Enforcing a maximum line length of 120

[john-science]

What is the change?

Using ruff to enforce a maximum line length of 120. Because ruff is doing the enforcing, GitHub CI will catch and enforce this new rule in the future.

This was mostly an automated process, though I did have to touch the pyproject.toml by hand, and I touched the release notes.

Why is the change being made?

This is purely for code quality and readability; lines that are longer than 120 characters are pretty hard to read on a git diff or even the "GitHub Diff" UI in PRs. They also add to a lot of clutter and confusion in most IDEs.

We could also move the bar to 100 characters, which would be more standard. But... ruff tells me that would make changes to 768 files, and that feels like too big a pill to swallow.

---

Checklist

• This PR has only one purpose or idea.
• Tests have been added/updated to verify that the new/changed code works.

• The code style follows good practices.
• The commit message(s) follow good practices.

• The release notes (location doc/release/0.X.rst) are up-to-date with any important changes.
• The documentation is still up-to-date in the doc folder.
• If any requirements were affected, mention it in the release notes.
• The dependencies are still up-to-date in pyproject.toml.

[albeanth]

I reviewed the first half of the files changed. @opotowsky is doing the second half.

[opotowsky]

There are a ton of updates in here that increase line length. Can you undo all those changes? That creates more unnecessary work for us (it gets reviewed in this PR, then again in the 100 char line length limit PR).

[albeanth]

There are a ton of updates in here that increase line length. Can you undo all those changes? That creates more unnecessary work for us (it gets reviewed in this PR, then again in the 100 char line length limit PR).

This is true if the plan is to move the line length bar again. Which I don't really understand a motivation for. I can get trimming it to 120, but 100? Or 88? It seems unnecessary?

[john-science]

This is true if the plan is to move the line length bar again. Which I don't really understand a motivation for. I can get trimming it to 120, but 100? Or 88? It seems unnecessary?

Well, two weeks ago, we had lines that were >200 characters long. That is a bit too much, even for me.

I tend to work at 120. Though... 88 is by a WIDE margin the most common. For instance, that is the default in both black and ruff.

To be clear, I wasn't really interested in lowering the bar again. But I am by FAR in the minority here. I just happen to find that scientific code has long lines of math that mean stringent line-length limits make code harder to read and parse. Perhaps the "88 character limit" comes from website development and other work of that sort, where no individual line has a LOT of math to it.

[opotowsky]

I reviewed the bottom 26 files.

I originally commented that lines already adjusted to a 100 char limit shouldn't be increased to 120. But we discussed that and since we aren't probably ever going to enforce a 100 char line length, I was OK leaving those edits as-is.

I now noticed a bunch of places where a 120 char line was edited down to a 100 char line. I commented on it in one file, but once I saw how pervasive the inconsistency was I stopped commenting. It's not the biggest deal....after all, these lines technically follow the <120 rule. But how this got inconsistently applied is baffling to me. Is it ruff misbehaving? Your error?

[opotowsky]

same as below....seems like some line wraps get 100 chars applied and some get 120

[opotowsky]

same comment as below

[albeanth]

The first 26 files are good.

[albeanth]

I still approve the first 26 files. @opotowsky to approve the second half.

[opotowsky]

I can approve this once my review is responded to.

Looks like I have an outstanding british v american english spelling thing, and question for my review. John partially answered my questions about the 100 vs 120 char issue, but it seems the inconsistency is broader than the question that was answered (that related to impl tags).