Context: I am developing a Google Cloud repository integration for CrateDB.
The workings and intended purpose of BlobContainer.children are quite unclear to me. Multiple BlobContainers with a different base path could be created in the same bucket next to each other, why should these be nested?
The Azure and S3 implementation of BlobContainer.children does not seem to respect the description: “A child container is defined as a container whose path() method returns a path that has this containers path() return as its prefix and has one more path element than the current,” and instead looks much more like an alternative to listBlobs (The S3 implementation contains a horrible infinite while loop, but that’s another topic).
Is the children method supposed to return a list of ‘sub-directories’? (of course in a blob storage this is not a real thing). If it is not crucial to making snapshots work, I am tempted to leave it unimplemented.