A thin asyncio wrapper for Azure Storage

I wanted a low-latency asyncio library for accessing Azure Storage via persistent connections (essentially tables and queues) and the existing Python SDK (which is auto-generated from API specs) was both too high-level and not async-ready.

In comparison with its somewhat slow pace, this can now (as published) upload roughly 2.5K "hello world" blobs/s and enumerate roughly 3K/s across the Atlantic, which is pretty decent.

This is an intentionally low-level wrapper, and meant largely for my own consumption. However, pull requests are welcome.

Since this targets Python 3.6+ (Ubuntu still only ships with 3.6.9 in 2020), I'm making liberal use of type annotations (which are still not uniform) and f-strings.

• SAS Token support
• advanced message semantics (including queueing status codes)
• message peek/clear/update
• blob enumeration/creation/tier management
• blob retrieval/deletion
• blob container enumeration/creation/deletion
• queue metadata
• queue enumeration
• message queueing/retrieval/deletion
• queue creation/deletion
• table batch operations (batch update implemented, result parsing not yet)
• table entry creation/updating/deletion/querying (with EDM annotation of supported types)
• table creation/deletion/querying

• Python 3.6
• aiohttp
• ujson (optional)
• uvloop (optional)