nanlabs / automation-seed Goto Github PK

Describe the issue

I am having issues getting yarn test to run.
I probably don't have all the dependencies ?
505 5 $ yarn test:local
yarn run v1.22.19
error [email protected]: The platform "win32" is incompatible with this module.
error Commands cannot run with an incompatible environment.
info Visit https://yarnpkg.com/en/docs/cli/run for documentation about this command.
515 15 $ fnm --version
fnm 1.35.1

504 4 $ npm --version
10.2.0

505 5 $ node --version
v18.17.1

506 6 $ yarn --version
1.22.19

additional observations:
/c/repo/automation_seed_webdriverIO___main__ (main)
516 16 $ npm install
npm ERR! code EBADPLATFORM
npm ERR! notsup Unsupported platform for [email protected]: wanted {"os":"linux,darwin","cpu":"x64,arm64"} (current: {"os":"win32
","cpu":"x64"})
npm ERR! notsup Valid os: linux,darwin
npm ERR! notsup Actual os: win32
npm ERR! notsup Valid cpu: x64,arm64
npm ERR! notsup Actual cpu: x64

npm ERR! A complete log of this run can be found in: C:\Users\kmartin\AppData\Local\npm-cache_logs\2023-10-18T23_03_44_742Z-debug-0.
log

Links

README.md ??

Blocked by: #4

Place a file at bin/suite-spec-paths.ts with the following content:

import glob from 'glob';
import { config } from '../config/wdio.suites.conf';

export const getSuiteSpecFiles = (
  globs: string | string[],
): string[] => {
  const files = Array.isArray(globs) ? globs : [globs];
  const specs = files.reduce((acc, file) => [...acc, ...glob.sync(file)], [] as string[]);
  return specs;
};

const suiteName = process.argv[2];

if (!suiteName) {
  process.exit(0);
}

if (!config.suites) {
  process.exit(0);
}

const suite = config.suites[suiteName];

if (!suite) {
  process.exit(0);
}

const suiteChunk = getSuiteSpecFiles(suite);

for (const spec of suiteChunk) {
  console.log(spec);
}

and add a script in the package json as follows:

"suite:files": "ts-node ./bin/suite-spec-paths.ts"

Describe the feature

I propose switching our CI linting setup, which includes shellcheck, markdownlint, eslint, and others, to use super-linter exclusively for GitHub Actions. super-linter is a comprehensive and centralized linter that supports multiple languages, providing a unified linting experience.

Use Case

There are some benefits of super-linter:

• Unified Linting: super-linter supports a wide range of languages and linters, providing a unified linting solution for our GitHub Actions workflow.

• Easy Configuration: Simplifies configuration by having a single .github/linters configuration file.

• Consistent Output: Standardizes the output format for linting results across different languages and linters.

• Extensibility: Supports adding or customizing linters easily, making it adaptable to our specific requirements.

Proposed Solution

1. Configure super-linter:

   • Create or update the .github/workflows configuration file to specify the linters and configurations needed.

2. GitHub Actions Workflows:

   • Update the GitHub Actions workflow files to use super-linter for linting instead of the actual implementation

3. Testing:

   • Thoroughly test the linting process with super-linter in the GitHub Actions workflow to ensure it covers all relevant languages and provides accurate results. You can test this by submitting a PR

Other Information

No response

Acknowledgements

• I may be able to implement this feature request
• This feature might incur a breaking change

Version used

latest

Environment details (OS name and version, etc.)

any

We already have a workflow to convert all the TODOs to issues here. We need to have a similar behavior for this repo

Blocked by: #5

Update the suite execution in the reusable workflow to support multiple splits using the following code as reference:

- name: Execute test Suite
  run: |
    SPEC_FILES=$(yarn --silent suite:files ${{ inputs.suite }} | sort | awk "NR % $CI_TOTAL_JOBS == $CI_JOB_INDEX")
    echo "Number of specs to be executed: $(echo $SPEC_FILES | wc -w)"
    [[ -z "$SPEC_FILES" ]] && exit 0
    echo "Running specs: $SPEC_FILES"
    echo $SPEC_FILES | xargs yarn test:remote --spec
  if: ${{ inputs.suite != '' }}
  env:
    OUTPUT_DIR: '../../logs'
    WEBDRIVER_LOGLEVEL: ${{ inputs.log_level }}
    BROWSER_VISIBLE: ${{ inputs.browser_visible }}
    ENABLE_VIDEO: ${{ inputs.enable_video }}
    MAX_INSTANCES: ${{ inputs.max_instances }}
    CHROME: ${{ inputs.chrome_instances }}
    FIREFOX: ${{ inputs.firefox_instances }}
    WEBDRIVER_SPEC_FILE_RETRIES: 1
    CI_JOB_INDEX: ${{ inputs.spec_ci_index }}
    CI_TOTAL_JOBS: ${{ inputs.spec_ci_total }}

We have different workflows that are doing exactly the same. That is why we need to create a separate action and call it on each existing workflow.

• Create a workflow at .github/workflows/automation-call.yml with the workflow_call trigger

The new workflow should have the following header:

permissions:
  contents: write

on:
  # This will allow us to reuse this workflow from other repos
  workflow_call:
    inputs:
      ref:
        description: Specifies from which ref to execute the automation.
        required: false
        type: string
        default: ""

      log_level:
        description: Specifies the log level
        required: true
        type: string
        default: 'info'

      suite:
        description: Suite to be executed
        default: ''
        type: string
        required: false

      spec:
        description: Spec to be executed. (Won't be used if suite is specified)
        default: ''
        type: string
        required: false

      spec_ci_index:
        description: Specifies the index of the spec in the CI environment
        default: 0
        type: number
        required: false

      spec_ci_total:
        description: Specifies the total number of specs in the CI environment
        default: 1
        type: number
        required: false

      max_instances:
        description: Maximum number of instances to be executed
        default: ''
        type: string
        required: false

      chrome_instances:
        description: Amount of Chrome Instances to be used. Won't use Chrome when its value is a blank string
        default: '1'
        type: string
        required: false

      firefox_instances:
        description: Amount of Firefox Instances to be used. Won't use Firefox when its value is a blank string
        default: ''
        type: string
        required: false

      browser_visible:
        description: Specifies whether the browser is visible on the Remote Dashboard. If false, the browsers will run in headless mode.
        type: string
        default: 'true'
        required: false

      enable_video:
        description: Specifies whether browser is recorded or not. Will be disabled if the browsers run in headless mode.
        type: string
        default: 'true'
        required: false

      update_report:
        description: Specifies whether the report is updated or not.
        type: string
        default: 'false'
        required: false

Describe the issue

installing these as packages works around the issue but shouldn't paths to the browser driver successfully override the current project installed version ?

Using a project .npmrc file with chromedriver_filepath=/path/to/ChromeDriver.exe

Is that expected to override the current chrome driver in node modules ?
I have created a sh that appends it in path and added it to project .npmrc with no luck preventing it from picking up chromedriver version .114

per the chromedriver README.md node_modules\chromedriver\README.md
executing a npm install chromedriver got me to a different issue that I worked around by doing a new yarn for the installation

Links

node_modules\chromedriver\README.md