It's sort of subjective whether, once .gitignore files become long, they should be grouped. The intention seemed roughly to group it, and I think this may be preferred to sorting it or doing neither.

This divides the entries into groups and labels each group with a comment.

I've also included some minor changes, which can be omitted if desired (I can amend and force push or commit again):

• Remove /*.egg-info because /*egg-info covers it.
• Add __pycache__/ even though it should probably only contain *.pyc and *.pyo.
• Add monkeytype.sqlite3.* for compressed copies of that database. In #1725 I was tempted to remove test/tstrunner.py, but I'm glad I didn't, because it looks like MonkeyType may still be useful for next time work is done on improving type annotations. Actually, what seems most useful to me about it is that it detects cases where imports rewrite what modules are accessible where, causing a situation where from x.y import z imports z from a different y than import x.y or from x import y imports. Anyway, the sqlite database it outputs is about 2.2 GiB, but compresses down to about 27 MiB (when xz is used). So I think it makes sense to ignore such files, too, to make working with them easier.