add 'postinstall' to enable hooks after install

[kennytram]

Description

Enable Git hooks for Husky immediately after yarn install husky's doc

For current developers, please run yarn husky install to ensure that Husky functions correctly.

Type of change

• New feature (non-breaking change which adds functionality)

How Has This Been Tested?

• After yarn husky install
  -- Intentionally made lint mistakes in some files after yarn husky install
  -- Detected by Husky after push

• Before yarn husky install
  -- Intentionally made lint mistakes before yarn husky install
  -- Not detected by Husky after push

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing tests pass locally with my changes
• Any dependent changes have been merged and published in downstream modules

[cherylli]

not working for me

[cherylli]

So I fixed it now and working, it's a windows issue (for me at least)

I followed this
typicode/husky#850 (comment)

[smurph7894]

Just fyi, the husky script does not work with Github Desktop but I am able to do it using gitbash.

[smurph7894]

see FYI comment but otherwise is working for me.

[cherylli]

So current husky setup is not quite working for me, it goes like this,
I push, husky runs automatically, but the code BEFORE formatting was pushed (linting seems to be ignored - I don't see any errors when it had errors when I manually run yarn lint)
so I had to do another commit for the formatted code

So the correct workflow should be

1. issue push command
2. husky runs
3. push FORMATTED code

is this happening to anyone else? I haven't looked it up yet but I was wondering if it was just me or not

[kennytram]

So current husky setup is not quite working for me, it goes like this, I push, husky runs automatically, but the code BEFORE linting was pushed so I had to do another commit for the formatted code

So the correct workflow should be

1. issue push command
2. husky runs
3. push FORMATTED code

is this happening to anyone else? I haven't looked it up yet but I was wondering if it was just me or not

Pre-commit - Detect and format files after commit
Pre-push - Detect and format files after push

Although, it does seem strange why Husky is only doing changes after a push when it's supposed to be "pre-push". I'll look it up this weekend if no one else does.

edit: "When you run the git push command, Git triggers any pre-push hooks that are installed. These hooks are intended to inspect the changes that are about to be pushed and possibly prevent the push if something isn't right. However, at this stage, the changes have already been committed to your local repository and are ready to be pushed. If your pre-push hook makes additional changes to the code (like reformatting it), these changes will not be included in the push because they haven't been committed yet"

It seems like we would need to enter additional commands for it to work as you intended.
We can do something like this:

yarn format
yarn lint
git add . {note: if we want to add staged files only, we will need to install lint-staged}
git commit -m "formatted code"

This will make two separate commits in one single push command.

What do you think?

[cherylli]

What do you think?

that's basically what I have to do now, but manually, but we shouldn't have to make an extra commit to include husky changes

if you look at this as an example, I basically have to do 2 commits every time I push, and one just for the husky changes
dev...feature/sprints-endpoints

Is there any reason we can't use a pre commit hook? I've used pre commit hooks for format in all my other projects and no issues

[kennytram]

What do you think?

that's basically what I have to do now, but manually, but we shouldn't have to make an extra commit to include husky changes

if you look at this as an example, I basically have to do 2 commits every time I push, and one just for the husky changes

dev...feature/sprints-endpoints

Is there any reason we can't use a pre commit hook? I've used pre commit hooks for format in all my other projects and no issues

Dan doesn't like it 😄

"Pre-commit would get too annoying" in his exact words.

[cherylli]

What do you think?

that's basically what I have to do now, but manually, but we shouldn't have to make an extra commit to include husky changes
if you look at this as an example, I basically have to do 2 commits every time I push, and one just for the husky changes
dev...feature/sprints-endpoints
Is there any reason we can't use a pre commit hook? I've used pre commit hooks for format in all my other projects and no issues

Dan doesn't like it 😄

"Pre-commit would get too annoying" in his exact words.

he's not going to commit or push this repo much. double commits annoys me and clogging the commit history 😂

What does everyone else in the team think?

[kennytram]

edit: Nvm I realized my solution isn't the correct approach for pre-push. I reverted my recent changes.

[jenny-alexander]

I haven't tried any of this yet - but agree that it should be a pre-commit hook and we shouldn't push two commits for this.
If we want to keep our git commit history clean"er", it's possible to do a git rebase.

[Dan-Y-Ko]

What do you think?

that's basically what I have to do now, but manually, but we shouldn't have to make an extra commit to include husky changes

if you look at this as an example, I basically have to do 2 commits every time I push, and one just for the husky changes dev...feature/sprints-endpoints

Is there any reason we can't use a pre commit hook? I've used pre commit hooks for format in all my other projects and no issues

If you use a pre commit hook, literally every time you try to commit, you will have to fix any linting errors. This honestly is a just a huge hassle. For me personally, I always do multiple commits and then when I'm done, I'll push it up to the repo and before I push, I do a lint/format. I like having the check in a pre-push in case I or someone else forgets to do that. Having to do 1 extra commit is way better tradeoff for me than having to fix stuff every time I commit (I'm not really sure what you mean about double commit, it should only be 1 extra). That being said that part is not relevant to the husky workflow issue and can be tweaked for each team. When I tried to set it up on the fe, it didn't matter what I used, it wasn't working for other people so yea.

[cherylli]

works for me, thanks for fixing this

[jenny-alexander]

It works for me after I set the pre-commit script to executable. Approved!