feat: add encoder_by_name factory function

[sumanjeet0012]

Fixes #39

Description

This pull request adds an encoder_by_name() factory function, which enables creating an Encoder from a single-character prefix as well as the standard name, bringing the python implementation into alignment with the go-multibase reference implementation.

Changes

• Created encoder_by_name() in multibase/multibase.py which inspects the given string and resolves it if it matches either a name or a prefix.
• Exported encoder_by_name from __init__.py.
• Added tests for encoder_by_name demonstrating successful construction from prefix, name, and base256emoji, as well as proper exception handling for unsupported/invalid strings.

[acul71]

Review summary (maintainer)

Thanks for the clean encoder_by_name() addition — it matches issue #39 and go-multibase’s name/prefix factory pattern, keeps Encoder.__init__ backward compatible, and local/CI checks look good.

Blocker (must fix)

Missing newsfragment. Please add:

• newsfragments/39.feature.rst

with a short user-facing ReST note (trailing newline required), e.g. that encoder_by_name() creates an Encoder from an encoding name or multibase prefix character.

Also please address

1. Rebase onto origin/master — branch is 6 commits behind (dry-run merge currently clean, but please sync).
2. Input hardening — encoder_by_name(b"f") currently yields encoding == b"f"; non-str values like None raise AttributeError. Prefer requiring str or normalizing to the codec name.
3. Docs — README still only shows Encoder('base64'); add a brief encoder_by_name example.
4. Tests (nice-to-have) — assert .encode() via the factory, not only .encoding.

Notes

• No merge conflicts vs current origin/master.
• Informational: overlaps with PR #53 on multibase.py / tests.
• Full review: local downloads/AI-PR-REVIEWS/54/AI-PR-REVIEW-54-0.md.

Requesting changes until the newsfragment (and preferably rebase) is done.