canimus

Had a similar challenge when writing alphareader is in GitHub. HN comments helped me to think in multi-byte separators, and one thing is sure no matter which char you choose it will appear in the wrong place at some point.

Thanks for the post. Appreciated.

Thank you @eesmith. Comments appreciated, and PRs to the repo as well. ;-) The multi-byte is great catch! I made the wrong assumption, on single byte separators. Perhaps a library limitation if the we want to keep the logic simple. Ideas on the fix?