Re: Algorithm akin to namespace fixup for xslt?

> Given a deep tree of nodes, each element node within a namespace. A child
node inherits its parent namespace unless it specifically overrides it. Is
there an algorithm for choosing the "best" or perhaps "sparsest" set of
overriding namespace declarations, hence taking the most advantage of
inheritance.
>
> My application of this algorithm isn't for namespaces but seems very similar
in requirement.
>
>
That's a pretty sketchy description, not helped by the fact that you appended
text from another thread which I assume is to be ignored.

Let's try another formulation, which might or might not be what you intended:
Given a document with lots of xml:lang attributes, eliminate those that are
redundant.

That one's fairly easy, I think:

<xsl:mode on-no-match="shallow-copy"/>
<xsl:template match="@xml:lang[. eq ancestor::*[@xml:lang][1]/@xml:lang"/>

Michael Kay
Saxonica