In xml.etree.ElementTree bytes tag or attributes raises on serialization

[py-user]

BPO	28237
Nosy	@scoder, @py-user

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = <Date 2019-09-10.16:09:54.847>
created_at = <Date 2016-09-21.12:19:15.151>
labels = ['expert-XML', 'type-bug', 'library']
title = 'In xml.etree.ElementTree bytes tag or attributes raises on serialization'
updated_at = <Date 2019-09-10.16:09:54.834>
user = 'https://github.com/py-user'

bugs.python.org fields:

activity = <Date 2019-09-10.16:09:54.834>
actor = 'scoder'
assignee = 'none'
closed = True
closed_date = <Date 2019-09-10.16:09:54.847>
closer = 'scoder'
components = ['Library (Lib)', 'XML']
creation = <Date 2016-09-21.12:19:15.151>
creator = 'py.user'
dependencies = []
files = []
hgrepos = []
issue_num = 28237
keywords = []
message_count = 2.0
messages = ['277128', '351721']
nosy_count = 2.0
nosy_names = ['scoder', 'py.user']
pr_nums = []
priority = 'normal'
resolution = 'wont fix'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'behavior'
url = 'https://bugs.python.org/issue28237'
versions = ['Python 3.6']

[py-user]

https://docs.python.org/3/library/xml.etree.elementtree.html#xml.etree.ElementTree.Element
"The element name, attribute names, and attribute values can be either bytestrings or Unicode strings."

The element name, attribute names, and attribute values can have bytes type, but they can't be serialized:

>>> import xml.etree.ElementTree as etree
>>>
>>> root = etree.Element(b'x')
>>> root
<Element b'x' at 0xb739934c>
>>>
>>> elem = etree.SubElement(root, b'y', {b'a': b'b'})
>>> elem
<Element b'y' at 0xb7399374>
>>> elem.attrib
{b'a': b'b'}
>>>
>>> etree.dump(root)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib/python3.3/xml/etree/ElementTree.py", line 1224, in dump
    elem.write(sys.stdout, encoding="unicode")
  File "/usr/lib/python3.3/xml/etree/ElementTree.py", line 826, in write
    qnames, namespaces = _namespaces(self._root, default_namespace)
  File "/usr/lib/python3.3/xml/etree/ElementTree.py", line 937, in _namespaces
    _raise_serialization_error(tag)
  File "/usr/lib/python3.3/xml/etree/ElementTree.py", line 1105, in _raise_serialization_error
    "cannot serialize %r (type %s)" % (text, type(text).__name__)
TypeError: cannot serialize b'x' (type bytes)
>>>
>>> etree.tostring(root)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib/python3.3/xml/etree/ElementTree.py", line 1171, in tostring
    ElementTree(element).write(stream, encoding, method=method)
  File "/usr/lib/python3.3/xml/etree/ElementTree.py", line 826, in write
    qnames, namespaces = _namespaces(self._root, default_namespace)
  File "/usr/lib/python3.3/xml/etree/ElementTree.py", line 937, in _namespaces
    _raise_serialization_error(tag)
  File "/usr/lib/python3.3/xml/etree/ElementTree.py", line 1105, in _raise_serialization_error
    "cannot serialize %r (type %s)" % (text, type(text).__name__)
TypeError: cannot serialize b'x' (type bytes)
>>>

Also attribute name can be serialized, but it holds the letter "b" and single quotes:

>>> import xml.etree.ElementTree as etree
>>> 
>>> e = etree.Element('a', {b'x': '1'})
>>> etree.tostring(e)
b'<a b\'x\'="1" />'
>>>

And same try with site package lxml works fine for all cases because it converts bytes to unicode strings right away:

>>> import lxml.etree
>>>
>>> root = lxml.etree.Element(b'x')
>>> root
<Element x at 0xb6ff00cc>
>>>
>>> elem = lxml.etree.SubElement(root, b'y', {b'a': b'b'})
>>> elem
<Element y at 0xb73a834c>
>>>
>>> elem.attrib
{'a': 'b'}
>>>

[scoder]

Arguably, writing out "b'x'" as attribute name instead of raising an exception isn't ideal. However, OTOH, I think it's reasonable to accept anything that is serialisable as a string, not just strings. That makes it difficult to draw a line.

In any case, it's ok for ElementTree to not allow bytes in general.

I'll close this issue, because I don't think it's worth being called a bug.