`elementtree` can abort on invalid XML bytes with UTF encoding

[StanFromIreland]

Crash report

What happened?

$ ./python -c "from io import BytesIO; from xml.etree.ElementTree import parse; parse(BytesIO(b'<?xml version=\"\" encoding=\"UTF\"?><a> \xd0 '))"
python: Objects/call.c:342: _PyObject_Call: Assertion `!_PyErr_Occurred(tstate)' failed.
Aborted                    (core dumped) ./python -c "from io import BytesIO; from xml.etree.ElementTree import parse; parse(BytesIO(b'<?xml version=\"\" encoding=\"UTF\"?><a> \xd0 '))"

Found by OSS-Fuzz in #517455264. I presumed this regression was introduced by 8ab7b43.

CC @serhiy-storchaka

CPython versions tested on:

CPython main branch

Operating systems tested on:

No response

Output from running 'python -VV' on the command line:

No response

Linked PRs

• gh-150560: Fix crash in XML parser on invalid XML with multi-byte encoding #150568

[vstinner]

The pyexpat_encoding_convert() callback is called twice:

• The first call fails with an exception and returns -1
• The second call fails with an assertion error since it's called with an exception set.