[RESOLVED] How to create a Table of Contents for a tagged HTML File?

tcarnes

I've got HTML object (generated by XML and XSL via XSLTProcessor) that comes with <a name=""> tags at all the major divisions. What I need to do is create a Table of Contents before I show the HTML.

I've got the major stuff working, in that I create an unordered list (which gets changed into a horizontal menu via CSS). Each item in the list is named from the

<a name="section-1"></a>

tags. This ends up giving me a list like:

<div id="nav">
  <ul>
    <li><a href="#section-1">Section-1</a></li>
    <li><a href="#section-2">Section-2</a></li>
  </ul>
</div>

What I want to end up with is something more like this:

<div id="nav">
  <ul>
    <li><a href="#section-1">1  INTRODUCTION</a></li>
    <li><a href="#section-2">2  INSTALLATION</a></li>
  </ul>
</div>

where the item names come from the

<h1>1  INTRODUCTION</h1>

tags that follow the

<a name="section-1"></a>

tags.

The HTML stream I'm working with looks like this:

<a name="section-5"></a><p></p>
<h1>2	INSTALLATION</h1>

I can't always guarantee the exact sequence of html nodes, but the <h1> node always follows closely behind the related <a name= ""> node. Sometimes there is the single <p></p> and sometimes a pair of breaks <br /><br />.

Here is my current code that creates an array of <a> nodes and then iterates through the ones with name="section-":

$proc = new XSLTProcessor();
$proc->importStylesheet($xsl);
$xml = new DOMDocument();
if (file_exists($xmlFile))
{
	$xml->load($xmlFile);
	$domTranObj = $proc->transformToDoc($xml);
	$linkPairs = array();
	$arrayAnodes = $domTranObj->getElementsByTagName('a');
	if (count($arrayAnodes) > 0) 
	{
		echo '<div id="nav"><ul>';
		foreach ($arrayAnodes as $anode) 
		{
			if ($nodeName = $anode->hasAttribute('name'))
			{
				if (($nodeNameVal = $anode->getAttribute('name')) &&
				    (strpos($nodeNameVal,'section') !== false) &&
				    (strpos($nodeNameVal,'.') === false)) /* I only want major divisions not any subdivsions*/
				{
					echo '<li><a href="#'.$nodeNameVal.'">'.ucwords($nodeNameVal).'</a></li>';
				}
			}
		}
		echo '</div>';
	}
}

Can anyone help me to get the contents of the <h1> nodes that follow each <a name=""> nodes?

Thanks for any help you can offer!

Weedpacket

You've got the source XML; wouldn't it be easier to generate the contents list from this (using XSLT, even)? It sounds like the markup would be more consistent.

But the first <h1> elements following each <a name=""> element would have an XPath specification of

//a[@name]/following-sibling::h1[1]

Take the list returned by that, iterate through its nodes, and get the textContent of each node.

tcarnes

Weedpacket;10911500 wrote:
You've got the source XML; wouldn't it be easier to generate the contents list from this (using XSLT, even)? It sounds like the markup would be more consistent.

The <a> tags are inserted during the XSLT process, so I'm forced to deal with the resulting HTML.

Thanks for pointing me to the XPath specs. Using that I was able to collect an array of <a name="???"> tags which I could then build into an unordered list that CSS made into an horizontal menu.

$proc = new XSLTProcessor();
$proc->importStylesheet($xsl);
$xml = new DOMDocument();
if (file_exists($xmlFile))
{
    $xml->load($xmlFile);
    $domTranObj = $proc->transformToDoc($xml);
    $linkPairs = array();
    $arrayAnodes = $domTranObj->getElementsByTagName('a');
    if (count($arrayAnodes) > 0)
    {
        echo '<div id="nav"><ul>';
        foreach ($arrayAnodes as $anode)
        {
            if ($nodeName = $anode->hasAttribute('name'))
            {
                if (($nodeNameVal = $anode->getAttribute('name')) &&
                    (strpos($nodeNameVal,'section') !== false) &&
                    (strpos($nodeNameVal,'.') === false)) /* I only want major divisions not any subdivsions*/
                {
			$xpath = new DOMXPath($domTranObj);
			$elements = $xpath->query('//a[@name="'.$nodeNameVal.'"]/following-sibling::h1[1]'); //find the trailing <h1> tags
			foreach ($elements as $bNode) 
			{
				 //create pairs of <h1>Titles and their matching <a> tags
				 //use trim to trim all leading and following whitespace, numbers, colons, periods, and commas
				$linkPairs[trim(ucwords($bNode->nodeValue)," \t\n\r\0\x0b0..9:.,")] = $nodeNameVal;
			}
		}
	}
	if (count($linkPairs) > 0) 
	{
		//Now that we have our pairs, let's build our unordered list:
		echo '<div id="labelnav"><a name="toc"></a><ul>';
		foreach ($linkPairs as $key=>$value) {
			if ($key <> '') {
				echo '<li><a href="#'.$value.'">'.$key.'</a></li>';
			}
		}
		echo '</ul></div>';
	}
}

Weedpacket

Glad you got it working.

tcarnes wrote:
The <a> tags are inserted during the XSLT process, so I'm forced to deal with the resulting HTML.

I figured that's where the <a> tags come from; but I was thinking of an additional XSLT process alongside the other one. Something vaguely like this:

$proc = new XSLTProcessor();
$proc->importStylesheet($xsl);

$listproc = new XSLTPrrocessor();
$listproc->importStylesheet('makecontentslist.xsl');

$xml = new DOMDocument();
if (file_exists($xmlFile))
{
    $xml->load($xmlFile);
    $domTranObj = $proc->transformToDoc($xml);
    $contents = $listproc->transformToDoc($xml);

Where $contents->asXML() would end up as "<div id='labelnav'>......</div>";

tcarnes

Weedpacket;10912071 wrote:
Glad you got it working.

Thanks!

I figured that's where the <a> tags come from; but I was thinking of an additional XSLT process alongside the other one....
Where $contents->asXML() would end up as "<div id='labelnav'>......</div>";

Yes, you are right. This is the right way to go. Although I have no idea how to start writing my own xsl file. What does that entail?