I'd written a Perl script to generate a table of contents from HTML pages which is working fine (and generating valid HTML) except for that the Perl output is removing closing tags for some elements like p. This is not validating against DocType of strict.
Please scroll down the post to see the Perl code.
What should I do to correct it?
#!/usr/bin/perl -w
#Copyright anurag gupta ; free to use under GNU GPL License
use strict;
use feature "switch";
use Common;
use HTML::Element;
use HTML::TreeBuilder;
#"F:/anurag/work/indiacustomercare/airtel/recharge.html";
my $filename="F:/tmp/t9.html";
my $index=0;
my $labelprefix="anu555ltg-";
my $tocIndex=100001;
my $toc;
my @stack;
my $prevHtag="h2";
sub hTagEncountered($)
{
my $hTag=shift;
my $currLevel=(split //, $hTag)[1];
given($hTag)
{
when(/h1/)
{
break;
}
default{
my $countCurr= (split /h/,$hTag)[1];
my $countPrev= (split /h/,$prevHtag)[1];
if($countCurr>$countPrev)
{
push @stack,($currLevel);
$toc.="<ul>";
}
elsif($countCurr<$countPrev)
{
# Now check in the stack
while ( @stack and $currLevel < $stack[$#stack])
{
pop @stack;
$toc.="</ul>";
}
}
}
}
$prevHtag=$hTag;
}
sub getLabel
{
my $name=$labelprefix.++$tocIndex;
}
sub traversehtml
{
my $node=$_[0开发者_运维问答];
# $node->dump();
# print "-----------------\n";
# print $node->tag()."\n";
# print ref($node),"->\n";
if((ref(\$node) ne "SCALAR" )and ($node->tag() =~m/^h[2-7]$/i)) #it's an H Element!
{
my @h = $node->content_list();
if(@h==1 and ref(\$h[0]) eq "SCALAR") #H1 contains simple string and nothing else
{
hTagEncountered($node->tag());
my $label=getLabel();
my $a = HTML::Element->new('a', name => $label);
my $text=$node->as_trimmed_text();
$a->push_content($text);
$node->delete_content();
$text=HTML::Entities::encode_entities($text);
$node->push_content($a);
$toc.=<<EOF;
<li><a href="#$label">$text</a>
EOF
}
elsif ( @h==1 and ($h[0]->tag() eq "a")) # <h1><a href="abc.com">ttt</a></h1> case
{
#See if any previous label already exists
my $prevlabel = $h[0]->attr("name");
$h[0]->attr("name",undef) if(defined($prevlabel) and $prevlabel=~m/$labelprefix/); #delete previous name tag if any
#set the new label
my $label=getLabel();
$h[0]->attr("name",$label);
hTagEncountered($node->tag());
my $text=HTML::Entities::encode_entities($node->as_trimmed_text());
$toc.=<<EOF;
<li><a href="#$label">$text</a>
EOF
}
elsif (@h>1) #<h1>some text here<a href="abc.com">ttt</a></h1> case
{
die "h1 must not contain any html elements";
}
}
my @h = $node->content_list();
foreach my $item (@h)
{
if(ref(\$item) ne "SCALAR") {traversehtml($item); } #skip scalar items
}
}
die "File $filename not found" if !-r $filename;
my $tree = HTML::TreeBuilder->new();
$tree->parse_file($filename);
my @h = $tree->content_list();
traversehtml($h[1]);
while(pop @stack)
{
$toc.="</ul>";
}
$toc="<ul>$toc</ul>";
print qq{<div id="icctoc"><h2>TOC</h2>$toc</div>};
my @list1=$tree->content_list();
my @list2=$list1[1]->content_list();
for(my $i=0;$i<@list2;++$i){
if(ref(\$list2[$i]) eq "SCALAR")
{
print $list2[$i]
}
else{
print $list2[$i]->as_HTML();
}
}
# Finally:
Try passing {}
for the \%optional_end_tags
argument to as_HTML
. See the documentation for details.
精彩评论