X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FHval.pm;h=2e883f810023f2a0bb3314e418437cca469365e0;hb=95bdac7f09c69036efed537a4d03d5bdd2ae4eb6;hp=652aef3daee55ab926e31262e0b722b1efbe568e;hpb=c952a7208d9941495a9efbe83d0e4af866f05a64;p=public-inbox.git

diff --git a/lib/PublicInbox/Hval.pm b/lib/PublicInbox/Hval.pm
index 652aef3d..2e883f81 100644
--- a/lib/PublicInbox/Hval.pm
+++ b/lib/PublicInbox/Hval.pm
@@ -1,5 +1,5 @@
-# Copyright (C) 2014-2015 all contributors <meta@public-inbox.org>
-# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
+# Copyright (C) 2014-2020 all contributors <meta@public-inbox.org>
+# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 #
 # represents a header value in various forms.  Used for HTML generation
 # in our web interface(s)
@@ -7,18 +7,15 @@ package PublicInbox::Hval;
 use strict;
 use warnings;
 use Encode qw(find_encoding);
-use URI::Escape qw(uri_escape_utf8);
-use PublicInbox::MID qw/mid_clean/;
+use PublicInbox::MID qw/mid_clean mid_escape/;
 use base qw/Exporter/;
-our @EXPORT_OK = qw/ascii_html/;
-
-# for user-generated content (UGC) which may have excessively long lines
-# and screw up rendering on some browsers.  This is the only CSS style
-# feature we use.
-use constant STYLE => '<style>pre{white-space:pre-wrap}</style>';
-
+our @EXPORT_OK = qw/ascii_html obfuscate_addrs to_filename src_escape
+		to_attr prurl/;
 my $enc_ascii = find_encoding('us-ascii');
 
+# safe-ish acceptable filename pattern for portability
+our $FN = '[a-zA-Z0-9][a-zA-Z0-9_\-\.]+[a-zA-Z0-9]'; # needs \z anchor
+
 sub new {
 	my ($class, $raw, $href) = @_;
 
@@ -32,17 +29,23 @@ sub new {
 
 sub new_msgid {
 	my ($class, $msgid) = @_;
-	$msgid = mid_clean($msgid);
-	$class->new($msgid, $msgid);
+	$class->new($msgid, mid_escape($msgid));
 }
 
-sub new_oneline {
-	my ($class, $raw) = @_;
-	$raw = '' unless defined $raw;
-	$raw =~ tr/\t\n / /s; # squeeze spaces
-	$raw =~ tr/\r//d; # kill CR
-	$class->new($raw);
-}
+# some of these overrides are standard C escapes so they're
+# easy-to-understand when rendered.
+my %escape_sequence = (
+	"\x00" => '\\0', # NUL
+	"\x07" => '\\a', # bell
+	"\x08" => '\\b', # backspace
+	"\x09" => "\t", # obvious to show as-is
+	"\x0a" => "\n", # obvious to show as-is
+	"\x0b" => '\\v', # vertical tab
+	"\x0c" => '\\f', # form feed
+	"\x0d" => '\\r', # carriage ret (not preceding \n)
+	"\x1b" => '^[', # ASCII escape (mutt seems to escape this way)
+	"\x7f" => '\\x7f', # DEL
+);
 
 my %xhtml_map = (
 	'"' => '&#34;',
@@ -52,15 +55,26 @@ my %xhtml_map = (
 	'>' => '&gt;',
 );
 
+$xhtml_map{chr($_)} = sprintf('\\x%02x', $_) for (0..31);
+%xhtml_map = (%xhtml_map, %escape_sequence);
+
+# for post-processing the output of highlight.pm and perhaps other
+# highlighers in the future
+sub src_escape ($) {
+	$_[0] =~ s/\r\n/\n/sg;
+	$_[0] =~ s/&apos;/&#39;/sg; # workaround https://bugs.debian.org/927409
+	$_[0] =~ s/([\x7f\x00-\x1f])/$xhtml_map{$1}/sge;
+	$_[0] = $enc_ascii->encode($_[0], Encode::HTMLCREF);
+}
+
 sub ascii_html {
 	my ($s) = @_;
 	$s =~ s/\r\n/\n/sg; # fixup bad line endings
-	$s =~ s/([<>&'"])/$xhtml_map{$1}/ge;
+	$s =~ s/([<>&'"\x7f\x00-\x1f])/$xhtml_map{$1}/sge;
 	$enc_ascii->encode($s, Encode::HTMLCREF);
 }
 
 sub as_html { ascii_html($_[0]->{raw}) }
-sub as_href { ascii_html(uri_escape_utf8($_[0]->{href})) }
 
 sub raw {
 	if (defined $_[1]) {
@@ -70,9 +84,68 @@ sub raw {
 	}
 }
 
-sub prurl {
+# returns a protocol-relative URL string
+sub prurl ($$) {
 	my ($env, $u) = @_;
+	if (ref($u) eq 'ARRAY') {
+		my $h = $env->{HTTP_HOST} // $env->{SERVER_NAME};
+		my @host_match = grep(/\b\Q$h\E\b/, @$u);
+		$u = $host_match[0] // $u->[0];
+		# fall through to below:
+	}
 	index($u, '//') == 0 ? "$env->{'psgi.url_scheme'}:$u" : $u;
 }
 
+# for misguided people who believe in this stuff, give them a
+# substitution for '.'
+# &#8228; &#183; and &#890; were also candidates:
+#   https://public-inbox.org/meta/20170615015250.GA6484@starla/
+# However, &#8226; was chosen to make copy+paste errors more obvious
+sub obfuscate_addrs ($$;$) {
+	my $ibx = $_[0];
+	my $repl = $_[2] || '&#8226;';
+	my $re = $ibx->{-no_obfuscate_re}; # regex of domains
+	my $addrs = $ibx->{-no_obfuscate}; # { adddress => 1 }
+	$_[1] =~ s/(([\w\.\+=\-]+)\@([\w\-]+\.[\w\.\-]+))/
+		my ($addr, $user, $domain) = ($1, $2, $3);
+		if ($addrs->{$addr} || ((defined $re && $domain =~ $re))) {
+			$addr;
+		} else {
+			$domain =~ s!([^\.]+)\.!$1$repl!;
+			$user . '@' . $domain
+		}
+		/sge;
+}
+
+# like format_sanitized_subject in git.git pretty.c with '%f' format string
+sub to_filename ($) {
+	my ($s, undef) = split(/\n/, $_[0]);
+	$s =~ s/[^A-Za-z0-9_\.]+/-/g;
+	$s =~ tr/././s;
+	$s =~ s/[\.\-]+\z//;
+	$s =~ s/\A[\.\-]+//;
+	$s
+}
+
+# convert a filename (or any string) to HTML attribute
+
+my %ESCAPES = map { chr($_) => sprintf('::%02x', $_) } (0..255);
+$ESCAPES{'/'} = ':'; # common
+
+sub to_attr ($) {
+	my ($str) = @_;
+
+	# git would never do this to us:
+	return if index($str, '//') >= 0;
+
+	my $first = '';
+	utf8::encode($str); # to octets
+	if ($str =~ s/\A([^A-Ya-z])//ms) { # start with a letter
+		  $first = sprintf('Z%02x', ord($1));
+	}
+	$str =~ s/([^A-Za-z0-9_\.\-])/$ESCAPES{$1}/egms;
+	utf8::decode($str); # allow wide chars
+	$first . $str;
+}
+
 1;