Fix username anonymization #70
Conversation
mabulgu
changed the title
ansible_anonymizer/anonymizer.py
Outdated
| r"(?P<before>[c-z]:\\users\\)(?P<user_name>\w{,255})", | ||
| r"(?P<before>/(home|Users)/)(?P<user_name>[a-z0-9_-]{,255})", | ||
| r"(?P<before>[c-z]:\\users\\)(?P<user_name>([a-z0-9_-]|{{\s*.*?\s*}})\w{,255})", | ||
| r"(?P<before>/(home|Users)/)(?P<user_name>([a-z0-9_-]|{{\s*.*?\s*}})[a-z0-9_-]{,255})", |
There was a problem hiding this comment.
Why is [a-z0-9_-] for? We've got already a pattern to match the login name.
There was a problem hiding this comment.
If I dont put it with an or (|), it will match the jinjas but not the regular texts:
https://regex101.com/r/y2o0d9/1
However, upon your comment the second one seems redundant. removing it seems to be working: https://regex101.com/r/54R0ZI/1
Let me check with the code and update it if the tests work.
There was a problem hiding this comment.
On Windows, a login name can come with non ASCII characters. [a-z0-9_-] is more strict than just \w which accepts any Unicode characters.
>>> import re
>>> re.match(r"[a-z0-9_-]+", "aaa")
<re.Match object; span=(0, 3), match='aaa'>
>>> re.match(r"[a-z0-9_-]+", "ááá")
There was a problem hiding this comment.
On Windows, a login name can come with non ASCII characters.
[a-z0-9_-]is more strict than just\wwhich accepts any Unicode characters.>>> import re >>> re.match(r"[a-z0-9_-]+", "aaa") <re.Match object; span=(0, 3), match='aaa'> >>> re.match(r"[a-z0-9_-]+", "ááá")
@goneri I must have preserved that \w for the windows version. Note that the code updates are different than the above regex links, which I created before I understood your point better.
|
Kudos, SonarCloud Quality Gate passed! |
|
Thank you @mabulgu |
Bump version and add changelog for #70 fix
This PR fixes the username anonymization in a path by ignoring the Jinja templates in the name fields.
Fixes https://issues.redhat.com/browse/AAP-14989.