11-18-perf_do_not_canocalize_the_entry_path - Branch - oxc-project/oxc-resolver

perf: do not canonicalize the entry path

#848Merged

Comparing

11-18-perf_do_not_canocalize_the_entry_path

(

ea425f6

) with

main

(

8e826e7

)

+8%

Improvements: 3

Untouched: 8

Skipped: 5

No successful run was found on main ( ea425f6) during the generation of this report, so 8e826e7 was used instead as the comparison base. There might be some changes unrelated to this pull request in this report.

Benchmarks

Skipped (5)

Improved

resolver_real[multi-thread]

benches/resolver.rs::resolver::bench_resolver_real

+8%

508.2 µs470.3 µs

resolver_real[resolve from symlinks]

benches/resolver.rs::resolver::bench_resolver_real

+4%

60.7 ms58.5 ms

resolver_memory[resolve from symlinks]

benches/resolver.rs::resolver::bench_resolver_memory

+3%

60.5 ms58.5 ms

Passed

resolver_real[single-thread]

benches/resolver.rs::resolver::bench_resolver_real

+2%

473.3 µs464.5 µs

resolver_memory[single-thread]

benches/resolver.rs::resolver::bench_resolver_memory

+1%

475.1 µs469.7 µs

resolver_memory[multi-thread]

benches/resolver.rs::resolver::bench_resolver_memory

+1%

480.9 µs477.7 µs

resolver_memory[drop]

benches/resolver.rs::resolver::bench_resolver_memory

84.9 ms84.8 ms

large

benches/resolver.rs::resolver::bench_package_json_deserialization::package_json_deserialization

-1%

26.4 µs26.5 µs

complex_real

benches/resolver.rs::resolver::bench_package_json_deserialization::package_json_deserialization

-1%

22.2 µs22.4 µs

medium

benches/resolver.rs::resolver::bench_package_json_deserialization::package_json_deserialization

-1%

17.1 µs17.3 µs

small

benches/resolver.rs::resolver::bench_package_json_deserialization::package_json_deserialization

-2%

11.6 µs11.9 µs

Commits

Click on a commit to change the comparison range

Base

main

8e826e7

+8.08%

perf: do not canonicalize the entry path (#848) This was introduced in `https://github.com/unrs/unrs-resolver/pull/125` The PR description and test case do not explain the reasoning behind this change, and I don't see how it reflects a real-world scenario in modern package managers.

ea425f6

7 days ago

by Boshen

Home Terms Privacy Docs