Re: [xsl] cleanup of <div>-elements

["Michael Kay michaelkay90@xxxxxxxxx" <xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx>]

> In XSLT it is often easy to implement if it is easy to define.

Indeed, this applies to any language. But XSLT does have an advantage here,
which is that you can incrementally add rules as you discover more and more
oddities in the input. If you're dealing with a large number of input files
then the big problem becomes quality control: if you've got more documents
than you can "eyeball", you're going to need to do some automated checking
that you haven't introduced corruptions. It might be an idea to make this
exercise test-driven: start by defining the requirements in terms of
executable tests that the output must satisfy, and only then start writing the
code to actually do the transformations.

Michael Kay
Saxonica