26 lines
952 B
Python
26 lines
952 B
Python
|
from __future__ import annotations
|
||
|
|
|
||
|
|
from dataclasses import dataclass, field
|
||
|
|
from datetime import datetime
|
||
|
|
from typing import Optional
|
||
|
|
|
||
|
|
|
||
|
|
@dataclass
|
||
|
|
class Envelope:
|
||
|
|
"""Frozen cross-source envelope — the only schema contract locked in upfront.
|
||
|
|
|
||
|
|
Additive: new fields may be added but existing ones must not be removed or renamed.
|
||
|
|
All fields mirror the `envelope` table in kb-postgres (init/001_envelope.sql).
|
||
|
|
"""
|
||
|
|
|
||
|
|
id: str
|
||
|
|
source: str # fastmail | gmail | …
|
||
|
|
ts: datetime # message date, must be UTC-aware
|
||
|
|
raw_ref: str # relative path to .eml in archive
|
||
|
|
geo: Optional[dict] = None # null for mails; filled by layer-3 enrich
|
||
|
|
entities: list = field(default_factory=list)
|
||
|
|
|
||
|
|
def __post_init__(self) -> None:
|
||
|
|
if self.ts.tzinfo is None:
|
||
|
|
raise ValueError("Envelope.ts must be timezone-aware (UTC)")
|