1 # Copyright (C) 2014-2021 all contributors <meta@public-inbox.org>
2 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
4 # represents a header value in various forms. Used for HTML generation
5 # in our web interface(s)
6 package PublicInbox::Hval;
9 use Encode qw(find_encoding);
10 use PublicInbox::MID qw/mid_clean mid_escape/;
11 use base qw/Exporter/;
12 our @EXPORT_OK = qw/ascii_html obfuscate_addrs to_filename src_escape
13 to_attr prurl mid_href fmt_ts ts2str/;
14 use POSIX qw(strftime);
15 my $enc_ascii = find_encoding('us-ascii');
17 # safe-ish acceptable filename pattern for portability
18 our $FN = '[a-zA-Z0-9][a-zA-Z0-9_\-\.]+[a-zA-Z0-9]'; # needs \z anchor
20 sub mid_href { ascii_html(mid_escape($_[0])) }
22 # some of these overrides are standard C escapes so they're
23 # easy-to-understand when rendered.
24 my %escape_sequence = (
25 "\x00" => '\\0', # NUL
26 "\x07" => '\\a', # bell
27 "\x08" => '\\b', # backspace
28 "\x09" => "\t", # obvious to show as-is
29 "\x0a" => "\n", # obvious to show as-is
30 "\x0b" => '\\v', # vertical tab
31 "\x0c" => '\\f', # form feed
32 "\x0d" => '\\r', # carriage ret (not preceding \n)
33 "\x1b" => '^[', # ASCII escape (mutt seems to escape this way)
34 "\x7f" => '\\x7f', # DEL
45 $xhtml_map{chr($_)} = sprintf('\\x%02x', $_) for (0..31);
46 %xhtml_map = (%xhtml_map, %escape_sequence);
48 # for post-processing the output of highlight.pm and perhaps other
49 # highlighers in the future
51 $_[0] =~ s/\r\n/\n/sg;
52 $_[0] =~ s/'/'/sg; # workaround https://bugs.debian.org/927409
53 $_[0] =~ s/([\x7f\x00-\x1f])/$xhtml_map{$1}/sge;
54 $_[0] = $enc_ascii->encode($_[0], Encode::HTMLCREF);
59 $s =~ s/([<>&'"\x7f\x00-\x1f])/$xhtml_map{$1}/sge;
60 $enc_ascii->encode($s, Encode::HTMLCREF);
63 # returns a protocol-relative URL string
66 if (ref($u) eq 'ARRAY') {
67 my $h = $env->{HTTP_HOST} // $env->{SERVER_NAME};
68 my @host_match = grep(/\b\Q$h\E\b/, @$u);
69 $u = $host_match[0] // $u->[0];
70 # fall through to below:
72 index($u, '//') == 0 ? "$env->{'psgi.url_scheme'}:$u" : $u;
75 # for misguided people who believe in this stuff, give them a
76 # substitution for '.'
77 # ․ · and ͺ were also candidates:
78 # https://public-inbox.org/meta/20170615015250.GA6484@starla/
79 # However, • was chosen to make copy+paste errors more obvious
80 sub obfuscate_addrs ($$;$) {
82 my $repl = $_[2] // '•';
83 my $re = $ibx->{-no_obfuscate_re}; # regex of domains
84 my $addrs = $ibx->{-no_obfuscate}; # { $address => 1 }
85 $_[1] =~ s#(\S+)\@([\w\-]+\.[\w\.\-]+)#
86 my ($pfx, $domain) = ($1, $2);
87 if (index($pfx, '://') > 0 || $pfx !~ s/([\w\.\+=\-]+)\z//) {
91 my $addr = "$user\@$domain";
92 if ($addrs->{$addr} || ((defined($re) &&
96 $domain =~ s!([^\.]+)\.!$1$repl!;
97 $pfx . $user . '@' . $domain
103 # like format_sanitized_subject in git.git pretty.c with '%f' format string
104 sub to_filename ($) {
105 my $s = (split(/\n/, $_[0]))[0] // return; # empty string => undef
106 $s =~ s/[^A-Za-z0-9_\.]+/-/g;
110 $s eq '' ? undef : $s;
113 # convert a filename (or any string) to HTML attribute
115 my %ESCAPES = map { chr($_) => sprintf('::%02x', $_) } (0..255);
116 $ESCAPES{'/'} = ':'; # common
121 # git would never do this to us:
122 return if index($str, '//') >= 0;
125 utf8::encode($str); # to octets
126 if ($str =~ s/\A([^A-Ya-z])//ms) { # start with a letter
127 $first = sprintf('Z%02x', ord($1));
129 $str =~ s/([^A-Za-z0-9_\.\-])/$ESCAPES{$1}/egms;
130 utf8::decode($str); # allow wide chars
134 # for the t= query parameter passed to overview DB
135 sub ts2str ($) { strftime('%Y%m%d%H%M%S', gmtime($_[0])) };
137 # human-friendly format
138 sub fmt_ts ($) { strftime('%Y-%m-%d %k:%M', gmtime($_[0])) }