Post-download actions sometimes need to be taken on downloaded files
(unzipping, conversion to a more efficient format, etc). If these actions are
time or memory consuming, it might be worth doing them only once after the file
is downloaded. This is a way of trading disk space for computation time.
pooch.retrieve accept the
argument to handle these situations.
Processors are Python callable objects (like functions or classes with a
__call__ method) that are executed after a file is downloaded to perform
these actions. They must have the following format:
def myprocessor(fname, action, pooch): ''' Processes the downloaded file and returns a new file name. The function **must** take as arguments (in order): fname : str The full path of the file in the local data storage action : str Either: "download" (file doesn't exist and will be downloaded), "update" (file is outdated and will be downloaded), or "fetch" (file exists and is updated so no download is necessary). pooch : pooch.Pooch The instance of the Pooch class that is calling this function. The return value can be anything but is usually a full path to a file (or list of files). This is what will be returned by Pooch.fetch and pooch.retrieve in place of the original file path. ''' ... return full_path
The processor is executed after a file downloaded attempted (whether the download actually happens or not) and before returning the path to the downloaded file. The processor lets us intercept the returned path, perform actions, and possibly return a different path.
Pooch provides built-in processors for common tasks, like decompressing files and unpacking tar and zip archives. See the API Reference for a full list.
Let’s say our data file is actually a zip (or tar) archive with a collection of files.
We may want to store an unpacked version of the archive or extract just a
single file from it. We can do both operations with the
For example, to extract a single file from a zip archive:
from pooch import Unzip def fetch_zipped_file(): """ Load a large zipped sample data as a pandas.DataFrame. """ # Extract the file "actual-data-file.txt" from the archive unpack = Unzip(members=["actual-data-file.txt"]) # Pass in the processor to unzip the data file fnames = GOODBOY.fetch("zipped-data-file.zip", processor=unpack) # Returns the paths of all extract members (in our case, only one) fname = fnames # fname is now the path of the unzipped file ("actual-data-file.txt") # which can be loaded by pandas directly data = pandas.read_csv(fname) return data
Or to extract all files into a folder and return the path to each file:
def fetch_zipped_archive(): """ Load all files from a zipped archive. """ # Pass in the processor to unzip the data file fnames = GOODBOY.fetch("zipped-archive.zip", processor=Unzip()) data = [pandas.read_csv(fname) for fname in fnames] return data
pooch.Untar to do the exact same for tar archives (with optional
If you have a compressed file that is not an archive (zip or tar), you can use
pooch.Decompress to decompress it after download. For example, large
binary files can be compressed with
gzip to reduce download times but will
need to be decompressed before loading, which can be slow. You can trade
storage space for speed by keeping a decompressed copy of the file:
from pooch import Decompress def fetch_compressed_file(): """ Load a large binary file that has been gzip compressed. """ # Pass in the processor to decompress the file on download fname = GOODBOY.fetch("large-binary-file.npy.gz", processor=Decompress()) # The file returned is the decompressed version which can be loaded by # numpy data = numpy.load(fname) return data
"large-binary-file.npy.gz.decomp" as the
decompressed file name by default. You can change this behaviour by passing a
file name instead:
import os from pooch import Decompress def fetch_compressed_file(): """ Load a large binary file that has been gzip compressed. """ # Pass in the processor to decompress the file on download fname = GOODBOY.fetch("large-binary-file.npy.gz", processor=Decompress(name="a-different-file-name.npy"), ) # The file returned is now named "a-different-file-name.npy" data = numpy.load(fname) return data
name can cause existing data to be lost! For example, if
a file already exists with the specified name it will be overwritten with
the new decompressed file content. Use this option with caution.
Let’s say we want to implement the
pooch.Unzip processor ourselves to
extract a single file from the archive. We could do that with the following
import os from zipfile import ZipFile def unpack(fname, action, pup): """ Post-processing hook to unzip a file and return the unzipped file name. Parameters ---------- fname : str Full path of the zipped file in local storage action : str One of "download" (file doesn't exist and will download), "update" (file is outdated and will download), and "fetch" (file exists and is updated so no download). pup : Pooch The instance of Pooch that called the processor function. Returns ------- fname : str The full path to the unzipped file. (Return the same fname is your processor doesn't modify the file). """ # Create a new name for the unzipped file. Appending something to the # name is a relatively safe way of making sure there are no clashes # with other files in the registry. unzipped = fname + ".unzipped" # Don't unzip if file already exists and is not being downloaded if action in ("update", "download") or not os.path.exists(unzipped): with ZipFile(fname, "r") as zip_file: # Extract the data file from within the archive with zip_file.open("actual-data-file.txt") as data_file: # Save it to our desired file name with open(unzipped, "wb") as output: output.write(data_file.read()) # Return the path of the unzipped file return unzipped def fetch_zipped_file(): """ Load a large zipped sample data as a pandas.DataFrame. """ # Pass in the processor to unzip the data file fname = GOODBOY.fetch("zipped-data-file.zip", processor=unpack) # fname is now the path of the unzipped file which can be loaded by # pandas directly data = pandas.read_csv(fname) return data
Similarly, you could build any custom processor function so long as it receives
fname, action, pup arguments. Example use cases for this would be:
Converting data from a download-friendly format (compressed and minimal file size) to a more user friendly format (easy to open and fast to load into memory).
Add missing metadata to data from public servers. You might be using public data that has known issues (poorly formated entries, missing metadata, etc) which can be fixed when the file is downloaded.
The main advantage to using a processor for these actions is that they are performed only when the file is downloaded. A modified version of the file can be kept on disk so that loading the file is easier. This is particularly convenient if the processor task takes a long time to run.