from hashlib import sha256
from math import sqrt
from typing import Any


def embed_text(text: str, size: int = 32) -> list[float]:
    """Deterministic local embedding used for offline demo and tests."""

    buckets = [0.0] * size
    for word in text.lower().split():
        digest = sha256(word.encode()).digest()
        buckets[digest[0] % size] += 1.0
    norm = sqrt(sum(v * v for v in buckets)) or 1.0
    return [v / norm for v in buckets]


class MemoryManager:
    """Stores semantic memories behind a Qdrant-like interface with local fallback."""

    def __init__(self) -> None:
        self._items: list[dict[str, Any]] = []

    def upsert(self, memory_type: str, text: str, metadata: dict[str, Any] | None = None) -> dict[str, Any]:
        item = {
            "id": sha256(f"{memory_type}:{text}".encode()).hexdigest()[:16],
            "type": memory_type,
            "text": text,
            "metadata": metadata or {},
            "embedding": embed_text(text),
        }
        self._items = [existing for existing in self._items if existing["id"] != item["id"]]
        self._items.append(item)
        return item

    def retrieve(self, query: str, limit: int = 5) -> list[dict[str, Any]]:
        query_vec = embed_text(query)

        def score(item: dict[str, Any]) -> float:
            return sum(a * b for a, b in zip(query_vec, item["embedding"], strict=True))

        ranked = sorted(self._items, key=score, reverse=True)
        return [{k: v for k, v in item.items() if k != "embedding"} | {"score": score(item)} for item in ranked[:limit]]

    def seed_demo(self) -> None:
        self.upsert("customer_preference", "Customer prefers white flowers and a classic minimal arrangement.", {"customer": "Demo"})
        self.upsert("vendor", "Aegean Blooms is the preferred florist for white flower arrangements.", {"vendor": "Aegean Blooms"})
        self.upsert("meeting", "Previous meeting agreed to compare venues before reserving July dates.", {"meeting_id": "prior-1"})