lxml - keep input symbols, disable entity conversion

Question

If the following string is read and output using lxml, the umlauts are converted to entities.

import xml.etree.ElementTree as ET

root = ET.fromstring("<r><s>Die Häuser haben Dächer.</s></r>")
as_text = ET.tostring(root).decode("utf-8")
print(as_text)

Output:

<r><s>Die H&#228;user haben D&#228;cher.</s></r>

Expected output:

<r><s>Die Häuser haben Dächer.</s></r>

The umlauts are just an example. I generally want to disable entity conversions and instead keep the raw input symbols.

Can I disable entity conversion? Is there a safe method to reconvert the entities?

Answer 1

The default encoding used by tostring() is ASCII in both ElementTree and lxml.

To get the expected output, you can use encoding="unicode":

as_text = ET.tostring(root, encoding="unicode")
print(as_text)

References:

• https://docs.python.org/3/library/xml.etree.elementtree.html#xml.etree.ElementTree.tostring
• https://lxml.de/api/lxml.etree-module.html#tostring