Skip to content

Improve email address parsing #14

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
dracos wants to merge 7 commits into
base: main
Choose a base branch
from
Open

Improve email address parsing #14

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
d77b859
Allow non-ASCII characters in email addresses.
dracos May 13, 2016
fa81a51
Bring regexes more in line with RFC5322.
dracos May 13, 2016
b142fb6
Try and prevent backtracking regex explosions.
dracos May 13, 2016
9066167
Use regex recusion for comment nesting.
dracos May 13, 2016
7205fc9
Don't extract "comments" from in quoted strings.
dracos May 14, 2016
c61bb28
Fetch phrase/user/host from main regex.
dracos May 14, 2016
5182577
Parse display name for multiple quoted strings.
dracos May 14, 2016
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 8 additions & 17 deletions lib/Email/Address.pm
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,6 @@ package Email::Address;

our $COMMENT_NEST_LEVEL ||= 1;
our $STRINGIFY ||= 'format';
our $COLLAPSE_SPACES = 1 unless defined $COLLAPSE_SPACES; # I miss //=
our $UNICODE ||= 0;

=head1 SYNOPSIS
Expand Down Expand Up @@ -37,25 +36,25 @@ for (1 .. $COMMENT_NEST_LEVEL) {
$ccontent = qr/$ctext|$quoted_pair|$comment/;
$comment = qr/\($ccontent*\)/;
}
my $cfws = qr/$comment|\s+/;
my $cfws = qr/(?>$comment|\s+)/;

@nwellnhof nwellnhof May 14, 2016

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this construct is only supported from 5.10 onwards.

@dracos dracos May 14, 2016

Copy link
Copy Markdown
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

http://perldoc.perl.org/5.8.8/perlre.html#(%3f%3epattern) says it is supported but experimental. I haven't got perl 5.8 installed to test, I'm afraid; later commits in this PR are definitely 5.10 only, but I hoped this one would be okay so that 5.8 could get the space backtracking improvements, if not the comment backtracking.


my $atext = qq/[^$CTL$special\\s]/;
my $atom = qr/$cfws*$atext+$cfws*/;
my $atom = qr/$cfws*(?>$atext+)$cfws*/;
my $dot_atom_text = qr/$atext+(?:\.$atext+)*/;
my $dot_atom = qr/$cfws*$dot_atom_text$cfws*/;
my $dot_atom = qr/$cfws*(?>$dot_atom_text)$cfws*/;

my $qtext = qr/[^$CTL\\"]/;
my $qcontent = qr/$qtext|$quoted_pair/;
my $quoted_string = qr/$cfws*"$qcontent*"$cfws*/;
my $quoted_string = qr/$cfws*"(?>$qcontent*)"$cfws*/;

my $word = qr/$atom|$quoted_string/;

my $obs_phrase = qr/$word(?:$word|\.|$cfws)*/;
my $phrase = qr/$obs_phrase|(?:$word+)/;
my $obs_phrase = qr/$word(?>(?:$word|\.|$cfws)*)/;
my $phrase = qr/$obs_phrase|(?>$word+)/;

my $local_part = qr/$dot_atom|$quoted_string/;
my $dtext = qr/[^$CTL\[\]\\]/;
my $domain_literal = qr/$cfws*\[$dtext*\]$cfws*/;
my $domain_literal = qr/$cfws*\[(?>$dtext*)\]$cfws*/;
my $domain = qr/$dot_atom|$domain_literal/;

my $display_name = $phrase;
Expand Down Expand Up @@ -109,7 +108,7 @@ following comment.

our $addr_spec = qr/$local_part\@$domain/;
our $angle_addr = qr/$cfws*<$addr_spec>$cfws*/;
our $name_addr = qr/(?>$display_name?)$angle_addr/;
our $name_addr = qr/$display_name?$angle_addr/;
our $mailbox = qr/$name_addr|$addr_spec/;

sub _PHRASE () { 0 }
Expand Down Expand Up @@ -150,12 +149,6 @@ C<$Email::Address::COMMENT_NEST_LEVEL> package variable to allow more.

The reason for this hardly-limiting limitation is simple: efficiency.

Long strings of whitespace can be problematic for this module to parse, a bug
which has not yet been adequately addressed. The default behavior is now to
collapse multiple spaces into a single space, which avoids this problem. To
prevent this behavior, set C<$Email::Address::COLLAPSE_SPACES> to zero. This
variable will go away when the bug is resolved properly.

By default, this module mandates that email addresses be ASCII only, and any
non-ASCII content will cause a blank result. This matches RFCs 822, 2822, and
5322. If you wish to allow UTF-8 characters in email, as per RFCs 5335 and
Expand Down Expand Up @@ -187,8 +180,6 @@ sub parse {
my ($class, $line) = @_;
return unless $line;

$line =~ s/[ \t]+/ /g if $COLLAPSE_SPACES;

if (my @cached = $class->__get_cached_parse($line)) {
return @cached;
}
Expand Down
15 changes: 15 additions & 0 deletions t/speed.t
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,15 @@
#!perl
use strict;

use Email::Address;
use Test::More tests => 1;

my $email = "\"Hello\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\" <\@m>";
my ($ea) = Email::Address->parse($email);

is(
$ea,
undef,
'Bad address does not parse, but is not really slow'
);