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
For the new workspace API we want a way to have a process that explicitly allows to store a result in a dedicated workspace instead of in "the normal user space"
We thought about having it as part of save_result, but then you either mix the storage location with the file formats or you get a number of parameters that depend on each other in save_result, both not ideal.
Backends may not do the intermediate step to store in the normal user space and directly store in the workspace although both processes are present.
The process usually only makes sense in batch jobs, I think.
NOTE: This process may have implications on save_result!
Summary
Exports a result to a registered workspace
Description
...
Parameters
data (or so?)
One issue is that save_result returns a boolean, so we need to look into that. Maybe it should return something else that we can feed into export_to_workspace as data? Maybe a list of files? Or just a generic "result" subtype?
name
Name of the registered workspace.
Type: string
Subtype: workspace-name
Optional: no
folder (or prefix?)
Stores the batch job with the given prefix / to the given folder in the workspace.
Default to null, which uses a unique identifier as prefix (e.g. batch job id).
An empty string stores them in the "root folder".
Q: Should the process by default override files or not?
Type: string or null
Optional: yes, default: null
Return Value
boolean?
Categories (optional)
export
Links to additional resources (optional)
Link to workspace API
Details about exceptions mentioned in the description (optional)
FileExists
The text was updated successfully, but these errors were encountered:
Should we have most of the discussion in this issue: Open-EO/openeo-api#376
Because there we already collected a lot of thoughts and requirements. I'm also not really clear yet on what the new workspace API looks like.
Proposed Process ID: export_to_workspace
Context
For the new workspace API we want a way to have a process that explicitly allows to store a result in a dedicated workspace instead of in "the normal user space"
We thought about having it as part of save_result, but then you either mix the storage location with the file formats or you get a number of parameters that depend on each other in save_result, both not ideal.
Backends may not do the intermediate step to store in the normal user space and directly store in the workspace although both processes are present.
The process usually only makes sense in batch jobs, I think.
NOTE: This process may have implications on save_result!
Summary
Exports a result to a registered workspace
Description
...
Parameters
data
(or so?)One issue is that save_result returns a boolean, so we need to look into that. Maybe it should return something else that we can feed into export_to_workspace as data? Maybe a list of files? Or just a generic "result" subtype?
name
Name of the registered workspace.
Type: string
Subtype: workspace-name
Optional: no
folder
(or prefix?)Stores the batch job with the given prefix / to the given folder in the workspace.
Default to
null
, which uses a unique identifier as prefix (e.g. batch job id).An empty string stores them in the "root folder".
Q: Should the process by default override files or not?
Type: string or null
Optional: yes, default: null
Return Value
boolean?
Categories (optional)
Links to additional resources (optional)
Details about exceptions mentioned in the description (optional)
The text was updated successfully, but these errors were encountered: