This issue is a follow-up of #23.
Once basic handling of Directory input parameters for EOAP-compliant stage-in is implemented, we can consider user-friendly automated processing options to present staged-in data to the notebook code.
-
The simplest implementation would be just to read each catalog.json into a pystac Catalog object and place it in the global namespace for the user to traverse. This should probably be implemented as well as any more sophisticated solutions, to ensure that everything is accessible even if xcengine's heuristics don't pick it up.
-
xcengine could traverse the STAC catalogue itself, pulling out likely-looking assets (probably by MIME type) and populating a dictionary for the user. This would be more convenient for the user than the rather boilerplatey process of digging through a STAC hierarchy with pystac.
See commits in cc69d74 and 2bdddff for some initial work in this direction.
This issue is a follow-up of #23.
Once basic handling of
Directoryinput parameters for EOAP-compliant stage-in is implemented, we can consider user-friendly automated processing options to present staged-in data to the notebook code.The simplest implementation would be just to read each
catalog.jsoninto a pystacCatalogobject and place it in the global namespace for the user to traverse. This should probably be implemented as well as any more sophisticated solutions, to ensure that everything is accessible even if xcengine's heuristics don't pick it up.xcengine could traverse the STAC catalogue itself, pulling out likely-looking assets (probably by MIME type) and populating a dictionary for the user. This would be more convenient for the user than the rather boilerplatey process of digging through a STAC hierarchy with pystac.
See commits in cc69d74 and 2bdddff for some initial work in this direction.