Is there a preferred way to run a benchmark that takes one or more files as inputs? (The input files are going to be checked into the repository, but the benchmark binary would ideally take files to process as command line arguments.)
Right now I am using include_str! to embed the test files in the benchmarks, but this is a little inflexible for my taste and also means I can't measure the effect of different ways of reading files from disk.
1 post - 1 participant


