Closed hadley closed 2 years ago
Yes, but this is due to the code in tools, I am not sure we should change this, in general I am trying to make sure our code stays as close to CRANs as we realistically can so we can reproduce their results.
It is objectively wrong though, because following a redirected link will preserve the fragment. Maybe another patch to R-core?
yeah I could look into it, will keep this open.
The standard says that the fragment should be preserved if the new url does not have a fragment: https://www.rfc-editor.org/rfc/rfc7231#section-7.1.2
Submitted patch to base R: https://bugs.r-project.org/show_bug.cgi?id=18316
I'll add this to urlchecker, nevertheless. If base is not patched then we'll need to update urltools.R
from R's SVN more carefully in the future.
This is now fixed in (tomorrow's) R-devel: https://github.com/wch/r-source/commit/2dc39e30df390b48218f9b062f829bbc64e347b2
The suggested replacement for
https://www.tidyverse.org/lifecycle/#maturing
ishttps://lifecycle.r-lib.org/articles/stages.html
, which loses the url fragment