Arvados

Idea #20852

Updated by Brett Smith about 1 year ago

A lot of code in the "Python SDK cookbook":https://doc.arvados.org/v2.6/sdk/python/cookbook.html that coordinates low-level API calls should just be methods that are defined in the SDK itself. Why force every user to copy-paste these code snippets when we can just give them the method directly? 

 For recipes that I think are strong candidates for inclusion, here's are some rough sketches of what the method signatures could look like, along with implementation notes. 

 <pre><code class="python"> 
 class Collection: 
     def download(self, src: PathLike, dst: PathLike | None=None) -> None: ... 
     # If src is a PurePath, generate a manifest path with .as_posix(). 
     # If dst is None, default to the src basename. 

     def upload(self, src: PathLike, dst: PathLike | None=None) -> None: ... 
     # Same argument handling as download, except now we use dst.as_posix(). 

     def iter_files(self, depth: int | None=None) -> Iterator[ArvadosFile]: ... 
     # depth limits recursion, None means no limit. 

     def iter_streams(self, depth: int | None=None) -> Iterator[Collection]: ... 

     def iter_contents(self, depth: int | None=None) -> Iterator[ArvadosFile | Collection]: ... 

     def copy( 
             self, 
             source: PathLike | IO, 
             target_path: PathLike | None=None, 
             source_collection: Collection | None=None, 
             overwrite: bool=False, 
     ) -> None: ... 
     # Proposed expansion of the existing copy method signature. 
     # Explicitly passing an empty target_path should not be required, that can 
     # be the default if not specified. 
     # If source is an IO, just copy that object directly. target_path defaults 
     # to `source.name`. If source_collection is not None, that's a ValueError. 
     # If source is an ArvadosFile, can do manifest manipulation directly. 
     # If source is PathLike, follows the current implementation. 
     # With these changes, the current "copy file" recipe could be simplified to: 
     with src_coll.open('ExamplePath') as src_file: 
         dst_coll.copy(src_file) 

     def sharing_link(self, expires_at: datetime.datetime | datetime.timedelta ) sharing_link(self) -> str: ... 
     # expires_at is used for the created token. If it's a timedelta, 
     # that's relative to datetime.datetime.now(). 

 


 # I'm not sure what module these go in. 
 def container_started( 
         client: ArvadosAPIClient, 
         container: str | Container | ContainerRequest, 
 ) -> bool: ... 
 # If container is a str, it's a UUID for one of the other types, fetch that. 
 # If it's a container request, return False if container_uuid is None, else fetch that. 
 # `return status not in {'Queued', 'Locked'}` 

 def container_finished( 
         client: ArvadosAPIClient, 
         container: str | Container | ContainerRequest, 
 ) -> bool: ... 
 # Same dance as above, `return status in {'Cancelled', 'Complete'}` 

 def container_succeeded( 
         client: ArvadosAPIClient, 
         container: str | Container | ContainerRequest, 
         success: typing.Container[int]=frozenset([0]), 
 ) -> bool: ... 
 # Same dance as above, `return status == 'Complete' and exit_code in success` 

 def child_requests( 
         client: ArvadosAPIClient, 
         container: str | Container | ContainerRequest, 
         filters: list[list[str]]=[], 
         select: list[str]=[], 
 ) -> Iterator[ContainerRequest]: ... 
 # Same dance as above, return the obvious `keyset_list_all` call. 

 def child_containers( 
         client: ArvadosAPIClient, 
         container: str | Container | ContainerRequest, 
         request_filters: list[list[str]]=[], 
         container_filters: list[list[str]]=[], 
         select: list[str]=[], 
 ) -> Iterator[Container]: ... 
 # Same dance as above, call `child_requests` to get container_uuids, then 
 # return the obvious `keyset_list_all` call with those UUIDs. 
 # Note that `child_requests` can be called with a very narrow `select` for 
 # optimization, and can add the filter `container_uuid is not null`. 
 </code></pre>

Back