You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to Biocaml
Hi everyone,
I've just pushed a small benchmark on line splitting, comparing core, sosa and container. Results (shown below) depend on the actual compilation target (bytecode or native), but show that core's implementation performs best. sosa has some room for improvement I think [0]. Specializing core's implementation for looking for newline character yields a small something, which makes it worth being added in `Line` IMO.
I did that in preparation of moving some modules to a `base` library, that would be contain all "pure" code as we discussed last month. Let me know if you want to participate.
Cheers,
ph.
P.S. output is generated by core_bench and provide execution time, words allocated in minor heap, in major heap, number of promotions and fraction of the slowest execution time.