Get XML nodes specified by attributes

S3NTIN3L

Below is my XML. I need to get each tag and format it in HTML. I am using PHP with foreach loops to do this as shown below:

<blog>
<title>H.M.S Endeaver</title>
<date>2011-06-03</date>
<body>text</body>
<tags>
<tag id="1">H.M.S Endeaver</tag>
<tag id="2">model-boat</tag>
</tags>
</blog>

$doc = new DOMDocument(); 
$doc->load($file); 								  

$dataset = $doc->getElementsByTagName("blog"); 

foreach($dataset as $row) 
{ 
	$title = $row->getElementsByTagName("title")->item(0)->nodeValue;
	$date = $row->getElementsByTagName("date")->item(0)->nodeValue; 
	$body = $row->getElementsByTagName("body")->item(0)->nodeValue; 									
	$tags = $row->getElementsByTagName("tags"); 

.....

foreach($tags as $tag) 
{
	$tags = $tag->getElementsByTagName("tag")->item(0)->nodeValue;
	echo "<a href='tag'>$tags</a>";
}
}

The problem is I dont know how to get all the different tag elements in the second foreach loop - would I use the attributes (id)?

Also is this the best approach to take i.e. could I have coded this better?

Thanks anyone

Weedpacket

Have a look at the [man]DOMXPath[/man] methods. Using GetElementById won't work; just because an attribute's name is "id" doesn't mean its type must be ID.

S3NTIN3L

Do you think it would be sloppy to change:

<tag id="1">H.M.S Endeaver</tag>
<tag id="2">model-boat</tag>

To:

<tag_1>H.M.S Endeaver</tag_1>
<tag_2>model-boat</tag_2>

And use a for loop with this inside:

"tag_".$i

Because I cannot find how to manipulate the attributes from your post.

Cheers

Weedpacket

I think yes. It would be sloppy. If you can't handle attributes (where an XPath expression like [font=monospace]"//tag[@id='1']"[/font] would select all [font=monospace]<tag id="1">[/font] elements), then you'd be better off with something like [font=monospace]<tag><id>1</id>HMS Endeavour</tag>[/font].

johanafm

And as long as you're dealing with ids, you really should know what element it relates to. There can be only one id of a certain value across the entire document. So getElementById('1') will give the same result as xpath //tag[@id='1']. Because if there may be a tag element with id attribute of '1', then there cannot be some other element with id '1'. Unless of course you assign numbers as ids to all your elements regardless of what they are using the same counter. But that seems unlikely.

And while you're at it, XML id attribute must match the NCName specification, which states that it cannot start with a digit. It must in fact start with a letter or underscore

http://www.w3.org/TR/1999/REC-xml-names-19990114/#NT-NCName wrote:
[4] NCName ::= (Letter | '') (NCNameChar) / An XML Name, minus the ":" */
[5] NCNameChar ::= Letter | Digit | '.' | '-' | '' | CombiningChar | Extender

Weedpacket

johanafm wrote:
There can be only one id of a certain value across the entire document.

Not true; see my previous reply but one. While strongly recommended that if you have an attribute named "id" then its values should be ID typed (and therefore unique throughout the document), and vice versa, it's not required.

getElementById() will not work without a formal definition of the XML document structure that specifies which attribute it should use.

Because the formal definition might not be available, and because some XML dialects actually forbid formal definition of their structure, the xml namespace was given its own [font=monospace]xml:id[/font] attribute. getElementById() will work with that.

johanafm

You are of course correct.