Code refactor

[paescuj]

Various optimizations:

• Check all inputs before doing any API requests
• Stricter check for doNotSkip input
• Exclude current workflow run from list of all workflow runs
• Get rid of redundant and unnecessary lines
• Add some explaining comments
• Fix bug with empty head_repository
• Action might be a bit faster 🚀

Breaking changes:

• Previously, on input and API request errors the action might have exited successfully but with outputs should_skip = false and reason = no_workflow_information. Now, the action will exit with an error instead.

  To make sure that subsequent jobs will not be skipped in such cases it's recommended to set continue-on-error: true in the skip-duplicate-actions job.
  Note: This has already been recommended in the past regardless of this change.

• The structure of the skipped_by output (also in paths_filter) has slightly changed for the sake of consistency and alignment with GitHub API:

  • runId is now id
  • html_url is now htmlUrl
  • Full example:

    {
      "id": 3099801293,
      "runNumber": 930,
      "event": "pull_request",
      "treeHash": "179eceb71f2caf9f60b93d00818cc48ef41e057f",
      "commitHash": "9c6c638244dbc141ef254b676e80cb8beaa63be0",
      "status": "completed",
      "conclusion": "success",
      "htmlUrl": "https://github.com/fkirc/skip-duplicate-actions/actions/runs/3099801293",
      "branch": "code-refactor",
      "repo": "paescuj/skip-duplicate-actions",
      "workflowId": 2640563,
      "createdAt": "2022-09-21T17:21:55Z"
    }

Note: This refactoring also took place to be better able to address #264 (will follow in a separate pull request). Currently unsolvable...

[fkirc]

The new code looks clean, I did not have the opportunity for detailed tests, but I think we should merge it and let the community test.

[paescuj]

The new code looks clean, I did not have the opportunity for detailed tests, but I think we should merge it and let the community test.

Thank you very much for your quick review!
I don't expect many users to be affected by those breaking changes, anyway...