#37 Use directories/archives for INPUTFILES and OUTPUTFILES

[Anonymous]

Originally created by: msieb...@gmail.com

*To whom is this feature relevant? (remove the other line(s))*
-users

*Please provide a one line description of the feature:*
Implement ability to handle directories (and/or archive files) in
*::INPUTFILES::* and ::OUTPUTFILES::*

*How would the new feature work (appear in use)?*
As with copying files; just with directories.

*Please provide any additional information below.*
When specifying a directory for either input or output, the entire
directory should be copied over. Some care should perhaps be takes so the
user does not copy previous result-sets over.

Along the same lines; an idea could be to handle archive files
(tar/zip/gzip), as with URLs and the like. Ex. the following would
uncompress the zip-file in the given directory before execution:
::INPUTFILE::
project/data.zip project/data/

[Anonymous]

Originally posted by: jonas.ba...@gmail.com

the basic recursive directory is something we should implement when we find the time.
It would be useful although manual packing workarounds exist.

Implicit un-/packing would be nice, but it is less important. We would need a more
explicit way of specifying un-/packing as he proposed scheme collides with implicit
destination path.

The https transfer of files with curl and optional availability of packing software
on the resource both make it a bit harder to implement.

Status: Accepted

[Anonymous]

Originally posted by: jonas.ba...@gmail.com

support for wild cards in paths where it makes sense would also be nice