Improving Git Blame

When running sqlfmt for the first time, or when running an upgraded version of sqlfmt after a style change, you are likely to produce a large number of changes in your codebase.

Some teams find this problemmatic, because they regularly use git blame as part of their workflow, to see a history of changes and their linked commits.

Fortunately, git makes it possible to ignore specific commits when viewing the blame for a file. To use this feature:

1. Commit all other changes before running sqlfmt on your project.
2. Run sqlfmt . to format all files in your project.
3. Create a new commit with only the formatting changes, with git commit -am "chore: apply sqlfmt".
4. Note the SHA hash returned by the previous command (or use git show -q to view it), since you'll need it later.

Now that you have an isolated commit with only formatting changes, you can instruct git blame to ignore that commit. The simplest way to do this is by passing the SHA to the --ignore-revs option:

git blame --ignore-revs 6ea1f05cc48469c09ba539b6498bd9f484aa63d3

This is fine for a one-off use case, but it's brittle and hard to collaborate with teammates, who will also want to ignore this commit. A more robust option is to create a file named .git-blame-ignore-revs in the root of your project. That file should contain a list of commits (and optionally comments) for git blame to ignore:

.git-blame-ignore-revs

# chore: apply sqlfmt v0.18.0
6ea1f05cc48469c09ba539b6498bd9f484aa63d3

tip

You can name your file anything, but if you use .git-blame-ignore-revs and locate that file in the root of your project, GitHub will automatically use that file for its own Blame feature.

Then to view the blame, ignoring commits in the file:

git blame --ignore-revs-file .git-blame-ignore-revs

That's fine, but we can do better. There is a configuration option that allows us to set this filename once, rather than pass it as an option every time we call git blame:

git config blame.ignoreRevsFile .git-blame-ignore-revs
git blame