Update upstream source from tag 'upstream/3.0.0'

Update to upstream version '3.0.0'
with Debian dir eeb9d119ee0e5e23b0dbf4c6e07ceae6bad0bac9

14 files changed, 128 insertions, 200 deletions

diff --git a/Changes b/Changes
index 37d6e96..40d96cd 100644
--- a/Changes
+++ b/Changes
@@ -1,8 +1,8 @@
 Revision history for HTML-Restrict
 
-v2.5.0    2019-02-08 22:18:11Z
-        - Strip some control characters from links (GH#34) (Olaf Alders)
-        - Enable empty_element_tags in HTML::Parser (GH#35) (Olaf Alders)
+v3.0.0    2019-03-11 13:52:08Z
+        - Better fix for handling malformed tags. Removes max_parser_loops(),
+          which was introduced in 2.4.0 (GH#37) (Graham Knop)
 
 v2.4.1    2019-02-05 14:13:16Z
         - Bump version of Type::Tiny to 1.002001. (GH#33) (Olaf Alders).  Issue
diff --git a/MANIFEST b/MANIFEST
index 730056a..e2732ec 100644
--- a/MANIFEST
+++ b/MANIFEST
@@ -14,7 +14,6 @@ dist.ini
 examples/naughty-strings.pl
 examples/sanitize-file.pl
 lib/HTML/Restrict.pm
-lib/HTML/Restrict/Types.pm
 perlcriticrc
 perltidyrc
 t/00-load.t
diff --git a/META.json b/META.json
index de1c81d..a7b9504 100644
--- a/META.json
+++ b/META.json
@@ -48,7 +48,8 @@
             "Test::Pod::Coverage" : "1.08",
             "Test::Spelling" : "0.12",
             "Test::Synopsis" : "0",
-            "Test::Vars" : "0.014"
+            "Test::Vars" : "0.014",
+            "warnings" : "0"
          }
       },
       "runtime" : {
@@ -61,15 +62,13 @@
             "Moo" : "1.002000",
             "Scalar::Util" : "0",
             "Sub::Quote" : "0",
-            "Type::Library" : "0",
             "Type::Tiny" : "1.002001",
-            "Type::Utils" : "0",
+            "Types::Standard" : "1.000001",
             "URI" : "0",
             "namespace::clean" : "0",
             "perl" : "5.006",
             "strict" : "0",
-            "version" : "0",
-            "warnings" : "0"
+            "version" : "0"
          }
       },
       "test" : {
@@ -81,7 +80,8 @@
             "File::Spec" : "0",
             "Test::Fatal" : "0",
             "Test::More" : "0",
-            "perl" : "5.006"
+            "perl" : "5.006",
+            "warnings" : "0"
          }
       }
    },
@@ -97,7 +97,7 @@
          "web" : "https://github.com/oalders/html-restrict"
       }
    },
-   "version" : "v2.5.0",
+   "version" : "v3.0.0",
    "x_Dist_Zilla" : {
       "perl" : {
          "version" : "5.026001"
@@ -138,7 +138,7 @@
          {
             "class" : "Dist::Zilla::Plugin::MAXMIND::TidyAll",
             "name" : "@Author::OALDERS/MAXMIND::TidyAll",
-            "version" : "0.83"
+            "version" : "0.13"
          },
          {
             "class" : "Dist::Zilla::Plugin::AutoPrereqs",
@@ -629,7 +629,7 @@
                   "branch" : null,
                   "changelog" : "Changes",
                   "signed" : 0,
-                  "tag" : "vv2.5.0",
+                  "tag" : "vv3.0.0",
                   "tag_format" : "v%v",
                   "tag_message" : "v%v"
                },
@@ -723,6 +723,28 @@
             "version" : "6.012"
          },
          {
+            "class" : "Dist::Zilla::Plugin::Git::Tag",
+            "config" : {
+               "Dist::Zilla::Plugin::Git::Tag" : {
+                  "branch" : null,
+                  "changelog" : "Changes",
+                  "signed" : 0,
+                  "tag" : "v3.0.0",
+                  "tag_format" : "%v",
+                  "tag_message" : "%v"
+               },
+               "Dist::Zilla::Role::Git::Repo" : {
+                  "git_version" : "2.20.1",
+                  "repo_root" : "."
+               },
+               "Dist::Zilla::Role::Git::StringFormatter" : {
+                  "time_zone" : "local"
+               }
+            },
+            "name" : "Git::Tag",
+            "version" : "2.045"
+         },
+         {
             "class" : "Dist::Zilla::Plugin::FinderCode",
             "name" : ":InstallModules",
             "version" : "6.012"
@@ -795,6 +817,6 @@
       "skaufman <sam@socialflow.com>"
    ],
    "x_generated_by_perl" : "v5.26.1",
-   "x_serialization_backend" : "Cpanel::JSON::XS version 4.04"
+   "x_serialization_backend" : "Cpanel::JSON::XS version 4.02"
 }
 
diff --git a/META.yml b/META.yml
index 9830551..059147a 100644
--- a/META.yml
+++ b/META.yml
@@ -8,6 +8,7 @@ build_requires:
   Test::Fatal: '0'
   Test::More: '0'
   perl: '5.006'
+  warnings: '0'
 configure_requires:
   ExtUtils::MakeMaker: '0'
   perl: '5.006'
@@ -32,20 +33,18 @@ requires:
   Moo: '1.002000'
   Scalar::Util: '0'
   Sub::Quote: '0'
-  Type::Library: '0'
   Type::Tiny: '1.002001'
-  Type::Utils: '0'
+  Types::Standard: '1.000001'
   URI: '0'
   namespace::clean: '0'
   perl: '5.006'
   strict: '0'
   version: '0'
-  warnings: '0'
 resources:
   bugtracker: https://github.com/oalders/html-restrict/issues
   homepage: https://github.com/oalders/html-restrict
   repository: https://github.com/oalders/html-restrict.git
-version: v2.5.0
+version: v3.0.0
 x_Dist_Zilla:
   perl:
     version: '5.026001'
@@ -78,7 +77,7 @@ x_Dist_Zilla:
     -
       class: Dist::Zilla::Plugin::MAXMIND::TidyAll
       name: '@Author::OALDERS/MAXMIND::TidyAll'
-      version: '0.83'
+      version: '0.13'
     -
       class: Dist::Zilla::Plugin::AutoPrereqs
       name: '@Author::OALDERS/AutoPrereqs'
@@ -463,7 +462,7 @@ x_Dist_Zilla:
           branch: ~
           changelog: Changes
           signed: 0
-          tag: vv2.5.0
+          tag: vv3.0.0
           tag_format: v%v
           tag_message: v%v
         Dist::Zilla::Role::Git::Repo:
@@ -531,6 +530,23 @@ x_Dist_Zilla:
       name: Prereqs
       version: '6.012'
     -
+      class: Dist::Zilla::Plugin::Git::Tag
+      config:
+        Dist::Zilla::Plugin::Git::Tag:
+          branch: ~
+          changelog: Changes
+          signed: 0
+          tag: v3.0.0
+          tag_format: '%v'
+          tag_message: '%v'
+        Dist::Zilla::Role::Git::Repo:
+          git_version: 2.20.1
+          repo_root: .
+        Dist::Zilla::Role::Git::StringFormatter:
+          time_zone: local
+      name: Git::Tag
+      version: '2.045'
+    -
       class: Dist::Zilla::Plugin::FinderCode
       name: ':InstallModules'
       version: '6.012'
@@ -588,4 +604,4 @@ x_contributors:
   - 'perlpong <calyx238@gmail.com>'
   - 'skaufman <sam@socialflow.com>'
 x_generated_by_perl: v5.26.1
-x_serialization_backend: 'YAML::Tiny version 1.70'
+x_serialization_backend: 'YAML::Tiny version 1.73'
diff --git a/Makefile.PL b/Makefile.PL
index 34e8077..16e61d2 100644
--- a/Makefile.PL
+++ b/Makefile.PL
@@ -25,22 +25,21 @@ my %WriteMakefileArgs = (
     "Moo" => "1.002000",
     "Scalar::Util" => 0,
     "Sub::Quote" => 0,
-    "Type::Library" => 0,
     "Type::Tiny" => "1.002001",
-    "Type::Utils" => 0,
+    "Types::Standard" => "1.000001",
     "URI" => 0,
     "namespace::clean" => 0,
     "strict" => 0,
-    "version" => 0,
-    "warnings" => 0
+    "version" => 0
   },
   "TEST_REQUIRES" => {
     "ExtUtils::MakeMaker" => 0,
     "File::Spec" => 0,
     "Test::Fatal" => 0,
-    "Test::More" => 0
+    "Test::More" => 0,
+    "warnings" => 0
   },
-  "VERSION" => "v2.5.0",
+  "VERSION" => "v3.0.0",
   "test" => {
     "TESTS" => "t/*.t"
   }
@@ -60,9 +59,8 @@ my %FallbackPrereqs = (
   "Sub::Quote" => 0,
   "Test::Fatal" => 0,
   "Test::More" => 0,
-  "Type::Library" => 0,
   "Type::Tiny" => "1.002001",
-  "Type::Utils" => 0,
+  "Types::Standard" => "1.000001",
   "URI" => 0,
   "namespace::clean" => 0,
   "strict" => 0,
diff --git a/README.md b/README.md
index 18d67b9..b47aa9c 100644
--- a/README.md
+++ b/README.md
@@ -4,7 +4,7 @@ HTML::Restrict - Strip unwanted HTML tags and attributes
 
 # VERSION
 
-version v2.5.0
+version v3.0.0
 
 # SYNOPSIS
 
@@ -233,30 +233,6 @@ HTML::Restrict recognizes:
         $html = $hr->process( $html );
         # $html is now: "<!-- comments! -->foo"
 
-- max\_parser\_loops => \[Integer\]
-
-    Defaults to 25.  Should never be less than 2.
-
-    As of v2.4.0, calling `process()` will force the parser to clean the text
-    multiple times, stopping only once the text is no longer changed or once
-    `max_parser_loops` has been reached.
-
-    The reason for this is that [HTML::Parser](https://metacpan.org/pod/HTML::Parser) could take malformed HTML and turn
-    it into well formed HTML.  This can defeat our processing logic and allow
-    malicious input to be returned.  In order to mitigate this, we will clean all
-    input at least two times.  If the second attempt at cleaning does not match
-    the previous attempt, we will make a third attempt and so on.  This helps to
-    ensure that we get the expected output.
-
-    If we are unable to get unchanged values after reaching `max_parser_loops`, an
-    exception will be thrown.  Returning partially cleaned text would be wrong, as
-    would be returning `undef` or an empty string.  Throwing an exception forces
-    the user to choose the appropriate way of dealing with this.
-
-    If you choose to set this value, please note that it can be no less than 2, or
-    the parser will never be able to make a comparison with a previous value.  An
-    exception will be thrown if you attempt to set this to a value less than 2.
-
 - replace\_img => \[0|1|CodeRef\]
 
     Set the value to true if you'd like to have img tags replaced with
diff --git a/cpanfile b/cpanfile
index 1cf9d0d..6e5d895 100644
--- a/cpanfile
+++ b/cpanfile
@@ -6,15 +6,13 @@ requires "List::Util" => "1.33";
 requires "Moo" => "1.002000";
 requires "Scalar::Util" => "0";
 requires "Sub::Quote" => "0";
-requires "Type::Library" => "0";
 requires "Type::Tiny" => "1.002001";
-requires "Type::Utils" => "0";
+requires "Types::Standard" => "1.000001";
 requires "URI" => "0";
 requires "namespace::clean" => "0";
 requires "perl" => "5.006";
 requires "strict" => "0";
 requires "version" => "0";
-requires "warnings" => "0";
 
 on 'test' => sub {
   requires "ExtUtils::MakeMaker" => "0";
@@ -22,6 +20,7 @@ on 'test' => sub {
   requires "Test::Fatal" => "0";
   requires "Test::More" => "0";
   requires "perl" => "5.006";
+  requires "warnings" => "0";
 };
 
 on 'test' => sub {
@@ -51,6 +50,7 @@ on 'develop' => sub {
   requires "Test::Spelling" => "0.12";
   requires "Test::Synopsis" => "0";
   requires "Test::Vars" => "0.014";
+  requires "warnings" => "0";
 };
 
 on 'develop' => sub {
diff --git a/dist.ini b/dist.ini
index c8bcf1b..30e9483 100644
--- a/dist.ini
+++ b/dist.ini
@@ -9,3 +9,7 @@ main_module = lib/HTML/Restrict.pm
 
 [Prereqs]
 Type::Tiny = 1.002001
+
+[Git::Tag]
+tag_format  = %v
+tag_message = %v
diff --git a/lib/HTML/Restrict.pm b/lib/HTML/Restrict.pm
index 6c2424d..fce0fba 100644
--- a/lib/HTML/Restrict.pm
+++ b/lib/HTML/Restrict.pm
@@ -4,20 +4,13 @@ use 5.006;
 package HTML::Restrict;
 
 use version;
-our $VERSION = 'v2.5.0';
+our $VERSION = 'v3.0.0';
 
 use Carp qw( croak );
 use Data::Dump qw( dump );
-use HTML::Entities qw( encode_entities );
 use HTML::Parser ();
-use HTML::Restrict::Types qw(
-    ArrayRef
-    Bool
-    CodeRef
-    HashRef
-    Int
-    MaxParserLoops
-);
+use HTML::Entities qw( encode_entities );
+use Types::Standard 1.000001 qw[ Bool HashRef ArrayRef CodeRef ];
 use List::Util 1.33 qw( any none );
 use Scalar::Util qw( reftype weaken );
 use Sub::Quote 'quote_sub';
@@ -44,12 +37,6 @@ has debug => (
     default => 0,
 );
 
-has max_parser_loops => (
-    is      => 'rw',
-    isa     => MaxParserLoops,
-    default => 25,
-);
-
 has parser => (
     is      => 'ro',
     lazy    => 1,
@@ -280,6 +267,7 @@ sub _build_parser {
             sub {
                 my ( $p, $text ) = @_;
                 print "text: $text\n" if $self->debug;
+                $text = _fix_text_encoding($text);
                 if ( !@{ $self->_stripper_stack } ) {
                     $self->_processed( ( $self->_processed || q{} ) . $text );
                 }
@@ -315,29 +303,6 @@ sub _build_parser {
 sub process {
     my $self = shift;
 
-    my $cleaned = $self->_process(@_);
-    return $cleaned if !$cleaned;
-
-    my $previous_iteration = $cleaned;
-
-    my $i = 1;    # We already cleaned once just above
-
-    while ( $i < $self->max_parser_loops ) {
-        $i++;
-        my $new = $self->_process($previous_iteration);
-        last if $new eq $previous_iteration;
-        if ( $i == $self->max_parser_loops ) {
-            die sprintf( 'Could not clean input after %s attempts', $i );
-        }
-
-        $previous_iteration = $new;
-    }
-    return $previous_iteration;
-}
-
-sub _process {
-    my $self = shift;
-
     # returns undef if no value was passed
     return if !@_;
     return $_[0] if !$_[0];
@@ -388,6 +353,39 @@ sub _delete_tag_from_stack {
     return;
 }
 
+# regex for entities that don't require a terminating semicolon
+my ($short_entity_re)
+    = map qr/$_/i,
+    join '|',
+    '#x[0-9a-f]+',
+    '#[0-9]+',
+    grep !/;\z/,
+    sort keys %HTML::Entities::entity2char;
+
+# semicolon required
+my ($complete_entity_re)
+    = map qr/$_/i,
+    join '|',
+    grep /;\z/,
+    sort keys %HTML::Entities::entity2char;
+
+sub _fix_text_encoding {
+    my $text = shift;
+    $text =~ s{
+        &
+        (?:
+          ($short_entity_re);?
+        |
+          ($complete_entity_re)
+        )?
+    }{
+          defined $1  ? "&$1;"
+        : defined $2  ? "&$2"
+                      : "&amp;"
+    }xgie;
+    return encode_entities( $text, '<>' );
+}
+
 1;    # End of HTML::Restrict
 
 # ABSTRACT: Strip unwanted HTML tags and attributes
@@ -404,7 +402,7 @@ HTML::Restrict - Strip unwanted HTML tags and attributes
 
 =head1 VERSION
 
-version v2.5.0
+version v3.0.0
 
 =head1 SYNOPSIS
 
@@ -635,30 +633,6 @@ feature is off by default.
     $html = $hr->process( $html );
     # $html is now: "<!-- comments! -->foo"
 
-=item * max_parser_loops => [Integer]
-
-Defaults to 25.  Should never be less than 2.
-
-As of v2.4.0, calling C<process()> will force the parser to clean the text
-multiple times, stopping only once the text is no longer changed or once
-C<max_parser_loops> has been reached.
-
-The reason for this is that L<HTML::Parser> could take malformed HTML and turn
-it into well formed HTML.  This can defeat our processing logic and allow
-malicious input to be returned.  In order to mitigate this, we will clean all
-input at least two times.  If the second attempt at cleaning does not match
-the previous attempt, we will make a third attempt and so on.  This helps to
-ensure that we get the expected output.
-
-If we are unable to get unchanged values after reaching C<max_parser_loops>, an
-exception will be thrown.  Returning partially cleaned text would be wrong, as
-would be returning C<undef> or an empty string.  Throwing an exception forces
-the user to choose the appropriate way of dealing with this.
-
-If you choose to set this value, please note that it can be no less than 2, or
-the parser will never be able to make a comparison with a previous value.  An
-exception will be thrown if you attempt to set this to a value less than 2.
-
 =item * replace_img => [0|1|CodeRef]
 
 Set the value to true if you'd like to have img tags replaced with
diff --git a/lib/HTML/Restrict/Types.pm b/lib/HTML/Restrict/Types.pm
deleted file mode 100644
index ba80a1e..0000000
--- a/lib/HTML/Restrict/Types.pm
+++ /dev/null
@@ -1,50 +0,0 @@
-package HTML::Restrict::Types;
-our $VERSION = 'v2.5.0';
-use strict;
-use warnings;
-
-use Type::Library -base;
-use Type::Utils ();
-
-BEGIN {
-    Type::Utils::extends( 'Types::Common::Numeric', 'Types::Standard', );
-}
-
-__PACKAGE__->add_type(
-    {
-        name       => 'MaxParserLoops',
-        parent     => PositiveInt,
-        constraint => '$_ >= 2',
-    }
-);
-
-1;
-
-=pod
-
-=encoding UTF-8
-
-=head1 NAME
-
-HTML::Restrict::Types - Type library for HTML::Restrict
-
-=head1 VERSION
-
-version v2.5.0
-
-=head1 AUTHOR
-
-Olaf Alders <olaf@wundercounter.com>
-
-=head1 COPYRIGHT AND LICENSE
-
-This software is copyright (c) 2013-2017 by Olaf Alders.
-
-This is free software; you can redistribute it and/or modify it under
-the same terms as the Perl 5 programming language system itself.
-
-=cut
-
-__END__
-
-# ABSTRACT: Type library for HTML::Restrict
diff --git a/t/00-report-prereqs.dd b/t/00-report-prereqs.dd
index ce615bf..31bf62d 100644
--- a/t/00-report-prereqs.dd
+++ b/t/00-report-prereqs.dd
@@ -26,7 +26,8 @@ do { my $x = {
                                       'Test::Pod::Coverage' => '1.08',
                                       'Test::Spelling' => '0.12',
                                       'Test::Synopsis' => '0',
-                                      'Test::Vars' => '0.014'
+                                      'Test::Vars' => '0.014',
+                                      'warnings' => '0'
                                     }
                     },
        'runtime' => {
@@ -39,15 +40,13 @@ do { my $x = {
                                       'Moo' => '1.002000',
                                       'Scalar::Util' => '0',
                                       'Sub::Quote' => '0',
-                                      'Type::Library' => '0',
                                       'Type::Tiny' => '1.002001',
-                                      'Type::Utils' => '0',
+                                      'Types::Standard' => '1.000001',
                                       'URI' => '0',
                                       'namespace::clean' => '0',
                                       'perl' => '5.006',
                                       'strict' => '0',
-                                      'version' => '0',
-                                      'warnings' => '0'
+                                      'version' => '0'
                                     }
                     },
        'test' => {
@@ -59,7 +58,8 @@ do { my $x = {
                                    'File::Spec' => '0',
                                    'Test::Fatal' => '0',
                                    'Test::More' => '0',
-                                   'perl' => '5.006'
+                                   'perl' => '5.006',
+                                   'warnings' => '0'
                                  }
                  }
      };
diff --git a/t/empty-element-tags.t b/t/empty-element-tags.t
index ce9cd9a..e6dc3f0 100644
--- a/t/empty-element-tags.t
+++ b/t/empty-element-tags.t
@@ -12,9 +12,9 @@ one element open & close break no space<br/>
 EOF
 
 my $after = <<'EOF';
-two element open & close break<br></br>
-one element open & close break <br>
-one element open & close break no space<br>
+two element open &amp; close break<br></br>
+one element open &amp; close break <br>
+one element open &amp; close break no space<br>
 EOF
 
 my $hr = HTML::Restrict->new(
diff --git a/t/malformed-html.t b/t/malformed-html.t
index afb1288..d838ec1 100755
--- a/t/malformed-html.t
+++ b/t/malformed-html.t
@@ -22,31 +22,21 @@ my $html = '<<input>div onmouseover="alert(1);">hover over me<<input>/div>';
 {
     my $hr = HTML::Restrict->new;
     is(
-        $hr->process($html), 'hover over me',
+        $hr->process(
+            '<<input>div onmouseover="alert(1);">hover over me<<input>/div>'),
+        '&lt;div onmouseover="alert(1);"&gt;hover over me&lt;/div&gt;',
         'malformed HTML is correctly cleaned'
     );
 }
 
 {
-    my $attempts = 2;
-    my $hr       = HTML::Restrict->new( max_parser_loops => $attempts );
-    like(
-        exception { $hr->process($html) },
-        qr/after $attempts attempts/,
-        'dies after max loops exceeded',
+    my $hr = HTML::Restrict->new;
+    is(
+        $hr->process(
+            '&<input></input>lt; &theta; &aMp; &#50; &#x50; &#xabg;'),
+        '&amp;lt; &theta; &aMp; &#50; &#x50; &#xab;g;',
+        'badly encoded entities corrected'
     );
-    $hr->max_parser_loops(3);
-    is( $hr->process('<foo>bar'), 'bar', 'can parse after caught exception' );
-}
-
-{
-    for my $i ( -1 .. 1 ) {
-        like(
-            exception { HTML::Restrict->new( max_parser_loops => $i ) },
-            qr/did not pass type constraint/i,
-            'max_parser_loops cannot be ' . $i,
-        );
-    }
 }
 
 done_testing();
diff --git a/xt/author/pod-spell.t b/xt/author/pod-spell.t
index 5a8e4d9..a813d52 100644
--- a/xt/author/pod-spell.t
+++ b/xt/author/pod-spell.t
@@ -38,7 +38,6 @@ Raybec
 Restrict
 Schmidt
 TerMarsch
-Types
 XSS
 benkasminbullock
 bolded