pycqa/flake8-json

JSON formatter for Flake8 output

flake8-json's Issues

Update codeclimate formatter to never show source

The codeclimate formatter breaks if the show source option is activated. This option would previously insert the source line, where the error occured as raw text into the json. This leads to parsing errors, when reading the generated json output. With the proposed changes, the source line will never be inserted.

Added severity field to meet GitLab spec - [merged]

In GitLab by @iainvm on Jul 26, 2021, 16:56

Merges 2-severity-field -> master

Closes #2

Fix json output when filenames contain special characters - [merged]

In GitLab by @asottile on Aug 11, 2019, 21:27

Merges invalid_filenames -> master

Before:

$ flake8 ../x/foo\"bar\".py --format=json | jq .
parse error: Invalid numeric literal at line 1, column 15

After:

$ flake8 ../x/foo\"bar\".py --format=json | jq .
{
  "../x/foo\"bar\".py": [
    {
      "code": "F401",
      "filename": "../x/foo\"bar\".py",
      "line_number": 1,
      "column_number": 1,
      "text": "'os' imported but unused",
      "physical_line": "import os\n"
    }
  ]
}

Report violations as noqa'd when using --disable-noqa

In GitLab by @pawamoy on Apr 6, 2021, 08:22

Hello @asottile, thanks for all the work you've done on flake8 and plugins :slight_smile:

I was wondering: does flake8 provide the necessary information for flake8-json to report a violation as disabled with a noqa comment, when using flake8's --disable-noqa option? Knowing if a violation is disabled would help generating reports without having to run flake8 twice (once normally, once by disabling noqas) to compute the difference, or without having to parse the noqa comments ourselves.

If flake8 itself does not provide that information, then I'll see if I can open a ticket on flake8's bugtracker instead :slight_smile:
Maybe you'd also like to share your opinion on the subject? Does what I'm trying to achieve make sense?

Pretty-printing JSON?

By default, the output is not pretty-printed. We would love to add that as a separate formatter option, though.

Hello, are you still interested in adding support for pretty-printing the output?
If so, I'd be happy to contribute this option :)

Add CodeClimate formatter - [merged]

In GitLab by @sigmavirus24 on Dec 1, 2020, 17:55

Merges code-climate -> master

The CodeClimate specification lists out the fields:
https://github.com/codeclimate/platform/blob/master/spec/analyzers/SPEC.md#data-types
the biggest concern I have is the fingerprint generation function
here.

Namely, I'm worried that using the line_number means this
functionality for CodeClimate won't work as expected.

Severity missing in codeclimate

In GitLab by @iainvm on Jul 26, 2021, 16:53

If the severity field was added to the codeclimate format, the output would meet the requirements to be used for GitLab's Code Coverage
https://docs.gitlab.com/ee/user/project/merge_requests/code_quality.html#implementing-a-custom-tool

Fix broken codeclimate format - [merged]

In GitLab by @NaughtMeta on Jan 18, 2021, 16:15

Merges metabytez-master-patch-71623 -> master

• Update invalid class reference in super call for CodeClimateJSON.start
• Update CodeClimateJSON._fingerprint to encode string data first, and return the hexdigest