[ARROW-13699] [Python][Doc] Refactor the FileSystem Interface documentation

• Type: Improvement
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Fix Version/s: 6.0.0
• Component/s: Documentation
• Labels: pull-request-available
• External issue URL: https://github.com/apache/arrow/issues/29334
• Assignee: Nathanael Leaute
• Reporter: Nathanael Leaute
• Created: 23/Aug/21 17:37
• Updated: 11/Jan/23 08:34
• Resolved: 15/Sep/21 14:53

Description

As a python developer working with different cloud vendors and their storages I'd like to quickly jump to code examples on how to read and write files for each filesystem.

The documentation concerned is the python/filesystem doc: https://arrow.apache.org/docs/python/filesystems.html

I find the information is a bit scattered and could be improved by having the following organisation.

Filesystem Interface

overview of the Pyarrow FS Interface

Usage

Local Filesystem

description

Writing files

code example

Listing files

code example

Reading files

code example

S3 Filesystem

description / configuration

Writing files

code example

Listing files

code example

Reading files

code example

Hadoop Filesystem

description / configuration

Writing files

code example

Listing files

code example

Reading files

code example

Extending to fsspec-compatible filesystems

description

Google Cloud Storage

code example

Azure

code example

That way a developer working on s3 can directly jump to the section of interest and start experimenting with the code examples.
Additionally if new python bindings are created for a "Arrow native" filesystem the documentation can be extended with a new section in same vein as the other.

Issue Links

• links to: GitHub Pull Request #11040