Closed juergenhoetzel closed 6 years ago
Oh, very nice. Thank you. Give me a day or two to look into and play around with this. 👍
Also looks like the link is wrong if the buffer being visited via tramp is opened via a symlinked path.
Also looks like the link is wrong if the buffer being visited via tramp is opened via a symlinked path.
Good catch: Also handled in the latest commit.
Hi, thanks.
Calling git-link
on tramp dired buffer produces wrong link. Maybe just need to call (file-remote-p file 'localname)
here?
Also can you squash this into a single or logical commits?
Hi, thanks.
Calling git-link on tramp dired buffer produces wrong link. Maybe just need to call (file-remote-p file 'localname) here?
I came up with a rather simple solution. There is no need to change git-link--relative-filename
: It's already Tramp-agnostic.
I just adjusted git-link--repo-root
to return tramp file names on remote hosts. In the latest (squashed) commit.
As already stated in #49,
process-lines
is not tramp-aware.Also a Tramp
(buffer-file-name)
has to be dissected.