dandi.download

Functions

download(urls, output_dir, *[, format, ...])
pairing(p, gen)

Classes

DLState(value)	An enumeration.
DownloadDirectory(filepath, digests)
DownloadExisting(value)	An enumeration.
DownloadFormat(value)	An enumeration.
DownloadProgress(state, downloaded, size)
Downloader(url, output_dir, existing, ...[, ...])	meta private:
ItemsSummary()	A helper "structure" to accumulate information about assets to be downloaded
PYOUTHelper()	Helper for PYOUT styling
PathType(value)	An enumeration.
ProgressCombiner(zarr_size, file_qty, files, ...)

class dandi.download.DLState(value)

An enumeration.

CHECKSUM_ERROR = 5

DONE = 6

DOWNLOADING = 2

ERROR = 4

SKIPPED = 3

STARTING = 1

class dandi.download.DownloadDirectory(filepath: str | Path, digests: dict[str, str])

append(blob: bytes) → None

digests

Expected hashes of the downloaded data, as a mapping from algorithm names to digests

dirpath

The working directory in which downloaded data will be temporarily stored

filepath

The path to which to save the file after downloading

fp: IO[bytes] | None

An open filehandle to writefile

lock: InterProcessLock | None

A fasteners.InterProcessLock on dirpath

offset: int | None

How much of the data has been downloaded so far

writefile

The file in dirpath to which data will be written as it is received

class dandi.download.DownloadExisting(value)

An enumeration.

ERROR = 'error'

OVERWRITE = 'overwrite'

OVERWRITE_DIFFERENT = 'overwrite-different'

REFRESH = 'refresh'

SKIP = 'skip'

class dandi.download.DownloadFormat(value)

An enumeration.

DEBUG = 'debug'

PYOUT = 'pyout'

class dandi.download.DownloadProgress(state: 'DLState' = <DLState.STARTING: 1>, downloaded: 'int' = 0, size: 'int | None' = None)

downloaded: int = 0

size: int | None = None

state: DLState = 1

class dandi.download.ItemsSummary

A helper “structure” to accumulate information about assets to be downloaded

To be used as a callback to IteratorWithAggregation

as_dict() → dict

class dandi.download.PYOUTHelper

Helper for PYOUT styling

Provides aggregation callbacks for PyOUT and also an iterator to be wrapped around iterating over assets, so it would get “totals” as soon as they are available.

agg_done(done_sizes: Iterator[int]) → list[str]

Formatter for “DONE” column

agg_files(*ignored: Any) → str

agg_size(sizes: Iterable[int]) → str | list[str]

Formatter for “size” column where it would show

how much is “active” (or done) +how much yet to be “shown”.

class dandi.download.PathType(value)

An enumeration.

EXACT = 'exact'

GLOB = 'glob'

class dandi.download.ProgressCombiner(zarr_size: 'int', file_qty: 'int', files: 'dict[str, DownloadProgress]' = <factory>, maxsize: 'int' = 0, prev_status: 'str' = '', yielded_size: 'bool' = False)

feed(path: str, status: dict) → Iterator[dict]

file_qty: int

files: dict[str, DownloadProgress]

get_done() → dict

maxsize: int = 0

Total size of all files that were not skipped and did not error out during download

property message: str

prev_status: str = ''

set_status(statusdict: dict) → None

yielded_size: bool = False

zarr_size: int

dandi.download.download(urls: str | Sequence[str], output_dir: str | Path, *, format: DownloadFormat = DownloadFormat.PYOUT, existing: DownloadExisting = DownloadExisting.ERROR, jobs: int = 1, jobs_per_zarr: int | None = None, get_metadata: bool = True, get_assets: bool = True, sync: bool = False, path_type: PathType = PathType.EXACT) → None

dandi.download.pairing(p: str, gen: Iterator[dict]) → Iterator[tuple[str, dict]]