We could do that. But having the whole directory readily available for manual interaction / debugging is also a plus.

Additionally, further tests might be added with non-empty files.

I'm thinking of two additional ways of handling it:

• Keep everything in master; use .ignore and counsel-rg for selecting the file list.
• Have the master git branch be the dependency of test instead of the other way around.

The current approach also helps us avoid including large files into the master branch. Suppose in the future we add a performance test to swiper with a 1Mb file. Then it's better to not pollute the master branch with it.