feat(ai): add tenant-safe hierarchical expense learning
This commit is contained in:
caoxiaozhu
@@ -0,0 +1,133 @@
|
||||
"""add hierarchical expense memory governance fields
|
||||
|
||||
Revision ID: 20260716_0007
|
||||
Revises: 20260716_0006
|
||||
Create Date: 2026-07-16 12:05:00
|
||||
"""
|
||||
|
||||
from collections.abc import Sequence
|
||||
|
||||
import sqlalchemy as sa
|
||||
|
||||
from alembic import op
|
||||
|
||||
revision: str = "20260716_0007"
|
||||
down_revision: str | None = "20260716_0006"
|
||||
branch_labels: str | Sequence[str] | None = None
|
||||
depends_on: str | Sequence[str] | None = None
|
||||
|
||||
|
||||
def upgrade() -> None:
|
||||
op.add_column(
|
||||
"memory_entries",
|
||||
sa.Column(
|
||||
"origin_type",
|
||||
sa.String(length=24),
|
||||
server_default="learned",
|
||||
nullable=False,
|
||||
),
|
||||
)
|
||||
op.add_column(
|
||||
"memory_entries",
|
||||
sa.Column("managed_by", sa.String(length=255), nullable=True),
|
||||
)
|
||||
op.add_column(
|
||||
"memory_entries",
|
||||
sa.Column("managed_at", sa.DateTime(timezone=True), nullable=True),
|
||||
)
|
||||
op.add_column(
|
||||
"memory_entries",
|
||||
sa.Column("management_reason", sa.String(length=255), nullable=True),
|
||||
)
|
||||
op.add_column(
|
||||
"memory_entries",
|
||||
sa.Column("policy_version", sa.String(length=64), nullable=True),
|
||||
)
|
||||
|
||||
op.drop_constraint(
|
||||
"ck_memory_entries_scope_type",
|
||||
"memory_entries",
|
||||
type_="check",
|
||||
)
|
||||
op.create_check_constraint(
|
||||
"ck_memory_entries_scope_type",
|
||||
"memory_entries",
|
||||
"scope_type IN ('user', 'department', 'enterprise')",
|
||||
)
|
||||
op.create_check_constraint(
|
||||
"ck_memory_entries_origin_type",
|
||||
"memory_entries",
|
||||
"origin_type IN ('learned', 'admin_managed')",
|
||||
)
|
||||
op.create_check_constraint(
|
||||
"ck_memory_entries_scope_origin",
|
||||
"memory_entries",
|
||||
"(scope_type = 'user' AND origin_type = 'learned') OR "
|
||||
"(scope_type IN ('department', 'enterprise') "
|
||||
"AND origin_type = 'admin_managed')",
|
||||
)
|
||||
op.create_check_constraint(
|
||||
"ck_memory_entries_enterprise_scope",
|
||||
"memory_entries",
|
||||
"scope_type != 'enterprise' OR scope_id = tenant_id",
|
||||
)
|
||||
op.create_check_constraint(
|
||||
"ck_memory_entries_management_audit",
|
||||
"memory_entries",
|
||||
"(origin_type = 'learned' AND managed_by IS NULL "
|
||||
"AND managed_at IS NULL AND management_reason IS NULL) OR "
|
||||
"(origin_type = 'admin_managed' AND managed_by IS NOT NULL "
|
||||
"AND length(trim(managed_by)) > 0 AND managed_at IS NOT NULL "
|
||||
"AND management_reason IS NOT NULL "
|
||||
"AND length(trim(management_reason)) > 0 "
|
||||
"AND policy_version IS NOT NULL AND length(trim(policy_version)) > 0)",
|
||||
)
|
||||
|
||||
|
||||
def downgrade() -> None:
|
||||
op.drop_constraint(
|
||||
"ck_memory_entries_management_audit",
|
||||
"memory_entries",
|
||||
type_="check",
|
||||
)
|
||||
op.drop_constraint(
|
||||
"ck_memory_entries_enterprise_scope",
|
||||
"memory_entries",
|
||||
type_="check",
|
||||
)
|
||||
op.drop_constraint(
|
||||
"ck_memory_entries_scope_origin",
|
||||
"memory_entries",
|
||||
type_="check",
|
||||
)
|
||||
op.drop_constraint(
|
||||
"ck_memory_entries_origin_type",
|
||||
"memory_entries",
|
||||
type_="check",
|
||||
)
|
||||
op.drop_constraint(
|
||||
"ck_memory_entries_scope_type",
|
||||
"memory_entries",
|
||||
type_="check",
|
||||
)
|
||||
# 0006 只认识个人记忆。组织记忆无法无损映射回旧结构，因此降级时
|
||||
# 先解除组织版本间的自引用并删除组织作用域记录，个人记忆完整保留。
|
||||
op.execute(
|
||||
"UPDATE memory_entries SET superseded_by_id = NULL "
|
||||
"WHERE scope_type IN ('department', 'enterprise')"
|
||||
)
|
||||
op.execute(
|
||||
"DELETE FROM memory_entries "
|
||||
"WHERE scope_type IN ('department', 'enterprise')"
|
||||
)
|
||||
op.create_check_constraint(
|
||||
"ck_memory_entries_scope_type",
|
||||
"memory_entries",
|
||||
"scope_type = 'user'",
|
||||
)
|
||||
|
||||
op.drop_column("memory_entries", "policy_version")
|
||||
op.drop_column("memory_entries", "management_reason")
|
||||
op.drop_column("memory_entries", "managed_at")
|
||||
op.drop_column("memory_entries", "managed_by")
|
||||
op.drop_column("memory_entries", "origin_type")
|
||||
@@ -0,0 +1,435 @@
|
||||
"""adopt tenant-safe historical case learning tables
|
||||
|
||||
Revision ID: 20260716_0008
|
||||
Revises: 20260716_0007
|
||||
Create Date: 2026-07-16 13:20:00
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from collections.abc import Sequence
|
||||
|
||||
import sqlalchemy as sa
|
||||
|
||||
from alembic import op
|
||||
|
||||
revision: str = "20260716_0008"
|
||||
down_revision: str | None = "20260716_0007"
|
||||
branch_labels: str | Sequence[str] | None = None
|
||||
depends_on: str | Sequence[str] | None = None
|
||||
|
||||
|
||||
def _require_postgresql() -> None:
|
||||
dialect_name = op.get_bind().dialect.name
|
||||
if dialect_name != "postgresql":
|
||||
raise RuntimeError(
|
||||
"20260716_0008 only supports PostgreSQL; "
|
||||
f"refusing to mutate {dialect_name} without transactional constraint DDL"
|
||||
)
|
||||
|
||||
|
||||
def _inspector() -> sa.Inspector:
|
||||
return sa.inspect(op.get_bind())
|
||||
|
||||
|
||||
def _table_exists(table_name: str) -> bool:
|
||||
return table_name in _inspector().get_table_names()
|
||||
|
||||
|
||||
def _columns(table_name: str) -> set[str]:
|
||||
return {str(item["name"]) for item in _inspector().get_columns(table_name)}
|
||||
|
||||
|
||||
def _indexes(table_name: str) -> dict[str, dict[str, object]]:
|
||||
return {
|
||||
str(item["name"]): item for item in _inspector().get_indexes(table_name) if item.get("name")
|
||||
}
|
||||
|
||||
|
||||
def _unique_constraints(table_name: str) -> dict[str, dict[str, object]]:
|
||||
return {
|
||||
str(item["name"]): item
|
||||
for item in _inspector().get_unique_constraints(table_name)
|
||||
if item.get("name")
|
||||
}
|
||||
|
||||
|
||||
def _foreign_keys(table_name: str) -> list[dict[str, object]]:
|
||||
return list(_inspector().get_foreign_keys(table_name))
|
||||
|
||||
|
||||
def _ensure_index(
|
||||
table_name: str,
|
||||
index_name: str,
|
||||
columns: list[str],
|
||||
*,
|
||||
unique: bool = False,
|
||||
) -> None:
|
||||
if index_name not in _indexes(table_name):
|
||||
op.create_index(index_name, table_name, columns, unique=unique)
|
||||
|
||||
|
||||
def _drop_legacy_single_key_uniqueness(table_name: str, key_column: str) -> None:
|
||||
for name, item in list(_unique_constraints(table_name).items()):
|
||||
if tuple(item.get("column_names") or ()) == (key_column,):
|
||||
op.drop_constraint(name, table_name, type_="unique")
|
||||
# PostgreSQL 会把唯一约束的支撑索引同时返回给 get_indexes；先删除约束并
|
||||
# 重新反射，避免对支撑索引执行 DROP INDEX 导致 dependency error。
|
||||
for name, item in list(_indexes(table_name).items()):
|
||||
if tuple(item.get("column_names") or ()) == (key_column,) and bool(item.get("unique")):
|
||||
op.drop_index(name, table_name=table_name)
|
||||
|
||||
|
||||
def _ensure_composite_unique(
|
||||
table_name: str,
|
||||
constraint_name: str,
|
||||
columns: list[str],
|
||||
) -> None:
|
||||
expected = tuple(columns)
|
||||
if any(
|
||||
tuple(item.get("column_names") or ()) == expected
|
||||
for item in _unique_constraints(table_name).values()
|
||||
):
|
||||
return
|
||||
op.create_unique_constraint(constraint_name, table_name, columns)
|
||||
|
||||
|
||||
def _drop_index_if_exists(table_name: str, index_name: str) -> None:
|
||||
if index_name in _indexes(table_name):
|
||||
op.drop_index(index_name, table_name=table_name)
|
||||
|
||||
|
||||
def _drop_unique_if_exists(table_name: str, constraint_name: str) -> None:
|
||||
if constraint_name in _unique_constraints(table_name):
|
||||
op.drop_constraint(constraint_name, table_name, type_="unique")
|
||||
|
||||
|
||||
def _require_lossless_default_tenant_downgrade(table_name: str) -> None:
|
||||
non_default_count = int(
|
||||
op.get_bind().scalar(
|
||||
sa.text(
|
||||
f"SELECT COUNT(*) FROM {table_name} "
|
||||
"WHERE tenant_id IS DISTINCT FROM 'default'"
|
||||
)
|
||||
)
|
||||
or 0
|
||||
)
|
||||
if non_default_count:
|
||||
raise RuntimeError(
|
||||
f"cannot downgrade {table_name}: non-default tenant data would lose isolation"
|
||||
)
|
||||
|
||||
|
||||
def _require_lossless_few_shot_downgrade() -> None:
|
||||
enriched_count = int(
|
||||
op.get_bind().scalar(
|
||||
sa.text(
|
||||
"SELECT COUNT(*) FROM few_shot_samples "
|
||||
"WHERE COALESCE(TRIM(policy_ref), '') <> '' "
|
||||
"OR COALESCE(TRIM(rule_version), '') <> ''"
|
||||
)
|
||||
)
|
||||
or 0
|
||||
)
|
||||
if enriched_count:
|
||||
raise RuntimeError(
|
||||
"cannot downgrade few_shot_samples: "
|
||||
f"{enriched_count} row(s) contain policy_ref or rule_version data"
|
||||
)
|
||||
|
||||
|
||||
def _drop_risk_observation_claim_foreign_keys() -> None:
|
||||
"""统一采用软引用，避免 Head 结构取决于 expense_claims 的创建时机。"""
|
||||
for item in _foreign_keys("risk_observations"):
|
||||
if (
|
||||
tuple(item.get("constrained_columns") or ()) == ("claim_id",)
|
||||
and item.get("referred_table") == "expense_claims"
|
||||
and item.get("name")
|
||||
):
|
||||
op.drop_constraint(
|
||||
str(item["name"]),
|
||||
"risk_observations",
|
||||
type_="foreignkey",
|
||||
)
|
||||
|
||||
|
||||
def _create_risk_observations() -> None:
|
||||
op.create_table(
|
||||
"risk_observations",
|
||||
sa.Column("id", sa.String(length=36), nullable=False),
|
||||
sa.Column("tenant_id", sa.String(length=64), server_default="default", nullable=False),
|
||||
sa.Column("observation_key", sa.String(length=160), nullable=False),
|
||||
sa.Column("subject_type", sa.String(length=50), nullable=False),
|
||||
sa.Column("subject_key", sa.String(length=160), nullable=False),
|
||||
sa.Column("subject_label", sa.String(length=160), nullable=False, server_default=""),
|
||||
# expense_claims 仍由 legacy bootstrap 创建，空库迁移到此版本时并不存在。
|
||||
# claim_id 因此是显式软引用，与 ORM 的 viewonly relationship 保持一致。
|
||||
sa.Column("claim_id", sa.String(length=36), nullable=True),
|
||||
sa.Column("claim_no", sa.String(length=80), nullable=False, server_default=""),
|
||||
sa.Column("run_id", sa.String(length=80), nullable=True),
|
||||
sa.Column("execution_log_id", sa.String(length=36), nullable=True),
|
||||
sa.Column("risk_type", sa.String(length=80), nullable=False),
|
||||
sa.Column("risk_signal", sa.String(length=100), nullable=False),
|
||||
sa.Column("title", sa.String(length=200), nullable=False, server_default=""),
|
||||
sa.Column("description", sa.Text(), nullable=False, server_default=""),
|
||||
sa.Column("risk_score", sa.Integer(), nullable=False, server_default="0"),
|
||||
sa.Column("risk_level", sa.String(length=20), nullable=False),
|
||||
sa.Column("confidence_score", sa.Float(), nullable=False, server_default="0"),
|
||||
sa.Column("control_stage", sa.String(length=50), nullable=False, server_default=""),
|
||||
sa.Column("control_mode", sa.String(length=50), nullable=False, server_default=""),
|
||||
sa.Column("automation_mode", sa.String(length=50), nullable=False, server_default=""),
|
||||
sa.Column("source", sa.String(length=60), nullable=False, server_default=""),
|
||||
sa.Column("algorithm_version", sa.String(length=80), nullable=False, server_default=""),
|
||||
sa.Column("status", sa.String(length=30), nullable=False, server_default="pending_review"),
|
||||
sa.Column(
|
||||
"feedback_status", sa.String(length=30), nullable=False, server_default="unreviewed"
|
||||
),
|
||||
sa.Column("contribution_scores_json", sa.JSON(), nullable=False),
|
||||
sa.Column("baseline_json", sa.JSON(), nullable=False),
|
||||
sa.Column("evidence_json", sa.JSON(), nullable=False),
|
||||
sa.Column("graph_node_keys_json", sa.JSON(), nullable=False),
|
||||
sa.Column("graph_edge_keys_json", sa.JSON(), nullable=False),
|
||||
sa.Column("policy_refs_json", sa.JSON(), nullable=False),
|
||||
sa.Column("similar_case_claim_ids_json", sa.JSON(), nullable=False),
|
||||
sa.Column("ontology_json", sa.JSON(), nullable=False),
|
||||
sa.Column("decision_trace_json", sa.JSON(), nullable=False),
|
||||
sa.Column(
|
||||
"created_at", sa.DateTime(timezone=True), server_default=sa.func.now(), nullable=False
|
||||
),
|
||||
sa.Column(
|
||||
"updated_at", sa.DateTime(timezone=True), server_default=sa.func.now(), nullable=False
|
||||
),
|
||||
sa.PrimaryKeyConstraint("id"),
|
||||
sa.UniqueConstraint("tenant_id", "observation_key", name="uq_risk_observations_tenant_key"),
|
||||
)
|
||||
|
||||
|
||||
def _adopt_risk_observations() -> None:
|
||||
_drop_risk_observation_claim_foreign_keys()
|
||||
if "tenant_id" not in _columns("risk_observations"):
|
||||
op.add_column(
|
||||
"risk_observations",
|
||||
sa.Column("tenant_id", sa.String(length=64), server_default="default", nullable=False),
|
||||
)
|
||||
op.execute(
|
||||
sa.text(
|
||||
"UPDATE risk_observations SET tenant_id = 'default' "
|
||||
"WHERE tenant_id IS NULL OR tenant_id = ''"
|
||||
)
|
||||
)
|
||||
_drop_legacy_single_key_uniqueness("risk_observations", "observation_key")
|
||||
_ensure_composite_unique(
|
||||
"risk_observations",
|
||||
"uq_risk_observations_tenant_key",
|
||||
["tenant_id", "observation_key"],
|
||||
)
|
||||
|
||||
|
||||
def _create_risk_observation_feedback() -> None:
|
||||
op.create_table(
|
||||
"risk_observation_feedback",
|
||||
sa.Column("id", sa.String(length=36), nullable=False),
|
||||
sa.Column("observation_id", sa.String(length=36), nullable=False),
|
||||
sa.Column("feedback_type", sa.String(length=30), nullable=False),
|
||||
sa.Column("action", sa.String(length=50), nullable=False, server_default=""),
|
||||
sa.Column("actor", sa.String(length=100), nullable=False, server_default=""),
|
||||
sa.Column("comment", sa.Text(), nullable=True),
|
||||
sa.Column("payload_json", sa.JSON(), nullable=False),
|
||||
sa.Column(
|
||||
"created_at", sa.DateTime(timezone=True), server_default=sa.func.now(), nullable=False
|
||||
),
|
||||
sa.ForeignKeyConstraint(["observation_id"], ["risk_observations.id"]),
|
||||
sa.PrimaryKeyConstraint("id"),
|
||||
)
|
||||
|
||||
|
||||
def _create_few_shot_samples() -> None:
|
||||
op.create_table(
|
||||
"few_shot_samples",
|
||||
sa.Column("id", sa.String(length=36), nullable=False),
|
||||
sa.Column("tenant_id", sa.String(length=64), server_default="default", nullable=False),
|
||||
sa.Column("sample_key", sa.String(length=160), nullable=False),
|
||||
sa.Column("source_observation_id", sa.String(length=36), nullable=True),
|
||||
sa.Column(
|
||||
"scene", sa.String(length=50), nullable=False, server_default="risk_rule_generation"
|
||||
),
|
||||
sa.Column("policy_ref", sa.String(length=160), nullable=False, server_default=""),
|
||||
sa.Column("rule_version", sa.String(length=80), nullable=False, server_default=""),
|
||||
sa.Column("domain", sa.String(length=50), nullable=False, server_default=""),
|
||||
sa.Column("risk_type", sa.String(length=80), nullable=False, server_default=""),
|
||||
sa.Column("risk_level", sa.String(length=20), nullable=False, server_default=""),
|
||||
sa.Column("label", sa.String(length=30), nullable=False, server_default="confirmed"),
|
||||
sa.Column("case_text", sa.Text(), nullable=False, server_default=""),
|
||||
sa.Column("conclusion_text", sa.Text(), nullable=False, server_default=""),
|
||||
sa.Column("payload_json", sa.JSON(), nullable=False),
|
||||
sa.Column("vector_id", sa.String(length=100), nullable=True),
|
||||
sa.Column("status", sa.String(length=20), nullable=False, server_default="active"),
|
||||
sa.Column("created_at", sa.DateTime(), server_default=sa.func.now(), nullable=False),
|
||||
sa.Column("updated_at", sa.DateTime(), server_default=sa.func.now(), nullable=False),
|
||||
sa.ForeignKeyConstraint(["source_observation_id"], ["risk_observations.id"]),
|
||||
sa.PrimaryKeyConstraint("id"),
|
||||
sa.UniqueConstraint("tenant_id", "sample_key", name="uq_few_shot_samples_tenant_key"),
|
||||
)
|
||||
|
||||
|
||||
def _adopt_few_shot_samples() -> None:
|
||||
existing = _columns("few_shot_samples")
|
||||
for column in (
|
||||
sa.Column("tenant_id", sa.String(length=64), server_default="default", nullable=False),
|
||||
sa.Column("policy_ref", sa.String(length=160), server_default="", nullable=False),
|
||||
sa.Column("rule_version", sa.String(length=80), server_default="", nullable=False),
|
||||
):
|
||||
if column.name not in existing:
|
||||
op.add_column("few_shot_samples", column)
|
||||
op.execute(
|
||||
sa.text(
|
||||
"UPDATE few_shot_samples SET tenant_id = 'default' "
|
||||
"WHERE tenant_id IS NULL OR tenant_id = ''"
|
||||
)
|
||||
)
|
||||
op.execute(sa.text("UPDATE few_shot_samples SET policy_ref = '' WHERE policy_ref IS NULL"))
|
||||
op.execute(sa.text("UPDATE few_shot_samples SET rule_version = '' WHERE rule_version IS NULL"))
|
||||
_drop_legacy_single_key_uniqueness("few_shot_samples", "sample_key")
|
||||
_ensure_composite_unique(
|
||||
"few_shot_samples",
|
||||
"uq_few_shot_samples_tenant_key",
|
||||
["tenant_id", "sample_key"],
|
||||
)
|
||||
|
||||
|
||||
def _ensure_risk_indexes() -> None:
|
||||
definitions = {
|
||||
"ix_risk_observations_tenant_id": ["tenant_id"],
|
||||
"ix_risk_observations_observation_key": ["observation_key"],
|
||||
"ix_risk_observations_subject_type": ["subject_type"],
|
||||
"ix_risk_observations_subject_key": ["subject_key"],
|
||||
"ix_risk_observations_claim_id": ["claim_id"],
|
||||
"ix_risk_observations_claim_no": ["claim_no"],
|
||||
"ix_risk_observations_run_id": ["run_id"],
|
||||
"ix_risk_observations_execution_log_id": ["execution_log_id"],
|
||||
"ix_risk_observations_risk_type": ["risk_type"],
|
||||
"ix_risk_observations_risk_signal": ["risk_signal"],
|
||||
"ix_risk_observations_risk_score": ["risk_score"],
|
||||
"ix_risk_observations_risk_level": ["risk_level"],
|
||||
"ix_risk_observations_source": ["source"],
|
||||
"ix_risk_observations_algorithm_version": ["algorithm_version"],
|
||||
"ix_risk_observations_status": ["status"],
|
||||
"ix_risk_observations_feedback_status": ["feedback_status"],
|
||||
"ix_risk_observations_subject": ["subject_type", "subject_key"],
|
||||
"ix_risk_observations_signal_level": ["risk_signal", "risk_level"],
|
||||
"ix_risk_observations_status_created": ["status", "created_at"],
|
||||
"ix_risk_observations_tenant_status": ["tenant_id", "status", "created_at"],
|
||||
}
|
||||
for name, columns in definitions.items():
|
||||
_ensure_index("risk_observations", name, columns)
|
||||
_ensure_index(
|
||||
"risk_observation_feedback",
|
||||
"ix_risk_observation_feedback_observation_id",
|
||||
["observation_id"],
|
||||
)
|
||||
_ensure_index(
|
||||
"risk_observation_feedback", "ix_risk_observation_feedback_feedback_type", ["feedback_type"]
|
||||
)
|
||||
_ensure_index(
|
||||
"risk_observation_feedback",
|
||||
"ix_risk_observation_feedback_type_created",
|
||||
["feedback_type", "created_at"],
|
||||
)
|
||||
|
||||
|
||||
def _ensure_few_shot_indexes() -> None:
|
||||
definitions = {
|
||||
"ix_few_shot_samples_tenant_id": ["tenant_id"],
|
||||
"ix_few_shot_samples_sample_key": ["sample_key"],
|
||||
"ix_few_shot_samples_source_observation_id": ["source_observation_id"],
|
||||
"ix_few_shot_samples_scene": ["scene"],
|
||||
"ix_few_shot_samples_policy_ref": ["policy_ref"],
|
||||
"ix_few_shot_samples_rule_version": ["rule_version"],
|
||||
"ix_few_shot_samples_domain": ["domain"],
|
||||
"ix_few_shot_samples_risk_type": ["risk_type"],
|
||||
"ix_few_shot_samples_label": ["label"],
|
||||
"ix_few_shot_samples_status": ["status"],
|
||||
"ix_few_shot_samples_scene_label": ["scene", "label"],
|
||||
"ix_few_shot_samples_domain_risk_type": ["domain", "risk_type"],
|
||||
"ix_few_shot_samples_tenant_rule_lookup": [
|
||||
"tenant_id",
|
||||
"scene",
|
||||
"policy_ref",
|
||||
"rule_version",
|
||||
"status",
|
||||
],
|
||||
}
|
||||
for name, columns in definitions.items():
|
||||
_ensure_index("few_shot_samples", name, columns)
|
||||
|
||||
|
||||
def upgrade() -> None:
|
||||
_require_postgresql()
|
||||
if _table_exists("risk_observations"):
|
||||
_adopt_risk_observations()
|
||||
else:
|
||||
_create_risk_observations()
|
||||
|
||||
if not _table_exists("risk_observation_feedback"):
|
||||
_create_risk_observation_feedback()
|
||||
|
||||
if _table_exists("few_shot_samples"):
|
||||
_adopt_few_shot_samples()
|
||||
else:
|
||||
_create_few_shot_samples()
|
||||
|
||||
_ensure_risk_indexes()
|
||||
_ensure_few_shot_indexes()
|
||||
|
||||
|
||||
def downgrade() -> None:
|
||||
_require_postgresql()
|
||||
# 0008 之前这三张表由旧 bootstrap 管理。降级必须恢复旧结构而不是删除
|
||||
# 已确认的风险反馈与 few-shot 数据；若存在非默认租户数据则无法无损合并，
|
||||
# 直接失败并保持当前版本，禁止用数据丢失换取“成功降级”。
|
||||
few_shot_exists = _table_exists("few_shot_samples")
|
||||
risk_observations_exists = _table_exists("risk_observations")
|
||||
# 所有数据安全检查必须先于 DDL；任一检查失败时完整保留 0008 结构。
|
||||
if few_shot_exists:
|
||||
_require_lossless_default_tenant_downgrade("few_shot_samples")
|
||||
_require_lossless_few_shot_downgrade()
|
||||
if risk_observations_exists:
|
||||
_require_lossless_default_tenant_downgrade("risk_observations")
|
||||
|
||||
if few_shot_exists:
|
||||
for index_name in (
|
||||
"ix_few_shot_samples_tenant_rule_lookup",
|
||||
"ix_few_shot_samples_tenant_id",
|
||||
"ix_few_shot_samples_policy_ref",
|
||||
"ix_few_shot_samples_rule_version",
|
||||
):
|
||||
_drop_index_if_exists("few_shot_samples", index_name)
|
||||
_drop_unique_if_exists(
|
||||
"few_shot_samples",
|
||||
"uq_few_shot_samples_tenant_key",
|
||||
)
|
||||
op.create_unique_constraint(
|
||||
"uq_few_shot_samples_sample_key",
|
||||
"few_shot_samples",
|
||||
["sample_key"],
|
||||
)
|
||||
op.drop_column("few_shot_samples", "rule_version")
|
||||
op.drop_column("few_shot_samples", "policy_ref")
|
||||
op.drop_column("few_shot_samples", "tenant_id")
|
||||
|
||||
if risk_observations_exists:
|
||||
for index_name in (
|
||||
"ix_risk_observations_tenant_status",
|
||||
"ix_risk_observations_tenant_id",
|
||||
):
|
||||
_drop_index_if_exists("risk_observations", index_name)
|
||||
_drop_unique_if_exists(
|
||||
"risk_observations",
|
||||
"uq_risk_observations_tenant_key",
|
||||
)
|
||||
op.create_unique_constraint(
|
||||
"uq_risk_observations_observation_key",
|
||||
"risk_observations",
|
||||
["observation_key"],
|
||||
)
|
||||
op.drop_column("risk_observations", "tenant_id")
|
||||
@@ -0,0 +1,127 @@
|
||||
"""serialize organization memory mutations and persist idempotency fingerprints
|
||||
|
||||
Revision ID: 20260716_0009
|
||||
Revises: 20260716_0008
|
||||
Create Date: 2026-07-16 14:20:00
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from collections.abc import Sequence
|
||||
|
||||
import sqlalchemy as sa
|
||||
|
||||
from alembic import op
|
||||
|
||||
revision: str = "20260716_0009"
|
||||
down_revision: str | None = "20260716_0008"
|
||||
branch_labels: str | Sequence[str] | None = None
|
||||
depends_on: str | Sequence[str] | None = None
|
||||
|
||||
|
||||
def _require_postgresql() -> None:
|
||||
dialect_name = op.get_bind().dialect.name
|
||||
if dialect_name != "postgresql":
|
||||
raise RuntimeError(
|
||||
"20260716_0009 only supports PostgreSQL; "
|
||||
f"refusing to mutate {dialect_name} without transactional constraint DDL"
|
||||
)
|
||||
|
||||
|
||||
def _require_unique_active_organization_scopes() -> None:
|
||||
duplicate_count = int(
|
||||
op.get_bind().scalar(
|
||||
sa.text(
|
||||
"SELECT COUNT(*) FROM ("
|
||||
"SELECT 1 FROM memory_entries "
|
||||
"WHERE status = 'active' "
|
||||
"AND scope_type IN ('department', 'enterprise') "
|
||||
"GROUP BY tenant_id, scope_type, scope_id, scene, field_key "
|
||||
"HAVING COUNT(*) > 1"
|
||||
") AS duplicate_active_organization_scopes"
|
||||
)
|
||||
)
|
||||
or 0
|
||||
)
|
||||
if duplicate_count:
|
||||
raise RuntimeError(
|
||||
"cannot upgrade organization memory idempotency: "
|
||||
f"found {duplicate_count} organization scope(s) with duplicate active memories"
|
||||
)
|
||||
|
||||
|
||||
def upgrade() -> None:
|
||||
_require_postgresql()
|
||||
# 必须在添加列或约束前完成数据预检，避免失败后留下半迁移结构。
|
||||
_require_unique_active_organization_scopes()
|
||||
op.add_column(
|
||||
"memory_entries",
|
||||
sa.Column("management_request_id", sa.String(length=120), nullable=True),
|
||||
)
|
||||
op.add_column(
|
||||
"memory_entries",
|
||||
sa.Column("management_payload_fingerprint", sa.String(length=80), nullable=True),
|
||||
)
|
||||
op.add_column(
|
||||
"memory_entries",
|
||||
sa.Column("revoke_request_id", sa.String(length=120), nullable=True),
|
||||
)
|
||||
op.add_column(
|
||||
"memory_entries",
|
||||
sa.Column("revoke_payload_fingerprint", sa.String(length=80), nullable=True),
|
||||
)
|
||||
op.create_check_constraint(
|
||||
"ck_memory_entries_management_idempotency_pair",
|
||||
"memory_entries",
|
||||
"(management_request_id IS NULL AND management_payload_fingerprint IS NULL) OR "
|
||||
"(management_request_id IS NOT NULL "
|
||||
"AND management_payload_fingerprint IS NOT NULL)",
|
||||
)
|
||||
op.create_check_constraint(
|
||||
"ck_memory_entries_revoke_idempotency_pair",
|
||||
"memory_entries",
|
||||
"(revoke_request_id IS NULL AND revoke_payload_fingerprint IS NULL) OR "
|
||||
"(revoke_request_id IS NOT NULL AND revoke_payload_fingerprint IS NOT NULL)",
|
||||
)
|
||||
op.create_index(
|
||||
"uq_memory_entries_management_request",
|
||||
"memory_entries",
|
||||
["tenant_id", "management_request_id"],
|
||||
unique=True,
|
||||
)
|
||||
op.create_index(
|
||||
"uq_memory_entries_revoke_request",
|
||||
"memory_entries",
|
||||
["tenant_id", "revoke_request_id"],
|
||||
unique=True,
|
||||
)
|
||||
op.create_index(
|
||||
"uq_memory_entries_active_scope",
|
||||
"memory_entries",
|
||||
["tenant_id", "scope_type", "scope_id", "scene", "field_key"],
|
||||
unique=True,
|
||||
postgresql_where=sa.text(
|
||||
"status = 'active' AND scope_type IN ('department', 'enterprise')"
|
||||
),
|
||||
)
|
||||
|
||||
|
||||
def downgrade() -> None:
|
||||
_require_postgresql()
|
||||
op.drop_index("uq_memory_entries_active_scope", table_name="memory_entries")
|
||||
op.drop_index("uq_memory_entries_revoke_request", table_name="memory_entries")
|
||||
op.drop_index("uq_memory_entries_management_request", table_name="memory_entries")
|
||||
op.drop_constraint(
|
||||
"ck_memory_entries_revoke_idempotency_pair",
|
||||
"memory_entries",
|
||||
type_="check",
|
||||
)
|
||||
op.drop_constraint(
|
||||
"ck_memory_entries_management_idempotency_pair",
|
||||
"memory_entries",
|
||||
type_="check",
|
||||
)
|
||||
op.drop_column("memory_entries", "revoke_payload_fingerprint")
|
||||
op.drop_column("memory_entries", "revoke_request_id")
|
||||
op.drop_column("memory_entries", "management_payload_fingerprint")
|
||||
op.drop_column("memory_entries", "management_request_id")
|
||||
Reference in New Issue
Block a user