lib/PublicInbox/Linkify.pm | 15 +++++++++++---- t/linkify.t | 19 +++++++++++++++++++ diff --git a/lib/PublicInbox/Linkify.pm b/lib/PublicInbox/Linkify.pm index aa472cdb7167473d2706e069073eb9ef7683aa51..71193bc28732f28ca715323b04d31e53467ab189 100644 --- a/lib/PublicInbox/Linkify.pm +++ b/lib/PublicInbox/Linkify.pm @@ -15,7 +15,7 @@ use warnings; use Digest::SHA qw/sha1_hex/; my $SALT = rand; -my $LINK_RE = qr{(\()?\b((?:ftps?|https?|nntps?|gopher):// +my $LINK_RE = qr{([\('!])?\b((?:ftps?|https?|nntps?|gopher):// [\@:\w\.-]+(?:/ (?:[a-z0-9\-\._~!\$\&\';\(\)\*\+,;=:@/%]*) (?:\?[a-z0-9\-\._~!\$\&\';\(\)\*\+,;=:@/%]+)? @@ -25,6 +25,14 @@ )}xi; sub new { bless {}, $_[0] } +# try to distinguish paired punctuation chars from the URL itself +# Maybe other languages/formats can be supported here, too... +my %pairs = ( + "(" => qr/(\)[\.,;\+]?)\z/, # Markdown (,), Ruby (+) (, for arrays) + "'" => qr/('[\.,;\+]?)\z/, # Perl / Ruby + "!" => qr/(![\.,;\+]?)\z/, # Perl / Ruby +); + sub linkify_1 { $_[1] =~ s^$LINK_RE^ my $beg = $1 || ''; @@ -35,9 +43,8 @@ # it's fairly common to end URLs in messages with # '.', ',' or ';' to denote the end of a statement; # assume the intent was to end the statement/sentence # in English - # Markdown compatibility: - if ($beg eq '(') { - if ($url =~ s/(\)[\.,;]?)\z//) { + if (defined(my $re = $pairs{$beg})) { + if ($url =~ s/$re//) { $end = $1; } } elsif ($url =~ s/([\.,;])\z//) { diff --git a/t/linkify.t b/t/linkify.t index f0b3a6d0b34bdb438a0fc940a9bbb3da2bbce87b..bef4ffd6fefb360533c007cb53a74428315f8f34 100644 --- a/t/linkify.t +++ b/t/linkify.t @@ -98,4 +98,23 @@ is($s, qq![markdown]($u).!, 'Markdown-compatible end of sentence'); } +# Perl and Ruby code compatibility +{ + my $l = PublicInbox::Linkify->new; + my $u = 'http://example.com/'; + foreach my $q ("'%s'", '"%s"', 'q!%s!', 'q(%s)') { + # Perl + my $s = sprintf("my \$var = $q;", $u); + $s = $l->linkify_1($s); + $s = $l->linkify_2($s); + like($s, qr/>\Q$u\Elinkify_1($s); + $s = $l->linkify_2($s); + like($s, qr/>\Q$u\E