hval: routines for attribute escaping

We'll use HTML attributes + anchor links to link to filenames
in coming commits.

diff --git a/lib/PublicInbox/Hval.pm b/lib/PublicInbox/Hval.pm
index d908bec2f3dbb61c714c40a835fb41936bb55ca1..c0db5667717ee46606e81b9cbff2d6412394ddf2 100644 (file)
--- a/lib/PublicInbox/Hval.pm
+++ b/lib/PublicInbox/Hval.pm
@@ -10,7 +10,7 @@ use Encode qw(find_encoding);
 use URI::Escape qw(uri_escape_utf8);
 use PublicInbox::MID qw/mid_clean/;
 use base qw/Exporter/;
-our @EXPORT_OK = qw/ascii_html utf8_html/;
+our @EXPORT_OK = qw/ascii_html utf8_html to_attr from_attr/;
 
 # for user-generated content (UGC) which may have excessively long lines
 # and screw up rendering on some browsers.  This is the only CSS style
@@ -97,4 +97,35 @@ sub prurl {
        index($u, '//') == 0 ? "$env->{'psgi.url_scheme'}:$u" : $u;
 }
 
+# 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:
+       die "invalid filename: $str" if index($str, '//') >= 0;
+
+       my $first = '';
+       if ($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;
+       $first . $str;
+}
+
+# reverse the result of to_attr
+sub from_attr ($) {
+       my ($str) = @_;
+       my $first = '';
+       if ($str =~ s/\AZ([a-f0-9]{2})//ms) {
+               $first = chr(hex($1));
+       }
+       $str =~ s!::([a-f0-9]{2})!chr(hex($1))!egms;
+       $str =~ tr!:!/!;
+       $first . $str;
+}
+
 1;

diff --git a/t/hval.t b/t/hval.t
new file mode 100644 (file)
index 0000000..f824752
--- /dev/null
+++ b/t/hval.t
@@ -0,0 +1,20 @@
+# Copyright (C) 2015 all contributors <meta@public-inbox.org>
+# License: AGPL-3.0+ (https://www.gnu.org/licenses/agpl-3.0.txt)
+use strict;
+use warnings;
+use Test::More;
+use PublicInbox::Hval qw(to_attr from_attr);
+
+foreach my $s ('Hello/World.pm', 'Zcat', 'hello world.c', 'Eléanor', '$at') {
+       my $attr = to_attr($s);
+       is(from_attr($attr), $s, "$s => $attr => $s round trips");
+}
+
+{
+       my $bad = eval { to_attr('foo//bar') };
+       my $err = $@;
+       ok(!$bad, 'double-slash rejected');
+       like($err, qr/invalid filename:/, 'got exception message');
+}
+
+done_testing();