fsspec: Filesystem interfaces for Python
Filesystem Spec (fsspec) is a project to provide a unified pythonic interface to
local, remote and embedded file systems and bytes storage.
Brief Overview
There are many places to store bytes, from in memory, to the local disk, cluster
distributed storage, to the cloud. Many files also contain internal mappings of names to bytes,
maybe in a hierarchical directory-oriented tree. Working with all these different
storage media, and their associated libraries, is a pain. fsspec exists to
provide a familiar API that will work the same whatever the storage backend.
As much as possible, we iron out the quirks specific to each implementation,
so you need do no more than provide credentials for each service you access
(if needed) and thereafter not have to worry about the implementation again.
Why
fsspec provides two main concepts: a set of filesystem classes with uniform APIs
(i.e., functions such as cp, rm, cat, mkdir, …) supplying operations on a range of
storage systems; and top-level convenience functions like fsspec.open(), to allow
you to quickly get from a URL to a file-like object that you can use with a third-party
library or your own code.
The section Background gives motivation and history of this project, but most users will want to skip straight to Usage to find out how to use the package and Features of fsspec to see the long list of added functionality included along with the basic file-system interface.
Who uses fsspec?
You can use fsspec’s file objects with any python function that accepts
file objects, because of duck typing.
You may well be using fsspec already without knowing it.
The following libraries use fsspec internally for path and file handling:
Dask, the parallel, out-of-core and distributed programming platform
Intake, the data source cataloguing and loading library and its plugins
pandas, the tabular data analysis package
xarray and zarr, multidimensional array storage and labelled operations
DVC, version control system for machine learning projects
Kedro, a Python framework for reproducible, maintainable and modular data science code
pyxet, a Python library for mounting and accessing very large datasets from XetHub
Huggingface🤗 Datasets, a popular library to load&manipulate data for Deep Learning models
fsspec filesystems are also supported by:
pyarrow, the in-memory data layout engine
petl, a general purpose package for extracting, transforming and loading tables of data.
… plus many more that we don’t know about.
Installation
fsspec can be installed from PyPI or conda and has no dependencies of its own
pip install fsspec
conda install -c conda-forge fsspec
Not all filesystem implementations are available without installing extra dependencies. For example to be able to access data in GCS, you can use the optional pip install syntax below, or install the specific package required
pip install fsspec[gcs]
conda install -c conda-forge gcsfs
fsspec attempts to provide the right message when you attempt to use a filesystem
for which you need additional dependencies.
The current list of known implementations can be found as follows
from fsspec.registry import known_implementations
known_implementations