Configure Codacy to stop flagging valid "assert" statements

[cclauss]

I'm submitting a ...

• bug report
• feature request

Current behavior:
Put a valid assert statement in Python code and Codacy will flag it:

• Use of assert detected. The enclosed code will be removed when compiling to optimized byte code.
  Pytest recommends the use of asset statements. The fact that the code will be removed is actually a feature, not an issue. We want to encourage the addition of new pytests.

Expected behavior:

• Codacy does not complain about valid Python.

Steps to fix:
https://support.codacy.com/hc/en-us/articles/207279979-Issues#2-remove-pattern

NOTES:

On Mon, Oct 30, 2017 at 12:52 PM, "Ruxandra Barbu" <ruxandra.barbu@codacy.intercom-mail.com>
wrote: 
Hi, thank you for contacting us.
Flagging the Python assert statements is a function of the Bandit linter itself.
You can ignore this issue in Codacy so it won´t affect the project. Please click here for
more details on how to ignore an issue/remove a pattern.
Please let me know if you have any other questions.

On Mon, Oct 30, 2017 at 01:55 AM, "cclauss" wrote: 
Why does Codacy flag the Python assert statements that pytest recommends? I know about the
optimization that eliminates the asserts commands from production code but 1) no one ever
does that and 2) if I did it, I would consider it a feature not a problem. Anyway, I how do
I configure Codacy to stop flagging the asserts that Pytest recommends?

[Dragneel7]

@cclauss I am in interested in working on this issue.

[cclauss]

That is great news! It would help a lot if adding new pytests was viewed as a positive contribution by Codacy. This is what their support folks emailed to me about “removing the assert pattern”::

Hi, thank you for contacting us.
Flagging the Python assert statements is a function of the Bandit linter itself.
You can ignore this issue in Codacy so it won´t affect the project.
Please click here for more details on how to ignore an issue/remove a pattern.
Please let me know if you have any other questions.

--
Ruxandra	Ruxandra from Codacy

https://support.codacy.com/hc/en-us/articles/207279979-Issues#2-remove-pattern

The process seems simple enough but it must be done by someone who has admin privileges of the project’s Codacy account.

[Dragneel7]

@cclauss will it be possible to work without "admin privileges"?

[cclauss]

I tried unsuccessfully to do what that web page suggested.

@mariobehling Do you know who has Codacy admin privileges for this repo?

[cclauss]

@vaibhavsingh97 Do you have admin rights to help solve issue?

[cclauss]

I found the solution!! Implemented in #385 Create a bandit.yml file at the root of the repo that contains:

skips: ['B101']

[toonarmycaptain]

It seems to me that the primary issue here is distinguishing between tests/testing code and project/production code. I don't want to turn off the checking for asserts in my project code, but I do want the asserts used (sensibly) in my tests to pass checks for asserts silently.

Seems to me a solution would be to have a separate category for testing code, or code that pytest would run as tests, so I can turn on/off the "Assert used" pattern separately for testing/main project code.

[manoadamro]

I actually stopped using codacy because of this, It was so frustrating that it couldn't figure out a file called tests/test_some_thing was a test file... I had 277 issues, all of them due to the use of assert.
surely any python file prefixed with test_ is a test file right? Especially if it's in a folder called tests

[toonarmycaptain]

@manoadamro Well unless you're working on assessments or worksheets maybe (eg test_templates, test_solutions), but once you run into this issue you'd think you'd change the filenames, or have all your tests under test_suite,
You can then do:

exclude_paths:
    - 'test_suite/**'

In your .codacy.yml file.

[agriyakhetarpal]

This comes up very high in a Google search, so I'm posting this here just in case this helps others:

One may add

assert_used:
  skips: ['*_test.py', '*test_*.py']

to their bandit.yml, which should help flag statements selectively.