build-small-hackathon/hackathon-advisor
16
1from __future__ import annotations2 3from hashlib import sha2564from pathlib import Path5 6from hackathon_advisor.data import ProjectIndex, normalize_vector, tokenize7 8 9def load_test_index() -> ProjectIndex:10 return ProjectIndex.from_files(11 Path("data/projects.json"),12 Path("data/project_index.json"),13 query_embedder=test_query_embedder,14 )15 16 17def test_query_embedder(text: str) -> tuple[float, ...]:18 vector = [0.0] * 76819 for token in tokenize(text):20 digest = sha256(token.encode("utf-8")).digest()21 index = int.from_bytes(digest[:2], "big") % len(vector)22 sign = 1.0 if digest[2] % 2 == 0 else -1.023 vector[index] += sign24 if not any(vector):25 vector[0] = 1.026 return normalize_vector(vector)27 