ansaurus

Question

Answer 1

+2 A:

You can use XSLT to manipulate HTML assuming the HTML is well formatted (as in the HTML document is a well-formed XML document in the strictest sense).

If you can confirm this, and your XSLT isn't working, maybe you should provide a more thorough snippet of both the HTML and XSLT documents so that we can figure it out.

Welbog 2009-10-28 19:48:41

Answer 2

+2 A:

As long as your XHTML document is well-formed, an XML parser will be able to read it, and therefore an XSLT engine will be able to transform it.

Assuming that, the most common causes of not being able to find elements in a document are:

Your XPath expression is being executed relative to a different node that what you thought it was going to be. What this means for your XSLT - check that your XSLT match patterns are correct relative to their templates.
You have not defined the namespace URI-to-prefix mappings in your XPath engine. What this means for your XSLT - make sure you have the xmlns="http://www.w3.org/1999/xhtml" namespace declared in your XSLT file, with or without a prefix.

If you post your XSLT I will be able to comment further.

Christian Hayter 2009-10-28 19:57:47

Answer 3

+3 A:

You're not using XPath correctly in your match attributes. You need the xmlns:xhtml="http://www.w3.org/1999/xhtml" attribute in your xsl:stylesheet element, and then you'll need to use the xhtml: prefix in your XPath expressions (you need a prefix; XPath does not obey default namespaces).

After this, you'll still get the problem that it will process everything else too. I don't know if there's a better solution to this, but I think you will need to explicitly process things on the path to the tbody element, something like

<xsl:template match="xhtml:html">
  <xsl:apply-templates select="xhtml:body"/>
</xsl:template>

and the same thing for body and so on until you get to your tbody match.

XPath also supports more complex matching than just a specific child as above. For instance, matching the third child div tag can be done with

<xsl:template match="xhtml:div[3]">

and matching an element with a specific attribute with

<xsl:template match="xhtml:div[@class='IMPORTANT']">

Here the [] surrounds an additional condition that needs to be fulfilled for the element to be considered a match. A plain number means to index into the matches and take only the one that has that index (the indexing is 1-based), an @ sign precedes an attribute, but you can have arbitrarily complex XPath in there, so you can match pretty much any substructure you'd like.

jk 2009-10-28 20:49:04

Thanks, that helped. Is there no way to search for an inner tag directly? either by just stating its name or full path?

gonso 2009-10-28 20:54:39

You can match it directly, but then you'll get the default behavior for everything else, which is to print their text content. And if you try to use `<xsl:template match="*"/>` to eliminate the default behavior, that's going to affect your root element too, so it won't process the document at all. So if you want processing on only a single element in the document, I think you do need to override the default behavior on all elements along the path to that element.

jk 2009-10-28 21:02:59

Almost there. I've added the HTML schema. My question is: How to dive into the right "div" tag. There are several, but Im just interested in some of them. Can I filter by occurrence (i.e. the third div tag) or by an attribute (i.e. div tag where class="IMPORTANT")?thanks again

gonso 2009-10-28 23:51:15

@jk, an easier approach is to do `<xsl:template match="*|text()"><xsl:apply-templates/></xsl:template>` - this removes all output for default rule, but ensures that the entire tree is processed (so other rules may still match).

Pavel Minaev 2009-10-29 07:37:54

@Pavel, Thanks! that really simplified my xsl. May I ask you how/why it works? How do you read the "*|text()" condition? Any tag or the text inside any tag?If that is the case, how come this rule doesn't take precedence over other tag-specific rules I've created? Like this: <xsl:template match="*|text()"> <xsl:apply-templates/> </xsl:template> <xsl:template match="xhtml:tbody"> process tbody </xsl:template></xsl:stylesheet>Thanks!

gonso 2009-10-29 10:41:27

ansaurus

tags:

views:

answers:

Can I parse an HTML using XSLT?

related questions