On 09.03.2022 14:43, rick@xxxxxxxxxxxxxx wrote:
>
> Thank you Martin and Wendell.B  I am incorporating this one:
>
> normalize-space(string-join(text()))=''
>

In your example, with the input beingB  <root>This has <b>some
</b>text.</root> and your template matching "root", do you want to
select only the text node children (i.e. "This has " and "text.") or all
text node descendants (i.e. "This has ", "some ", and "text.")?