I'm generating some HTML using ElementTree in a Python script. The relevant part looks like this:
separator = ET.SubElement(toc, 'span')
separator.set('class', 'separator')
separator.text = u" ⬩"
The problem is that the HTML entity is being encoded in the resulting XML output as  , which puts the literal text on the rendered web page, instead of inserting a non-breaking space.
I'm working around it by escaping a non-breaking space into the Python string like this: u"\u00A0⬩". This puts a literal non-breaking space in my HTML instead of the entity. Ultimately, it works because it is faithfully rendered as a non-breaking space, but it makes the source code hard to read because a non-breaking space looks like a regular space.
How can I get ElementTree to insert an HTML entity into my source code?
&characters are replaced by&during serialization (github.com/python/cpython/blob/main/Lib/xml/etree/…). See also stackoverflow.com/q/9276848/407651. lxml behaves in a similar way: stackoverflow.com/q/67299346/407651.