It would be really great if Lexper could provide even more accurate handling of canonical URL information. I may be misunderstanding this, but it seems there are the following points:
There are websites that have link rel=canonical specified as relative path and for them Lexper's canonical url is "incorrect", e.g.:
1) If the website always redirects to https and the link is passed with http protocol, Lexper's canonical url will be with http e.g. http://firstround.com/review/The-Arc-of-Company-Life-and-How-to-Prolong-It/
2) If the link redirects to some article on other domain, an "incorrect" link is returned, e.g. when https://bit.ly/2PqScH1 is parsed, the canonical url is https://bit.ly/review/The-Arc-of-Company-Life-and-How-to-Prolong-It/
There are websites that don't have link rel=canonical specified and Lexper's canonical url in this case points to original url, not to redirected url, e.g.
1) https://bit.ly/2DzViUc redirects to https://blog.posthaven.com/read-about-how-fly-has-helped-with-posthavens-infrastructure, however https://bit.ly/2DzViUc is returned as canonical url.
2) http->https redirect, http://sidekiq.org/products/enterprise.html is returned as canonical url, however it redirects to https://sidekiq.org/products/enterprise.html
Is this something that you could incorporate?
Thank you very much!