study-buddy / app /services /memory_candidates.py
GitHub Actions
deploy d092bea3608b7a29952f16357fda39b7a29e399b
2e818da
Raw
History Blame Contribute Delete
1.54 kB
"""Typed proposals emitted by Commit before either memory backend is mutated."""
from __future__ import annotations
import hashlib
import re
from datetime import datetime, timezone
from typing import Literal
from pydantic import BaseModel, ConfigDict, Field
class MemoryCandidate(BaseModel):
model_config = ConfigDict(strict=True, extra="forbid")
candidate_id: str
destination: Literal["project", "student"]
project_id: str
kind: str
statement: str
attribution: Literal[
"explicit_student",
"idea_observer_interaction",
"idea_observer_profile_proposal",
]
confidence: float = Field(ge=0.0, le=1.0)
evidence_ids: list[str] = Field(default_factory=list)
interaction_ids: list[str] = Field(default_factory=list)
observed_at: datetime = Field(default_factory=lambda: datetime.now(timezone.utc))
@property
def recurrence_key(self) -> str:
normalized = re.sub(r"\s+", " ", self.statement.casefold()).strip()
return hashlib.sha256(f"{self.kind}\0{normalized}".encode("utf-8")).hexdigest()
def make_candidate_id(
destination: str,
project_id: str,
kind: str,
statement: str,
) -> str:
normalized = re.sub(r"\s+", " ", statement.casefold()).strip()
digest = hashlib.sha256(
f"{destination}\0{project_id}\0{kind}\0{normalized}".encode("utf-8")
).hexdigest()
return f"mc_{digest}"
def interaction_ref(value: str) -> str:
return "ji_" + hashlib.sha256(value.strip().encode("utf-8")).hexdigest()[:24]