Revised WebGUI::HTML::filter "all" so that text does not run together when tags are removed. Added additional tests to HTML.t.

docs/changelog/7.x.x.txt

@@ -21,6 +21,8 @@
    doesn't work.
  - Fixed a cross-Matrix linking problem when you have two or more Matricies on
    one site with the same category names.
+ - fix: Revised WebGUI::HTML::filter "all" so that text does not run together when
+   tags are removed. Added additional tests to HTML.t. (Eric Kennedy)
 
 
 7.0.2

lib/WebGUI/HTML.pm

@@ -102,8 +102,42 @@ sub filter {
 	my $html = shift;
 	my $type = shift;
 	if ($type eq "all") {
-		my $filter = HTML::TagFilter->new(allow=>{'none'},strip_comments=>1);
-		$html = $filter->filter($html);
+		#Hash used to keep track of depth within tags
+		my %html_parser_inside_tag; 
+		#String containing text output from HTML::Parser
+		my $html_parser_text = "" ;
+		#Hash containing HTML tags (as keys) that create whitespace when rendered by the browser
+		my %html_parser_whitespace_tags = ('p'=>1, 'br'=>1, 'hr'=>1, 'td'=>1, 'th'=>1, 
+													  'tr'=>1, 'table'=>1, 'ul'=>1, 'li'=>1, 'div'=>1) ;
+		#HTML::Parser event handler called at the start and end of each HTML tag, adds whitespace (if necessary) 
+		#to output if the tag creates whitespace.  This was done to keep text from running together inappropriately.
+		my $html_parser_tag_sub = sub {
+			my($tag, $num) = @_;
+			$html_parser_inside_tag{$tag} += $num;
+			if ($html_parser_whitespace_tags{$tag} && 
+			    ($html_parser_text =~ /\S$/)) { #add space only if no preceeding space
+				$html_parser_text .= " "  ;
+   		}
+		} ;
+		#HTML::Parser event handler called with non-tag text (no tags)
+		my $html_parser_text_sub = sub {
+			return if $html_parser_inside_tag{script} || $html_parser_inside_tag{style}; # do not output text
+			$html_parser_text .= $_[0] ;
+		} ;
+		my $parser = HTML::Parser->new(api_version => 3,
+		  handlers => [start => [$html_parser_tag_sub, 
+   					 "tagname, '+1'"],
+			       end   => [$html_parser_tag_sub, 
+   					 "tagname, '-1'"],
+			       text  => [$html_parser_text_sub, 
+					 "text"]
+			      ],
+		  marked_sections => 1,
+		) ;
+		$parser->parse($html) ;
+		$parser->eof() ;
+		$html = $html_parser_text ;
+		$html =~ s/ / /ixsg ;
 		WebGUI::Macro::negate(\$html);
 	} elsif ($type eq "javascript") {
 		$html =~ s/\<script.*?\/script\>//ixsg;

t/HTML.t

@@ -22,6 +22,24 @@ use Test::More; # increment this value for each test you create
 my $session = WebGUI::Test->session;
 
 my @filterSets = (
+	{
+		inputText => q#<table border="0" cellpadding="0" cellspacing="0"><tbody><tr><td><img src="/uploads/abc.jpg" align="bottom" border="0" />&nbsp;</td><td style="padding-left: 10px">&nbsp;<!--StartFragment --> <center><b>ABC DEF</b></center><br />To provide cost<b> </b>efficient products that put you in c<b>ontrol</b> of your own future. <br /><br /><center><b>Our Company</b></center><ul class="BodyText"><li><a href="http://www.abc.com" target="_blank">ABC Company</a></li></ul><div align="center">Welcome to our Intranet! </div></td></tr></tbody></table>#,
+		output => q#    ABC DEF To provide cost efficient products that put you in control of your own future. Our Company ABC Company Welcome to our Intranet! #,
+		type => 'all',
+		comment => 'all filter complex HTML',
+	},
+	{
+		inputText => q!<tr style="height: 14px"><td style="border: medium none #d4d0c8; height: 14px; background-color: transparent" class="xl28"><b><font face="Arial" size="2">ABC</font></b></td><td style="border: medium none #d4d0c8; background-color: transparent" class="xl25"><b><font face="Arial" size="2"></font></b><br /></td><td style="border: medium none #d4d0c8; background-color: transparent" class="xl24"><font face="Arial" size="2">DEF</font></td><td style="border: medium none #d4d0c8; background-color: transparent" class="xl25"><font face="Arial" size="2"></font><br /></td><td style="border: medium none #d4d0c8; background-color: transparent" class="xl26"><b><font face="Arial" size="2">123</font></b></td></tr>!,
+		output => q!ABC DEF 123 !,
+		type => 'all',
+		comment => 'all filter complex HTML',
+	},
+	{
+		inputText => q!<div>Some text in a sentence.</div><div>Some more text in a sentence.</div>!,
+		output => q!Some text in a sentence. Some more text in a sentence. !,
+		type => 'all',
+		comment => 'all filter JT',
+	},
 	{
 		inputText => q!<p>Paragraph</p>^H();,^SQL("select * from users","^0;,^1;")!,
 		output => q!<p>Paragraph</p>&#94;H();,&#94;SQL("select * from users","&#94;0;,&#94;1;")!,
@@ -36,19 +54,19 @@ my @filterSets = (
 	},
 	{
 		inputText => q!<p>Paragraph</p>!,
-		output => q!Paragraph!,
+		output => q!Paragraph !,
 		type => 'all',
 		comment => 'all filter HTML',
 	},
 	{
 		inputText => q!<div class="something"><div style="float: left;">This <span>is some</span> <i>text</i> here. Let's&nbsp;see what we get when we add a table.<table><tr><td>a</td><td>b</td></tr><tr><td>c</td><td>d</td></tr></table></div><div>Here's a little more text with <b>bold</b>, <strong>strong</strong>, and <strong><b>bold strong</b></strong>.</div>!,
-		output => q!This is some text here. Let&#39;s&nbsp;see what we get when we add a table.abcdHere&#39;s a little more text with bold, strong, and bold strong.!,
+		output => q!This is some text here. Let's see what we get when we add a table. a b c d Here's a little more text with bold, strong, and bold strong. !,
 		type => 'all',
 		comment => 'all filter complex HTML',
 	},
 	{
 		inputText => q!<p>Paragraph</p>^H();!,
-		output => q!Paragraph&#94;H();!,
+		output => q!Paragraph &#94;H();!,
 		type => 'all',
 		comment => 'all filters macros and HTML',
 	},