Hi everyone, I was looking into evaluating the performance of some library I’m writing and I was able to write a dune test comparing it to other libraries fairly easily (with the benchmark package). However this requires to have all the pieces of code you want to test available, and in many cases I would like to compare several approaches for the same code which would most naturally be found in different commits or branches, and having them side by side would involve some tedious copying and renaming of things. Does anyone have a good workflow to share for how to do this? Ideally I would like to have an arbitrary collection of functions taken from different branches (not necessarily the same name in each branch) in one big benchmark.
Related topics
Topic | Replies | Views | Activity | |
---|---|---|---|---|
Custom tests template dune | 8 | 1732 | January 25, 2021 | |
Building Base and other Jane Street libraries from source | 4 | 1042 | May 19, 2020 | |
Dune shortcut for specifying a dependency between modules in 'src' and 'test' | 6 | 1310 | April 18, 2018 | |
Dune tests in subdirectories | 4 | 1373 | November 29, 2018 | |
Compile errors with `dune runtest` | 4 | 833 | August 24, 2018 |