feat: Make vicinity typing generic

[Pringled]

Vicinity has always accepted any item type at runtime, but the type for SimilarityItem was hardcoded to list[tuple[str, float]], making query() appear to return string items regardless of what was actually stored. This makes it impossible to use Vicinity with typed item types without resorting to cast() or type: ignore. This PR fixes that by changing the return type to a generic type.

For example, the following now works correctly:

from dataclasses import dataclass                                                         
from vicinity import Vicinity                                                             
import numpy as np                                                                      
                                                                                        
@dataclass                                                                              
class Document:
    title: str
    url: str
                                                                                        
docs = [Document("Cats", "/cats"), Document("Dogs", "/dogs")]                   
vectors = np.random.rand(2, 4).astype(np.float32)                                         
                                                                                        
index: Vicinity[Document] = Vicinity.from_vectors_and_items(vectors, docs)              
query = np.random.rand(1, 4).astype(np.float32)
                                                                                        
hits = index.query(query, k=1)[0]  # list[tuple[Document, float]]
for doc, distance in hits:                                                                
    print(doc.url)  # this would have failed before with a type error since doc was always assumed to be a str

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Files with missing lines	Coverage Δ
vicinity/datatypes.py	100.00% <100.00%> (ø)
vicinity/vicinity.py	95.42% <100.00%> (-3.13%)	⬇️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.