strip local variables in tracebacks
Project description
sensitive_variables - strip local variables in tracebacks
sensitive_variables
is a decorator you can apply to your functions to
prevent certain local variables from being read by debugging tools, such as the
Django crash reporter or Sentry.
Unlike Django's sensitive_variables
it is independent of the web framework
you use and also does not rely on debugging tools to know about the decorator
for things to work.
Usage:
Basic
from sentry_sdk import init
from sensitive_variables import sensitive_variables
init()
@sensitive_variables('password')
def login_user(username, password):
print("Logging in " + username + " with " + password)
# TypeError: unsupported operand type(s) for +: 'NoneType' and 'str'
login_user(None, "secret123")
results in:
Custom scrub function
sensitive_varibles
can receive a custom_scrub_fn parameter which will ba called for each local variable.
It receives the local value and variable name and must return value_has_changed, new_value
.
Where value_has_changed is a boolean which represents the value being changed or not and new_value is the new value.
You can use this to extend scrub for dictionaries and any other custom type.
Example:
from sentry_sdk import init
from sensitive_variables import sensitive_variables
init()
def my_scrub_fn(value, variable_name):
if variable_name == 'password':
return True, 'scrubbed-value'
return False, value
@sensitive_variables(custom_scrub_fn=my_scrub_fn)
def login_user(username, password):
print("Logging in " + username + " with " + password)
# TypeError: unsupported operand type(s) for +: 'NoneType' and 'str'
login_user(None, "secret123")
How does it work?
When the decorated function throws an exception, sensitive_variables
walks through the traceback, removes sensitive data from frame.f_locals
calling custom_scrub_fn so custom processing can be made and reraises the exception.
This is usually not problematic because a function that just threw an exception is unlikely to still use its local variables.
Why would I use this over Django's decorator?
Django has a decorator also called sensitive_variables
, which this package is inspired by. It sets an attribute on the function object that contains the variable names.
Debugging tools have to know about this attribute and respect it. For anything outside of the Django world, this is unlikely to be the case.
The decorator in this package will always work because it actually modifies your locals.
Why would I use this over Sentry's datascrubbing options?
-
This decorator does not couple your configuration for what is sensitive data to a specific crash reporting tool.
-
Behavior of the decorator is easily unit-testable (see
tests/
folder).
Why would I not use this?
This decorator inherently requires custom code for each Python implementation. Currently this is only tested against CPython 2.7, CPython 3.6, CPython 3.7, CPython 3.8 and PyPy 2.7.
License
Licensed under the MIT, see LICENSE
.
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file sensitive-variables-0.1.4.tar.gz
.
File metadata
- Download URL: sensitive-variables-0.1.4.tar.gz
- Upload date:
- Size: 5.7 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/3.1.1 pkginfo/1.5.0.1 requests/2.23.0 setuptools/45.2.0 requests-toolbelt/0.9.1 tqdm/4.46.0 CPython/3.7.5
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | c6cba828e3d40769dfdb38a41c20a27e2df3da8541f6934a52349a8250dad93f |
|
MD5 | 655af9a5b23ecb3129c437b3c7a5e5d9 |
|
BLAKE2b-256 | 1d85234bc922dd1b243240a2ce0cd88685980a6942429ad233667e11b1243437 |
File details
Details for the file sensitive_variables-0.1.4-py3-none-any.whl
.
File metadata
- Download URL: sensitive_variables-0.1.4-py3-none-any.whl
- Upload date:
- Size: 4.9 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/3.1.1 pkginfo/1.5.0.1 requests/2.23.0 setuptools/45.2.0 requests-toolbelt/0.9.1 tqdm/4.46.0 CPython/3.7.5
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 8534c557da3c3f03096a3eabbdecdf3a5b704430351bc237dcbc55081eef44df |
|
MD5 | e7ff19eb4546b0c54103f9ef15a21b68 |
|
BLAKE2b-256 | e1542b75f7c7ceb487d86be66e29a277ffc0ebc414eebb2d55b47abedae28ef4 |