Improve performance of spliceChangesIntoString

[ABuffSeagull]

The original algorithm caused a lot of memory pressure, especially in large files.
For a 91 KB file at my job, this change speeds it up from 101509ms to 1095ms.

[thecrypticace]

I opened a PR with these changes, some additional optimizations, tests, and benchmarks: #312

I would've edited this PR but GH wasn't allowing me to for some reason (it says I can edit but it wasn't letting me 🤷 )

Thank you for figuring this one out. Closing in favor of that PR (you're still credited btw!)

[shyakadavis]

I would've edited this PR but GH wasn't allowing me to for some reason (it says I can edit but it wasn't letting me 🤷 )

Hi,

I don't mean to spam, but I've seen some comments from huntabyte, saying that the issue here is caused by the fact that one opens a P.R from the main branch of a fork, targetting main on, well, the main repo.

A simple solution was always to open the P.R from a different branch, then maintainers can commit directly to it.

Cheers.

P.S. I found myself here after digging through the new release(s) (v0.6.8 & v0.6.7) which were breaking after renovate bot tried updating. Glad it's working now. 🙂

Cc: @thecrypticace @ABuffSeagull