Read empty GitHub files without the HTTP extra

[hoobnn]

Summary

Reading a zero-byte file with only fsspec[github] installed currently raises an ImportError asking for fsspec[http]. GitHub returns these files as encoding="base64" with an empty content string, which was being mistaken for content that needs a separate download.

Use the response encoding to recognize inline content, including empty files. Large files with encoding="none" and Git LFS pointers still use the HTTP download path.

Tests

• The new empty-file regression fails before the fix.
• test_github_paths.py and test_memory.py pass (85 tests); the new tests mock requests and need no network.
• A request for this repository's empty fsspec/implementations/__init__.py returns encoding="base64" with empty content. The live large-file and LFS tests in test_github.py also pass.

I used OpenAI Codex and Claude Code to investigate the issue, write the fix and tests, and run the checks. I reviewed the code and this description myself.

Assisted by Claude Code