WIP: V1 artifact files endpoint changes #658
Conversation
| @@ -46,6 +47,7 @@ | |||
| from .service import ( | |||
| RESOURCE_TYPE, | |||
| SEARCHABLE_FIELDS, | |||
| ArtifactIdContentsService, | |||
There was a problem hiding this comment.
causes an import error right now. I realize you haven't worked on the service layer yet, but I found it useful to create a class stub, so flask would run and I could view the endpoint and schema in swagger.
| attribute="relative_path", | ||
| metadata=dict(description="Relative path to the Artifact File."), | ||
| ) | ||
| artifactType = fields.String( |
There was a problem hiding this comment.
will need to coordinate artifact types with @hbooth 's work. Probably should be an enum see here for an example:
https://github.com/usnistgov/dioptra/blob/a661aaa168788dffb1d7dcb1c95952b4106e9b7e/src/dioptra/restapi/v1/schemas.py#L211C1-L229C1
| load_default=None, | ||
| ) | ||
|
|
||
|
|
||
| class ArtifactGetQueryParameters( | ||
| PagingQueryParametersSchema, | ||
| GroupIdQueryParametersSchema, | ||
| SearchQueryParametersSchema, |
There was a problem hiding this comment.
I think we should leave out Search and SortBy query parameters at least for now. These would require a custom implementation since search and sort for resources are implemented via the sqlalchemy.
| required=True, | ||
| ) | ||
| files = fields.Nested( | ||
| ArtifactFileSchema, |
There was a problem hiding this comment.
should be the paged schema.
| required=True, | ||
| ) | ||
| fileTree = fields.Nested( | ||
| ArtifactFileSchema, |
There was a problem hiding this comment.
need a schema for the file tree view. I'm not sure if we defined this yet. I'm also not sure how it would work with the paging parameters.
The intent of the file tree was to provide the frontend with a nested file tree structure that makes it easy to display in a file browser like UI. If we are concerned with the amount of data that might be returned and need paging, maybe we can instead remove fileTree and have the UI use files, but add a recurse boolean query param that determines if we want a flat list of all files, or if we just want the files in the directory in the path query param. This way the UI would work by making a call to /artifacts/{id}/contents?path=PATH&recurse=False, updating the path each time the user clicks a file/directory in the browser.
| ) | ||
|
|
||
|
|
||
| class DownloadPathQueryParametersSchema(Schema): |
There was a problem hiding this comment.
Consider renaming to something like ArtifactContentsQueryParameterSchema (especially if we add additional fields like recurse) since we are grouping a couple different parameters all related to the contents or files associated with the artifact.
|
|
||
|
|
||
| class DownloadPathQueryParametersSchema(Schema): | ||
| """A schema for adding path and downlaod query parameters to a resource endpoint.""" |
There was a problem hiding this comment.
minor typo: downlaod -> download
mtrapnell-nist
force-pushed
the
v1-artifact-files-schema
branch
from
0e6e374 to
a5749ef
Compare
Add initial schema and controller updates for handling browsing and retrieving artifact files.
This commit adds a service for retrieving Artifact file contents or a directory listing of the files. It also changes the Artifact service to allow for an Artifact to be uploaded to mlflow directly on creation. These features are not fully implemented.
Artifact file contents can be requested from mlflow using an artifact_id. If the artifact contains multiple files, it can either be sent as an archive or a file directory listing, depending on the download parameter.
…or ArtifactID endpoint
mtrapnell-nist
force-pushed
the
v1-artifact-files-schema
branch
from
9b1bdc2 to
674903c
Compare
No description provided.