Why not both? about semantic-pull-requests

[CecileRobertMichon]

+1 on this.

Our use case is the following: GitHub doesn't let you choose a default squash commit message. Squashing a pull request with one commit would use that commit’s message, but squashing a pull request with more than one commit will use the pull request’s title for the commit message. We want to make sure that every single external contributor PR in our repo gets merged with a valid squash commit message so checking both would be beneficial for us.

Ideally we would be able to check the squash commit message only (so PR title for multiple commits, commit for single commit).

[zeke]

Got it. There's an open issue about squashing single-commit PRs: #17

I am pretty busy right now, but would happily accept and ship a PR that addresses this.

[nikolasleblanc]

#46 @clakech I think you meant this PR of yours :)

[zeke]

Resolved by #46 -- this should be live within a few minutes.

[zeke]

👋 Thanks for reporting. You're right about the current behavior. Here's what we have now:

default behavior: validate title or at least one commit is semantic
titleOnly: true: Always validate the PR title, and ignore the commits
commitsOnly: true: Always validate all commits, and ignore the PR title

For what reason would you want to validate both the title and all commits? I'm not against it, but wondering about your use case.

[zeke]

@CecileRobertMichon would only allowing "squash merging" on the repo work for your case?

[CecileRobertMichon]

@zeke not exactly. We already only allow squash merging. The issue is that I can't control how many commits are in a pull request and GitHub is inconsistent when squash merging commits. If there is a single commit, the commit message comes from the commit. If there are multiple commits, the commit message comes from the PR title. So checking only the commits or only the title is not sufficient since either may be used at squash time.

[clakech]

We would like to have both to have consistency between PR titles and all commits

Here is a first implementation / proposal #42