Look down to the bottom of the synopsis in the HTML:TableExtract docs for everything you need. I just ran this

#!/usr/bin/perl -w

use strict;
use warnings;
use HTML::TableExtract qw(tree);
use Data::Dumper;

my $table_string = <<EOF;
<table width="100%" bgcolor="#ffffff">
  <tr>
    <td>Larry &amp; Gloria</td>
    <td>Mountain View</td>
    <td>California</td>
  </tr>
  <tr>
    <td><b>Tom</b></td>
    <td>Boulder</td>
    <td>Colorado</td>
  </tr>
  <tr>
    <td>Nathan &amp; Jenine</td>
    <td>Fort Collins</td>
    <td>Colorado</td>
  </tr>
</table>
EOF


 my $te = HTML::TableExtract->new(keep_html=>1);
 $te->parse($table_string);
 my $table = $te->first_table_found;
 my $table_tree = $table->tree;
my  $table_html = $table_tree->as_HTML;
my  $table_text = $table_tree->as_text;
 
 print Dumper($table_html),"\n";
 print Dumper($table_text),"\n";
[download]

I was too lazy to dig up a table earlier, but ran across one I could paste in while I was reading so I went ahead and tested it. If you change the keep_html=>1 to keep_html=>0 then the as_HTML will strip all the markup except the table tags, and the as_text will strip out all the table tags, too.

So, back to HTML::TableExtract -- the documentation makes it very clear that keep_html SHOULD keep the markup.

Thus, the question becomes, could there be an error in your code? We won't know the answer to that until you post it... and a bit of information about how you're acquiring the table in the first place.

Reading the same docs it looks like keep_html keeps any markup within the cell (e.g. text formatting) but not the tags that define the table. That it doesn't keep embedded tables within a cell makes me think it removes all table stucture tags

The synopsis does give a different method that looks like the right thing through:

$table_html = $table_tree->as_HTML;

I haven't tried either one- maybe later on today if the op doesn't solve it.

We appear to read the 2 line para on keep_html differently. IMO, it's open to multiple readings, but the next para (re strip...) seems more to support your view than mine.

So, I bet we can agree without dissent that the doc needs some improvement.

That said, let's think about Parse::HTML. It won't do the whole job, by itself, for OP (but as noted above, we don't know how OP is doing whatever led to the SOPW), but with no added code... and just a little additional, and a few tweaks to the code in the example entitled "The Identity Parser, it should be no great problem achieve OP's objective.

use HTML::TableExtract qw(tree);
my $te = HTML::TableExtract->new();
$te->parse($string_of_html);

my $table = $te->first_table_found;
my $table_tree = $table->tree;
my $table_html = $table_tree->as_HTML;
print $table_html;
[download]