I'm benchmarking on two repos (see script/bench.sh):
| Target | Time |
|---|---|
| (elm-spa-example)[https://github.com/rtfeldman/elm-spa-example] (224kb of Elm code) | 30.9 ms ± 1.3 ms |
| my main work repo (12mb of Elm code) | 2.055 s ± 0.041 s |
I want this to be much faster.
The old tool that this is replacing for me does more and runs in 1.210 s ± 0.016 s on the work repo... and it's written in Python (not that Python is necessarily slow but it's probably possible to get a faster result!)
It would probably also be OK to drop tree-sitter.
It's not really super essential for this task, since the regex I actually want is tiny: ^import ([A-Z][\w\d\.]+)
We can probably do better by reducing allocations etc but there are lower-hanging fruit.
I added crossbeam and used a lot more of ignore, so maybe LTO has an effect now?
| Target | Old Time | New Time | Speedup |
|---|---|---|---|
| elm-spa-example | 30.9 ms ± 1.3 ms | 29.7 ms ± 1.0 ms | Something like 1ms |
| work repo | 2.055 s ± 0.041s | 2.129 s ± 0.022 s | actually slowed down by like 0.1s? |
Build times are worse, but in a manageable way: a fresh build of cargo build --release now takes 2m19s (up 30s from 1m49s)
In the end, this is still not worth it, and I'm backing the change out.
ignore has a parallel walker implementation.
Here's what happens when we switch to walking in parallel:
| Target | Old Time | New Time | Speedup |
|---|---|---|---|
| elm-spa-example | 44.9ms ± 2ms | 30.9 ms ± 1.3 ms | ~14ms, or ~69% of the time |
| work repo | 3.920s ± 0.252s | 2.055 s ± 0.041 s | ~1.87s, or ~52% of the time |
Enabling link-time optimization with this in Cargo.toml:
[profile.release]
lto = true
... makes compile times go to like 2 minutes on my machine, with only a tiny speedup. "thin" LTO does not help either (although it's a slightly faster compile.)