Cache absolute_path to decrease allocations
#346
Conversation
While profiling local file handling, I noticed that resolving paths was taking a significant amount of time. It also caused quite a few allocations. By caching the path and using a constant value for the current directory, we can reduce the number of allocs by quite a lot. For example, when testing on the sentry documentation, we do 50,4% less allocations in total now. That's just a single test-case of course, but it's probably also helping in many other cases as well.
Another ~1.5% less allocs
|
I'm running into the same issues as in idanarye/rust-typed-builder#57 again. My suspicion is that this is actually an issue in cargo-publish-all. When comparing the compile commands between cargo publish and cargo-publish-all, I see some differences: cargo publish --manifest-path lychee-lib/Cargo.toml --dry-run: cargo-publish-all --dry-run --verbose: Namely, cargo-publish-all uses @dblock fyi |
Did that just reappear and is it reproducible on main? |
|
Oh I might not have expressed myself very well. Just to be clear, the old bug with rust-typed-builder is fixed and that fix still works. But after adding the |
I see. You should (re)open an issue in cargo-publish-all, cc: @idanarye |
|
Good point. I've added a comment to https://gitlab.com/torkleyy/cargo-publish-all/-/issues/3. |
|
There was no update on the issue yet. I've decided to disable the
|
* Cache `absolute_path` to decrease allocations While profiling local file handling, I noticed that resolving paths was taking a significant amount of time. It also caused quite a few allocations. By caching the path and using a constant value for the current directory, we can reduce the number of allocs by quite a lot. For example, when testing on the sentry documentation, we do 50,4% less allocations in total now. That's just a single test-case of course, but it's probably also helping in many other cases as well. * Defer to_string for attr.value to reduce allocs * Use Tendrils instead of Strings for parsing (another ~1.5% less allocs) * Move option parsing code into separate module * Handle base dir more correctly * Temporarily disable dry run
While profiling local file handling, I noticed that resolving paths was taking a
significant amount of time. It also caused quite a few allocations.
By caching the path and using a constant value for the current
directory, we can reduce the number of allocs by quite a lot (5%).
This is my second attempt at reducing allocs. The first one is here #345.