You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm working on building a dataset from .tgz archives using the replacement for add-archive-content demonstrated here in combination with the archivist special remote. The demo below works if the archive is a .tar.gz extension but not with .tgz. With .tgz, I need to configure the archivist.legacy-mode for a successful datalad get. Here's a quick demo:
% datalad configuration --scope localset datalad.archivist.legacy-mode=yes 1 !
set_configuration(ok): . [datalad.archivist.legacy-mode=yes]
% datalad get project/file1.txt
[INFO ] datalad-archives special remote is using an extraction cache under /playground/loj/abcd/tmp3/.git/datalad/tmp/archives/8bc4249de3. Remove it with DataLad's 'clean' command to save disk space. get(ok): project/file1.txt (file) [from archivist...]
Thanks a lot for the excellent report that made it easy to spot the issue. There are two things that can be done here. The problem is indeed the .tgz extension not being used to detect the archive type.
Fix 1:
You can declare the archive type in the URL. The adjusted addurls call that does this is:
I'm working on building a dataset from
.tgz
archives using the replacement foradd-archive-content
demonstrated here in combination with the archivist special remote. The demo below works if the archive is a.tar.gz
extension but not with.tgz
. With.tgz
, I need to configure thearchivist.legacy-mode
for a successfuldatalad get
. Here's a quick demo:datalad wtf
The text was updated successfully, but these errors were encountered: