repoview: separate concerns in _filteredrepotypes comment
The cited issue in Python bugtracker is closed, but hasn't been
fixed. We've been able to use the attached example and reproduce
it with Python 3.9.
The point where it turns from needless stress on the GC to
the an actual leak is when one factors in the fact that the GC
was before Python 3.4 unable to collect some types (see PEP 442).
Note that even with Python 2.7, the simple example of cycles
due to __mro__ are collectable. This was seen again with the
example attached on the CPython issue.
from __future__ import absolute_import, print_function
# Import something from Mercurial, so the module loader gets initialized.
from mercurial import pycompat
del pycompat # unused for now
from hgext.lfs import pointer
def tryparse(text):
r = {}
try:
r = pointer.deserialize(text)
print('ok')
except Exception as ex:
print((b'%s' % ex).decode('ascii'))
if r:
text2 = r.serialize()
if text2 != text:
print('reconstructed text differs')
return r
t = (
b'version https://git-lfs.github.com/spec/v1\n'
b'oid sha256:4d7a214614ab2935c943f9e0ff69d22eadbb8f32b1'
b'258daaa5e2ca24d17e2393\n'
b'size 12345\n'
b'x-foo extra-information\n'
)
tryparse(b'')
tryparse(t)
tryparse(t.replace(b'git-lfs', b'unknown'))
tryparse(t.replace(b'v1\n', b'v1\n\n'))
tryparse(t.replace(b'sha256', b'ahs256'))
tryparse(t.replace(b'sha256:', b''))
tryparse(t.replace(b'12345', b'0x12345'))
tryparse(t.replace(b'extra-information', b'extra\0information'))
tryparse(t.replace(b'extra-information', b'extra\ninformation'))
tryparse(t.replace(b'x-foo', b'x_foo'))
tryparse(t.replace(b'oid', b'blobid'))
tryparse(t.replace(b'size', b'size-bytes').replace(b'oid', b'object-id'))