Files
ss-tools/backend/src/core/mapping_service.py
busya 8c10632494 feat(semantic): curator-driven protocol hardening — decision memory + relation repair
- Add @RATIONALE/@REJECTED to 103+ C4/C5 contracts across backend core, services, API routes, and frontend models
- Fix 109 unresolved @RELATION edges (Auth.*, SupersetClient.*, AgentChat.*, ADR cross-refs)
- Add 13 @ingroup tags for DSA/HCA attention grouping
- Repair 29 stale graph edges via index rebuild
- Update .kilo agent prompts and skills for GRACE-Poly v2.6 compliance
- Git integration: merge routes, branch lifecycle, remote providers, UX components
- 0 broken anchor pairs, index rebuilt with 0 parse warnings
2026-07-02 08:53:19 +03:00

252 lines
12 KiB
Python

# #region IdMappingServiceModule [C:5] [TYPE Module] [SEMANTICS sqlalchemy, mapping, sync, superset, resource]
# @defgroup Core Module group.
#
# @BRIEF Service for tracking and synchronizing Superset Resource IDs (UUID <-> Integer ID)
# @LAYER Core
# @RELATION DEPENDS_ON -> [MappingModels]
# @RELATION DEPENDS_ON -> [LoggerModule]
# @RELATION DEPENDS_ON -> [AsyncJobRunner]
# @PRE Database session is valid and Superset client factory returns authenticated clients for requested environments.
# @POST Mapping synchronization and lookup APIs are available for environment-scoped UUID-to-integer resolution.
# @SIDE_EFFECT Reads/writes ResourceMapping rows, emits logs.
# @DATA_CONTRACT Input[environment_id, resource_type, uuid] -> Output[remote_integer_id|None]
# @TEST_DATA: mock_superset_resources -> {'chart': [{'id': 42, 'uuid': '1234', 'slice_name': 'test'}], 'dataset': [{'id': 99, 'uuid': '5678', 'table_name': 'data'}]}
#
# @INVARIANT sync_environment must handle remote API failures gracefully.
# @RATIONALE Centralizes UUID-to-integer ID resolution for Superset resources because the Superset API uses different ID schemes across endpoints (UUIDs for import/export, integer IDs for CRUD operations), enabling cross-environment migration with consistent resource references.
# @REJECTED BackgroundScheduler — was never started; replaced by AsyncJobRunner for async/sync bridge.
from datetime import UTC, datetime
from sqlalchemy.orm import Session
from src.core.cot_logger import seed_trace_id
from src.core.logger import belief_scope, logger
from src.models.mapping import ResourceMapping, ResourceType
# #region IdMappingService [C:5] [TYPE Class]
# @defgroup Core Module group.
# @BRIEF Service handling the cataloging and retrieval of remote Superset Integer IDs.
# @PRE db_session is an active SQLAlchemy Session bound to mapping tables.
# @POST Service instance provides environment-scoped mapping synchronization APIs.
# @RELATION DEPENDS_ON -> [MappingModels]
# @RELATION DEPENDS_ON -> [LoggerModule]
# @RELATION DEPENDS_ON -> [AsyncJobRunner]
# @INVARIANT self.db remains the authoritative session for all mapping operations.
# @SIDE_EFFECT Performs database writes during sync cycles.
# @DATA_CONTRACT Input[db_session] -> Output[IdMappingService]
#
# @TEST_CONTRACT IdMappingServiceModel ->
# {
# required_fields: {db_session: Session},
# invariants: [
# "sync_environment correctly creates or updates ResourceMapping records",
# "get_remote_id returns an integer or None",
# "get_remote_ids_batch returns a dictionary of valid UUIDs to integers"
# ]
# }
# @TEST_FIXTURE valid_mapping_service -> {"db_session": "MockSession()"}
# @TEST_EDGE sync_api_failure -> handles exception gracefully
# @TEST_EDGE get_remote_id_not_found -> returns None
# @TEST_EDGE get_batch_empty_list -> returns empty dict
# @TEST_INVARIANT resilient_fetching -> verifies: [sync_api_failure]
class IdMappingService:
# #region __init__ [TYPE Function]
# @PURPOSE: Initializes the mapping service.
def __init__(self, db_session: Session):
self.db = db_session
# #endregion __init__
# #region sync_environment [TYPE Function]
# @ingroup Core
# @PURPOSE: Fully synchronizes mapping for a specific environment.
# @PARAM environment_id (str) - Target environment ID.
# @PARAM superset_client - Instance capable of hitting the Superset API.
# @PRE environment_id exists in the database.
# @POST ResourceMapping records for the environment are created or updated.
async def sync_environment(
self, environment_id: str, superset_client, incremental: bool = False
) -> None:
"""
Polls the Superset APIs for the target environment and updates the local mapping table.
If incremental=True, only fetches items changed since the max last_synced_at date.
"""
seed_trace_id()
with belief_scope("IdMappingService.sync_environment"):
logger.reason(
f"Starting sync for environment {environment_id}",
extra={"src": "IdMappingService.sync_environment", "payload": {"environment_id": environment_id, "incremental": incremental}},
)
# Implementation Note: In a real scenario, superset_client needs to be an instance
# capable of auth & iteration over /api/v1/chart/, /api/v1/dataset/, /api/v1/dashboard/
# Here we structure the logic according to the spec.
types_to_poll = [
(ResourceType.CHART, "chart", "slice_name"),
(ResourceType.DATASET, "dataset", "table_name"),
(
ResourceType.DASHBOARD,
"dashboard",
"slug",
), # Note: dashboard slug or dashboard_title
]
total_synced = 0
total_deleted = 0
try:
for res_enum, endpoint, name_field in types_to_poll:
logger.debug(
"Polling %s endpoint", endpoint,
)
# Simulated API Fetch (Would be: superset_client.get(f"/api/v1/{endpoint}/")... )
# This relies on the superset API structure, e.g. { "result": [{"id": 1, "uuid": "...", name_field: "..."}] }
# We assume superset_client provides a generic method to fetch all pages.
try:
since_dttm = None
if incremental:
from sqlalchemy.sql import func
max_date = (
self.db.query(func.max(ResourceMapping.last_synced_at))
.filter(
ResourceMapping.environment_id == environment_id,
ResourceMapping.resource_type == res_enum,
)
.scalar()
)
if max_date:
# We subtract a bit for safety overlap
from datetime import timedelta
since_dttm = max_date - timedelta(minutes=5)
logger.debug(
"Incremental sync since %s", since_dttm,
)
resources = await superset_client.get_all_resources(
endpoint, since_dttm=since_dttm
)
# Track which UUIDs we see in this sync cycle
synced_uuids = set()
for res in resources:
res_uuid = res.get("uuid")
raw_id = res.get("id")
res_name = res.get(name_field)
if not res_uuid or raw_id is None:
continue
synced_uuids.add(res_uuid)
res_id = str(raw_id) # Store as string
# Upsert Logic
mapping = (
self.db.query(ResourceMapping)
.filter_by(
environment_id=environment_id,
resource_type=res_enum,
uuid=res_uuid,
)
.first()
)
if mapping:
mapping.remote_integer_id = res_id
mapping.resource_name = res_name
mapping.last_synced_at = datetime.now(UTC)
else:
new_mapping = ResourceMapping(
environment_id=environment_id,
resource_type=res_enum,
uuid=res_uuid,
remote_integer_id=res_id,
resource_name=res_name,
last_synced_at=datetime.now(UTC),
)
self.db.add(new_mapping)
total_synced += 1
# Delete stale mappings: rows for this env+type whose UUID
# was NOT returned by the API (resource was deleted remotely)
# We only do this on full syncs, because incremental syncs don't return all UUIDs
if not incremental:
stale_query = self.db.query(ResourceMapping).filter(
ResourceMapping.environment_id == environment_id,
ResourceMapping.resource_type == res_enum,
)
if synced_uuids:
stale_query = stale_query.filter(
ResourceMapping.uuid.notin_(synced_uuids)
)
deleted = stale_query.delete(synchronize_session="fetch")
if deleted:
total_deleted += deleted
logger.reason(
"Removed stale mappings",
payload={"deleted": deleted, "endpoint": endpoint, "environment_id": environment_id},
)
except Exception as loop_e:
logger.explore(
"Error polling endpoint",
error=str(loop_e),
payload={"endpoint": endpoint},
)
# Continue to next resource type instead of blowing up the whole sync
self.db.commit()
logger.reflect(
"Successfully synced environment",
payload={"synced": total_synced, "deleted_stale": total_deleted},
)
except Exception as e:
self.db.rollback()
logger.explore(
"Critical sync failure",
error=str(e),
)
raise
# #endregion sync_environment
# #region get_remote_id [TYPE Function]
# @ingroup Core
# @PURPOSE: Retrieves the remote integer ID for a given universal UUID.
# @PARAM environment_id (str)
# @PARAM resource_type (ResourceType)
# @PARAM uuid (str)
# @RETURN Optional[int]
def get_remote_id(
self, environment_id: str, resource_type: ResourceType, uuid: str
) -> int | None:
mapping = (
self.db.query(ResourceMapping)
.filter_by(
environment_id=environment_id, resource_type=resource_type, uuid=uuid
)
.first()
)
if mapping:
try:
return int(mapping.remote_integer_id)
except ValueError:
return None
return None
# #endregion get_remote_id
# #region get_remote_ids_batch [TYPE Function]
# @ingroup Core
# @PURPOSE: Retrieves remote integer IDs for a list of universal UUIDs efficiently.
# @PARAM environment_id (str)
# @PARAM resource_type (ResourceType)
# @PARAM uuids (List[str])
# @RETURN Dict[str, int] - Mapping of UUID -> Integer ID
def get_remote_ids_batch(
self, environment_id: str, resource_type: ResourceType, uuids: list[str]
) -> dict[str, int]:
if not uuids:
return {}
mappings = (
self.db.query(ResourceMapping)
.filter(
ResourceMapping.environment_id == environment_id,
ResourceMapping.resource_type == resource_type,
ResourceMapping.uuid.in_(uuids),
)
.all()
)
result = {}
for m in mappings:
try:
result[m.uuid] = int(m.remote_integer_id)
except ValueError:
logger.debug("Could not parse remote_integer_id for mapping %s (uuid=%s)", m.id, m.uuid)
return result
# #endregion get_remote_ids_batch
# #endregion IdMappingService
# #endregion IdMappingServiceModule