The benchmark measures and compares execution time and peak memory consumption of the current and new commonpath methods.
Measurement is performed using batches of 1000 paths. Each batch contains paths generated with the following variables:
- Length of each path part (
range(16, 65, 16)
) - Number of parts (
range(4, 17, 4)
) - Number of common parts (
range(parts_count + 1)
)
The batch is then split into chunks of equal size such that paths in each following chunk reduce the number of common parts so far by 1 up to the selected number [3], e.g. if the selected number of common parts is 1 then the batch would be "a/b/c", "a/b/c", "a/b/d", "a/b/d", "a/c/d", "a/c/d", "a/c/d".
Each batch is tested in two permutations: