Testing #23

[OnkarRuikar]

• Testing Revert "Test custom reviewdog" #23

Here we've tested:

# Because reviewdog submits suggestion if the
# OWNER ID of the PR's HEAD and the repository are the same, or
# the branch is created from the same repository.
# As a workaround, replace those OWNER IDs with `1`.
EVENT=$(jq '.pull_request.head.repo.owner.id = 1 | .pull_request.base.repo.owner.id = 1' diff/event.json)
echo "${EVENT}" > diff/event.json

[github-actions]

[mdn-linter] _{reported by reviewdog 🐶}

Suggested change

	## 开发网页游戏
	## 开发网页游戏

[github-actions]

[mdn-linter] _{reported by reviewdog 🐶}

Suggested change

	div {background-color: red;}
	div {
	background-color: red;
	}

[github-actions]

[mdn-linter] _{reported by reviewdog 🐶}

Suggested change

	* [关于 Emscripten](https://emscripten.org/docs/introducing_emscripten/about_emscripten.html)——简介和高级特性。
	+ [下载和安装](https://emscripten.org/docs/getting_started/downloads.html)——安装工具链。
	- [关于 Emscripten](https://emscripten.org/docs/introducing_emscripten/about_emscripten.html)——简介和高级特性。
	- [下载和安装](https://emscripten.org/docs/getting_started/downloads.html)——安装工具链。

[github-actions]

[mdn-linter] _{reported by reviewdog 🐶}

Suggested change

	* abc
	+ efg
	- abc
	- efg

[OnkarRuikar]

@yin1999 🎉 This is a clean solution because we get to keep the original reviewdog.

Could you implement this solution in mdn/content repo?

• Update contents of pr-check_markdownlint.yml with pr-review-lint.yml, because we are doing only linting in the first part.
• Rename second part to pr-lint-review-companion.yml.

[yin1999]

@yin1999 🎉 This is a clean solution because we get to keep the original reviewdog.

Could you implement this solution in mdn/content repo?

• Update contents of pr-check_markdownlint.yml with pr-review-lint.yml, because we are doing only linting in the first part.
• Rename second part to pr-lint-review-companion.yml.

Of course I could. One more thing, should we recommend users to submit all suggested changes at once in PR review. I'm not sure will the reviewdog create duplicate suggestions?

[OnkarRuikar]

I'm not sure will the reviewdog create duplicate suggestions?

It doesn't seem to. I've later pushed a new commit containing only invalid list markers. Revewdog only posted a comment related to that.
Also later I merged only one suggestion and the workflow didn't upload any artifact to for the second part. Could you remove the front-matter linter part from the first workflow, it is throwing errors? Also, now instead of dumping env log the diff file for debugging cat ${TMP_DIR}/diff.

should we recommend users to submit all suggested changes at once

We can recommend them to do it in a single batch from the files tab as a good practice. Because committing one by one triggers multiple workflows every time, don't know if GitHub cancels prior running workflows before starting workflows for new commits.

We can make this suggestion in PR review companion's message explaining what the auto review is doing and how to accept these changes.

[yin1999]

Thanks for the try and suggestions @OnkarRuikar 👍

[OnkarRuikar]

@yin1999 could you hold off on the PR in mdn/content repo? Another approach has been suggested by Leo let me try that first.

[yin1999]

@yin1999 could you hold off on the PR in mdn/content repo? Another approach has been suggested by Leo let me try that first.

Yes, please go ahead!

[yin1999]

Closing as the new approach (mdn/content#28363) works great.