This repository contains the source code for a profane content filter 🤬.

The GitHub Action: Profane content filter maintains over 4,900 swear words from nine different languages. For a list of all supported languages and swear words, see the data directory raw newline-delimited text files with each corresponding language's alphabetically sorted swear word list. This tool is used to scan issues, pull requests and comments in either for profanity. It can be configured to replace any profane content with a strategy and renders a complete job summary for tracking profanity filter results. This action can be useful for maintaining a professional and respectful environment in your GitHub repository.

But why is this important? Let's be honest, not everyone who creates issues or pull requests use appropriate language (it's not always rainbows and ponies, am I right?)

The following is an example of how to use the action as a standalone workflow:

# The name of the workflow
name: Profanity filter

# Trigger on issue or pull requests, that are opened, edited, or reopened
on:
  issue_comment:
    types: [created, edited]
  issues:
    types: [opened, edited, reopened]
  pull_request:
    types: [opened, edited, reopened]

# Required permissions
permissions:
  issues: write
  pull-requests: write

jobs:
  # Name the job whatever you'd like
  apply-filter:

    runs-on: ubuntu-latest

    steps:

    # Name the step anything that makes sense to you
    - name: Scan issue or pull request for profanity
      # Conditionally run the step if the actor isn't a bot
      if: ${{ github.actor != 'dependabot[bot]' && github.actor != 'github-actions[bot]' }}
      uses: IEvangelist/profanity-filter@main
      id: profanity-filter
      with:
        token: ${{ secrets.GITHUB_TOKEN }}
        # See https://bit.ly/potty-mouth-replacement-strategies
        replacement-strategy: Emoji # See Replacement strategy

If you already have an existing workflow that is triggered on/issues|pull_request/types/opened|edited|reopened feel free to simply add a step to the existing job:

- name: Scan issue or pull request for profanity
  # Conditionally run the step if the actor isn't a bot
  if: ${{ github.actor != 'dependabot[bot]' && github.actor != 'github-actions[bot]' }}
  # Use the profanity filter action
  uses: IEvangelist/profanity-filter@main
  id: profanity-filter
  with:
    token: ${{ secrets.GITHUB_TOKEN }}
    # See https://bit.ly/potty-mouth-replacement-strategies
    replacement-strategy: FirstLetterThenAsterisk

This action has several inputs. Only the token is required, and the replacement-strategy defaults to asterisk when not specified.

The following table describes each input:

Each replacement strategy corresponds to a different way of replacing profane content. The following represents the available replacement types:

The value of the replacement-strategy input is case-insensitive, and accepts hyphened alternatives (such as anger-emoji) as well.

This action will look for a label with the following verbatim name "profane content 🤬", if found this label is applied to any issue or pull request where profane content filtration occurs.

Consider the following automatically applied label to an issue that contains profane content:

When profane content is detected, the action will update the issue or pull request by:

• Replacing any found profane content with the configured replacement strategy.
• Reacting to the issue or pull request with the confused 😕 reaction.
• Conditionally applying the profane content 🤬 label if found in the repository.
• Reporting the profane content in the workflow summary as a detailed table.

Consider the following diagram, that represents the workflow of this GitHub action:

flowchart TD
    A(Issue or Pull Request)
      --> B[Contains Profane Content?]
    B -->|YES| C(Apply Filter)
      --> E(All swear words are filtered, for example sw**r)
      --> F(React to and label profane content)
      --o G[Job Summary]
    B --o|NO| D{{Stop}} ~~~A

The following is an example of a job summary that is rendered when profane content is detected: