156 lines
6.6 KiB
Python
156 lines
6.6 KiB
Python
|
"""画像 → 图投影：把 MySQL 里权威的画像与事实投影成 Neo4j 的节点与关系。
|
||
|
|
|
||
|
|
**为什么要有图**（这是它存在的理由，不是技术偏好）：投顾要靠"买过 A 的客户还买过什么"
|
||
|
|
这类多跳关系做组合推荐，风控要靠关系网络发现关联账户与资金链路；这些查询用关系库写
|
||
|
|
既别扭又慢，用图是顺手的事。
|
||
|
|
|
||
|
|
设计约束：
|
||
|
|
|
||
|
|
1. **单向投影**：MySQL 是唯一真相，图库只是派生视图。图数据损坏或模型变更时可以直接重建
|
||
|
|
（拿同一份画像重投影即可），因此这里**不需要与 MySQL 保持事务一致**。
|
||
|
|
2. **幂等**：全部用 MERGE，重复投影不会产生重复节点或关系。
|
||
|
|
3. **降级**：图库不可用时返回 `degraded`，**绝不把异常抛给画像重建这类主链路**
|
||
|
|
—— 投顾推荐可以暂时没有图，但不能因为图挂了就让画像更新失败。
|
||
|
|
4. **只投影"已确认"的事实**：与画像同源（`user_facts`），不直接读原始记忆，
|
||
|
|
保证图里的偏好标签与画像口径一致；否则同一件事在画像和图里会有两种说法。
|
||
|
|
"""
|
||
|
|
|
||
|
|
from dataclasses import dataclass
|
||
|
|
from typing import Any
|
||
|
|
|
||
|
|
from sqlalchemy import select, text
|
||
|
|
from sqlalchemy.ext.asyncio import AsyncSession
|
||
|
|
|
||
|
|
from app.model.profile import UserFact
|
||
|
|
from app.service.graph_model import merge_node_clause, node_spec, relation_allowed
|
||
|
|
from app.service.relationship_service import RelationshipService
|
||
|
|
|
||
|
|
# 事实键 → (关系类型, 目标节点类型)
|
||
|
|
FACT_PROJECTION: dict[str, tuple[str, str]] = {
|
||
|
|
"preference:risk_level": ("PREFERS", "tag"),
|
||
|
|
"preference:asset_class": ("PREFERS", "tag"),
|
||
|
|
"preference:horizon": ("HAS_GOAL", "tag"),
|
||
|
|
"profile:family": ("PREFERS", "tag"),
|
||
|
|
}
|
||
|
|
|
||
|
|
MAX_EDGES_PER_CUSTOMER = 200
|
||
|
|
|
||
|
|
|
||
|
|
@dataclass(frozen=True)
|
||
|
|
class ProjectionOutcome:
|
||
|
|
relations: int = 0
|
||
|
|
degraded: bool = False
|
||
|
|
reason: str = ""
|
||
|
|
|
||
|
|
|
||
|
|
class ProfileGraphProjectionService:
|
||
|
|
def __init__(
|
||
|
|
self, session: AsyncSession, relationships: RelationshipService | None
|
||
|
|
) -> None:
|
||
|
|
self.session = session
|
||
|
|
self.relationships = relationships
|
||
|
|
|
||
|
|
async def project_customer(self, customer_id: int) -> ProjectionOutcome:
|
||
|
|
"""把该客户的画像事实投影到图中。"""
|
||
|
|
if self.relationships is None:
|
||
|
|
return ProjectionOutcome(degraded=True, reason="graph_client_unavailable")
|
||
|
|
|
||
|
|
edges = await self._edges(customer_id)
|
||
|
|
written = 0
|
||
|
|
try:
|
||
|
|
for relation, target_type, target_id in edges[:MAX_EDGES_PER_CUSTOMER]:
|
||
|
|
await self._merge_edge(customer_id, relation, target_type, target_id)
|
||
|
|
written += 1
|
||
|
|
except Exception as exc:
|
||
|
|
# 图库故障：如实标记降级，不伪装成功（与 Milvus/Neo4j 投影清理同一取向）
|
||
|
|
return ProjectionOutcome(relations=written, degraded=True,
|
||
|
|
reason=f"graph_write_failed:{type(exc).__name__}")
|
||
|
|
return ProjectionOutcome(relations=written)
|
||
|
|
|
||
|
|
async def delete_customer(self, customer_id: int) -> ProjectionOutcome:
|
||
|
|
"""删除该客户在图中的节点及其全部关系。
|
||
|
|
|
||
|
|
这是此前缺失的**投影删除客户端**：记忆失效或客户销户后，若图里仍留着旧关系，
|
||
|
|
投顾会拿过期偏好做推荐、风控会看到已不存在的关系网络。
|
||
|
|
`DETACH DELETE` 同时清理关系，避免留下悬挂边。
|
||
|
|
"""
|
||
|
|
if self.relationships is None:
|
||
|
|
return ProjectionOutcome(degraded=True, reason="graph_client_unavailable")
|
||
|
|
try:
|
||
|
|
await self.relationships.driver.execute_query(
|
||
|
|
"MATCH (c:Customer {customer_id: $customer_id}) DETACH DELETE c",
|
||
|
|
customer_id=int(customer_id),
|
||
|
|
)
|
||
|
|
except Exception as exc:
|
||
|
|
return ProjectionOutcome(degraded=True,
|
||
|
|
reason=f"graph_delete_failed:{type(exc).__name__}")
|
||
|
|
return ProjectionOutcome()
|
||
|
|
|
||
|
|
# ---------- 内部 ----------
|
||
|
|
|
||
|
|
async def _edges(self, customer_id: int) -> list[tuple[str, str, str]]:
|
||
|
|
"""收集要投影的边：(关系, 目标节点类型, 目标节点标识)。"""
|
||
|
|
edges: list[tuple[str, str, str]] = []
|
||
|
|
facts = list(await self.session.scalars(
|
||
|
|
select(UserFact).where(UserFact.customer_id == customer_id)
|
||
|
|
))
|
||
|
|
for fact in facts:
|
||
|
|
key = str(fact.fact_key)
|
||
|
|
projection = FACT_PROJECTION.get(key)
|
||
|
|
if projection is None:
|
||
|
|
continue
|
||
|
|
relation, target_type = projection
|
||
|
|
edges.append((relation, target_type, f"{key}={self._text(fact.fact_value)}"))
|
||
|
|
for product_code in await self._holdings(customer_id):
|
||
|
|
edges.append(("HOLDS", "product", product_code))
|
||
|
|
return edges
|
||
|
|
|
||
|
|
async def _holdings(self, customer_id: int) -> list[str]:
|
||
|
|
"""客户当前持仓的产品代码。
|
||
|
|
|
||
|
|
交易模块尚未产生数据，或表结构与预期不符时返回空列表——持仓投影自动跳过，
|
||
|
|
而不是让整次投影失败。等交易侧就绪后无需改这里的代码。
|
||
|
|
"""
|
||
|
|
try:
|
||
|
|
rows = await self.session.execute(text(
|
||
|
|
"SELECT p.product_code FROM fin_holding h "
|
||
|
|
"JOIN fin_product p ON p.id = h.product_id "
|
||
|
|
"WHERE h.customer_id = :customer_id LIMIT 50"
|
||
|
|
), {"customer_id": customer_id})
|
||
|
|
except Exception:
|
||
|
|
return []
|
||
|
|
return [str(row[0]) for row in rows if row[0]]
|
||
|
|
|
||
|
|
async def _merge_edge(
|
||
|
|
self, customer_id: int, relation: str, target_type: str, target_id: str
|
||
|
|
) -> None:
|
||
|
|
if not relation_allowed(relation, "customer", target_type):
|
||
|
|
raise ValueError(f"relationship {relation} cannot connect customer -> {target_type}")
|
||
|
|
source = node_spec("customer")
|
||
|
|
target = node_spec(target_type)
|
||
|
|
query = (
|
||
|
|
f"{merge_node_clause('a', source, 'source_id')} "
|
||
|
|
f"{merge_node_clause('b', target, 'target_id')} "
|
||
|
|
f"MERGE (a)-[r:{relation}]->(b) "
|
||
|
|
"SET r.source = $source, r.updated_at = $updated_at"
|
||
|
|
)
|
||
|
|
await self.relationships.driver.execute_query( # type: ignore[union-attr]
|
||
|
|
query,
|
||
|
|
source_id=node_spec("customer").cast(customer_id),
|
||
|
|
target_id=target_id,
|
||
|
|
source="user_facts",
|
||
|
|
updated_at=_now_iso(),
|
||
|
|
)
|
||
|
|
|
||
|
|
@staticmethod
|
||
|
|
def _text(value: Any) -> str:
|
||
|
|
if isinstance(value, str):
|
||
|
|
return value.strip().strip('"')
|
||
|
|
return str(value)
|
||
|
|
|
||
|
|
|
||
|
|
def _now_iso() -> str:
|
||
|
|
from datetime import UTC, datetime
|
||
|
|
|
||
|
|
return datetime.now(UTC).replace(tzinfo=None).isoformat()
|