs; # [[anchor]]
+ s/\[https?:[^\[\]\s]+\s+([^\[\]]+)\]/$1/gs; # [url anchor]
+
+ # Convert all references to asterisks.
+ s@\s*<ref>\s*.*?</ref>@*@gs; # <ref> ... <ref> -> "*"
+ s@\n[ \t]*\d+\s*\^\s*http[^\s]+[ \t]*\n@\n@gs; # 1 ^ URL (a Reflist)
+
+ s@\[\[File:([^\|\]]+).*?\]\]@\n$1\n@gs; # [[File: X | ... ]]
+ s@\[\[Category:.*?\]\]@@gs; # omit categories
+
+ s/<[^<>]*>//gs; # Omit all remaining tags
+ s/\'{3,}//gs; # Omit ''' and ''''
+ s/\'\'/\"/gs; # '' -> "
+ s/\`\`/\"/gs; # `` -> "
+ s/\"\"+/\"/gs; # "" -> "
+
+ s/^[ \t]*[*#]+[ \t]*$//gm; # Omit lines with just * or # on them
+
+ # Omit trailing headlines with no text after them (e.g. == Notes ==)
+ 1 while (s/\n==+[ \t]*[^\n=]+[ \t]*==+\s*$/\n/s);
+ }
+
+
# 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;
+ s/([^\000-\176]+(\s*[^\000-\176]+)[^a-z\d]*)/\xAB...\xBB /g;
+# s/([^\000-\176]+(\s*[^\000-\176]+)[^a-z\d]*)/\xAB$1\xBB /g;
$_ .= "\n";
if (!defined($wrap_columns) || $wrap_columns > 0) {
$Text::Wrap::columns = ($wrap_columns || 72);
- $_ = wrap ("", " ", $_); # wrap the lines as a paragraph
- s/[ \t]+$//gm; # lose whitespace at end of line again
+ $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
}
+ s/^\n+//gs;
+
y/A-Za-z/N-ZA-Mn-za-m/ if ($nyarlathotep_p);
print STDOUT $_;
}
sub reformat_rss($) {
my ($body) = @_;
+ my $wiki_p = ($body =~ m@<generator>[^<>]*Wiki@si);
+
$body =~ s/(<(ITEM|ENTRY)\b)/\001\001$1/gsi;
my @items = split (/\001\001/, $body);
$title = rss_field_to_html ($title || '');
$body1 = rss_field_to_html ($body1 || '');
- reformat_html ("$title<P>$body1", 1);
+ $title = '' if ($body1 eq $title); # Identical in Twitter's atom feed.
+
+ reformat_html ("$title<P>$body1", $wiki_p ? 'wiki' : 'rss');
print "\n";
}
}
if ($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;
}
}
+# Figure out what the proxy server should be, either from environment
+# variables or by parsing the output of the (MacOS) program "scutil",
+# which tells us what the system-wide proxy settings are.
+#
+sub set_proxy($) {
+ my ($ua) = @_;
+
+ 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) = ($proxy_data =~ m/\bHTTPPort\s*:\s*([^\s]+)/s);
+ if ($server) {
+ # Note: this ignores the "ExceptionsList".
+ $ENV{http_proxy} = "http://" . $server . ($port ? ":$port" : "") . "/";
+ print STDERR "$progname: MacOS proxy: $ENV{http_proxy}\n"
+ if ($verbose > 2)
+ }
+ }
+
+ $ua->env_proxy();
+}
+
+
sub get_url_text($) {
my ($url) = @_;
- # historical suckage: the environment variable name is lower case.
- $http_proxy = $ENV{http_proxy} || $ENV{HTTP_PROXY};
+ my $ua = eval 'LWP::UserAgent->new';
- if ($http_proxy && $http_proxy =~ m@^http://([^/]*)/?$@ ) {
- # historical suckage: allow "http://host:port" as well as "host:port".
- $http_proxy = $1;
+ 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') {
elsif (m/^--?date$/) { $text_mode = 'date';
$load_p = 0; }
elsif (m/^--?text$/) { $text_mode = 'literal';
- $text_literal = shift @ARGV;
+ $text_literal = shift @ARGV || '';
$load_p = 0; }
elsif (m/^--?file$/) { $text_mode = 'file';
- $text_file = shift @ARGV;
+ $text_file = shift @ARGV || '';
$load_p = 0; }
elsif (m/^--?program$/) { $text_mode = 'program';
- $text_program = shift @ARGV;
+ $text_program = shift @ARGV || '';
$load_p = 0; }
elsif (m/^--?url$/) { $text_mode = 'url';
- $text_url = shift @ARGV;
+ $text_url = shift @ARGV || '';
$load_p = 0; }
elsif (m/^--?col(umn)?s?$/) { $wrap_columns = 0 + shift @ARGV; }
elsif (m/^--?cocoa$/) { $cocoa_id = shift @ARGV; }
}
output();
+
+
+ if (defined ($cocoa_id)) {
+ #
+ # On MacOS, sleep for 10 seconds between when the last output is
+ # printed, and when this process exits. This is because MacOS
+ # 10.5.0 and later broke ptys in a new and exciting way: basically,
+ # once the process at the end of the pty exits, you have exactly
+ # 1 second to read all the queued data off the pipe before it is
+ # summarily flushed.
+ #
+ # Many of the screen savers were written to depend on being able
+ # to read a small number of bytes, and continue reading until they
+ # reached EOF. This is no longer possible.
+ #
+ # Note that the current MacOS behavior has all four of these
+ # awesome properties: 1) Inconvenient; 2) Has no sane workaround;
+ # 3) Different behavior than MacOS 10.1 through 10.4; and 4)
+ # Different behavior than every other Unix in the world.
+ #
+ # See http://jwz.org/b/DHke, and for those of you inside Apple,
+ # "Problem ID 5606018".
+ #
+ # One workaround would be to rewrite the savers to have an
+ # internal buffer, and always read as much data as possible as
+ # soon as a pipe has input available. However, that's a lot more
+ # work, so instead, let's just not exit right away, and hope that
+ # 10 seconds is enough.
+ #
+ # This will solve the problem for invocations of xscreensaver-text
+ # that produce little output (e.g., date-mode); and won't solve it
+ # in cases where a large amount of text is generated in a short
+ # amount of time (e.g., url-mode.)
+ #
+ sleep (10);
+ }
}
main();