PYTHON-4468 Hide the value of sensitive subtype binary objects #1649
Conversation
TerryRPatterson
force-pushed
the
redact_sensitive_bytes
branch
2 times, most recently
from
0a1b266 to
8e52e2a
Compare
TerryRPatterson
force-pushed
the
redact_sensitive_bytes
branch
from
8e52e2a to
4bb92c2
Compare
There was a problem hiding this comment.
Thanks @TerryRPatterson, do you mind adding a mention of this change in the doc/changelog.rst?
bson/binary.py
Outdated
| @@ -364,4 +364,7 @@ def __ne__(self, other: Any) -> bool: | |||
| return not self == other | |||
|
|
|||
| def __repr__(self) -> str: | |||
| return f"Binary({bytes.__repr__(self)}, {self.__subtype})" | |||
| if self.__subtype == SENSITIVE_SUBTYPE: | |||
| return f"Binary('<redacted>', {self.__subtype})" | |||
There was a problem hiding this comment.
Since this is no longer an accurate representation of the object we need to surround it with <> so that repr+eval() will raise an error instead of silently changing the data, like this:
return f"<Binary(REDACTED, {self.__subtype})>"
Reference: https://docs.python.org/3/library/functions.html#repr
There was a problem hiding this comment.
I added an entry to the changelog. Since the the jira is tagged Fixes 4.7.3 I assumed that is the release. I don't know how to get a link to the release notes from jira so I didn't include the link.
TerryRPatterson
force-pushed
the
redact_sensitive_bytes
branch
from
4bb92c2 to
87c80c2
Compare
|
FYI since this is a minor breaking behavior change it can't go into 4.7.3 and has to wait for 4.8. |
I updated the changelog. |
PYTHON-4468
Summary
I noticed the
bson.binary.Binarytype include it'svalue even when the subtype is sensitive (8)
Changes in this PR
This PR changes the Binary type's repr to be
<Binary ### {self.__subtype}>when the subtype is 8.The
< >representation follow general guideline that when a repr can't be roundtripped byevalit should use< >instead of a constructor like syntax.