xpath: How do we select just the very last text node?

Question

How do I select the globally-last text node using xpath? I tried this, but it gives me the last node in every context of the document.

lxml.html.fromstring('1<a>2<b>3</b>4<c>5</c>6</a>').xpath('//text()[last()]')
['1', '3', '5', '6']

I can do this, but it's inefficient in both time and space, especially as the document gets large.

lxml.html.fromstring('1<a>2<b>3</b>4<c>5</c>6</a>').xpath('//text()[last()]')[-1]
'6'

I tried to use an index of -1, but that gives me an empty list. I tried to use some of the reverse axes (so that I could index with 1), but I couldn't get them to work in a global context.

Answer 1

Use:

(//text())[last()]

This is a FAQ.

Explanation:

The XPath pseudo-operator // has a lower precedence (priority), than the [] operator, which binds stronger to the node-test preceding it.

As in Math and in every language, the way to override default priority is by using brackets.