snark/html-laundry

HTML::Parser-based snippet cleaner and sanitizer

html-laundry's Issues

t/uri.t fails (with newest URI.pm?)

t/uri.t started to fail on my smoker systems:

#   Failed test 'Restricted characters left untouched in path'
#   at t/uri.t line 24.
#          got: '<img src="http://example.com/otter;/?:@&=+$,%5B%5D,.png" />'
#     expected: '<img src="http://example.com/otter;/?:@&=+$,[],.png" />'
# Looks like you failed 1 test of 36.
t/uri.t ............... 
Dubious, test returned 1 (wstat 256, 0x100)
Failed 1/36 subtests 

Statistical analysis suggests that this happens if new URI (>= 5.12) is installed:

****************************************************************
Regression 'mod:URI::Escape'
****************************************************************
Name           	       Theta	      StdErr	 T-stat
[0='const']    	      1.0000	      0.0000	88121074673334784.00
[1='eq_5.10']  	      0.0000	      0.0000	   0.70
[2='eq_5.12']  	     -1.0000	      0.0000	-21372499924770628.00
[3='eq_5.14']  	     -1.0000	      0.0000	-27066172930994744.00
[4='eq_5.15']  	     -1.0000	      0.0000	-8947339523702376.00
[5='eq_5.16']  	     -1.0000	      0.0000	-40055033942424904.00

R^2= 1.000, N= 140, K= 6
****************************************************************
(4)
****************************************************************
Regression 'mod:URI'
****************************************************************
Name           	       Theta	      StdErr	 T-stat
[0='const']    	      1.0000	      0.0000	70610503637522888.00
[1='eq_1.72']  	      0.0000	      0.0000	   0.00
[2='eq_1.73']  	     -0.0000	      0.0000	  -7.62
[3='eq_1.74']  	      0.0000	      0.0000	   3.19
[4='eq_1.76']  	      0.0000	      0.0000	   1.16
[5='eq_5.10']  	     -0.0000	      0.0000	  -1.03
[6='eq_5.12']  	     -1.0000	      0.0000	-20821901562797712.00
[7='eq_5.14']  	     -1.0000	      0.0000	-26134119890636624.00
[8='eq_5.15']  	     -1.0000	      0.0000	-8826312954690363.00
[9='eq_5.16']  	     -1.0000	      0.0000	-37635524894375792.00

R^2= 1.000, N= 140, K= 10
****************************************************************
(5)
****************************************************************
Regression 'mod:URI::Split'
****************************************************************
Name           	       Theta	      StdErr	 T-stat
[0='const']    	      1.0000	      0.0000	70610503637522888.00
[1='eq_1.72']  	      0.0000	      0.0000	   0.00
[2='eq_1.73']  	     -0.0000	      0.0000	  -7.62
[3='eq_1.74']  	      0.0000	      0.0000	   3.19
[4='eq_1.76']  	      0.0000	      0.0000	   1.16
[5='eq_5.10']  	     -0.0000	      0.0000	  -1.03
[6='eq_5.12']  	     -1.0000	      0.0000	-20821901562797712.00
[7='eq_5.14']  	     -1.0000	      0.0000	-26134119890636624.00
[8='eq_5.15']  	     -1.0000	      0.0000	-8826312954690363.00
[9='eq_5.16']  	     -1.0000	      0.0000	-37635524894375792.00

R^2= 1.000, N= 140, K= 10
****************************************************************

Can't handle character data

The following sample code:

use warnings;
use strict;
use utf8;
use HTML::Laundry;
my $html =<<EOF;
<p>出す医すと愚っと</p>
EOF
my $laundry = HTML::Laundry->new();
my $launout = $laundry->clean ($html);

causes a fatal error:

$ perl laundry-test.pl
Wide character in subroutine entry at /home/ben/software/install/lib/perl5/site_perl/5.18.2/HTML/Laundry.pm line 316.

Commenting out the "use utf8;" line removes the error.

Test Failure in add_elements.t

I tried installing HTML::Laundry and encountered a test failure. I notice several CPAN Testers see the same failure, for example http://www.cpantesters.org/cpan/report/74662a1e-6b16-11e1-93a2-983e49abe0af

I haven't looked at why this happens, but I thought you might appreciate me reporting the problem.

Thank you for your work on this module,
Tom