init: Prefect ETL sync project (RIIN/SDS → PG) with 12 sync flows
This commit is contained in:
@@ -0,0 +1 @@
|
||||
"""RIIN 同步 flows:order_sync / incremental_sync / recent_sync / refresh_pending / by_orders / verify。"""
|
||||
@@ -0,0 +1,45 @@
|
||||
"""RIIN 同步公共管道:映射订单 → 并发拉商品 → upsert 到 PG。"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from typing import Any
|
||||
|
||||
from flows.shared import fetch_goods_concurrent
|
||||
from flows.shared.db import SyncDB
|
||||
from flows.shared.mappers import map_riin_goods, map_riin_order
|
||||
from api import RiinDisClient
|
||||
|
||||
|
||||
def sync_riin_records(raw_orders: list[Any], client: RiinDisClient, db: SyncDB) -> dict[str, int]:
|
||||
order_inputs: list[Any] = []
|
||||
order_meta: dict[str, tuple[str, Any]] = {} # raw.id -> (orderNo, createdAt)
|
||||
goods_failed = 0
|
||||
|
||||
for raw in raw_orders:
|
||||
try:
|
||||
o = map_riin_order(raw)
|
||||
except Exception as err: # noqa: BLE001
|
||||
print(f"[RIIN] 映射订单失败 {getattr(raw, 'platform_oid', '?')}: {err}")
|
||||
continue
|
||||
order_inputs.append(o)
|
||||
if raw.id:
|
||||
order_meta[raw.id] = (o.orderNo, o.createdAt)
|
||||
|
||||
goods_inputs: list[Any] = []
|
||||
if order_meta:
|
||||
goods_map = fetch_goods_concurrent(client, list(order_meta.keys()))
|
||||
for oid, (order_no, created_at) in order_meta.items():
|
||||
items = goods_map.get(oid, [])
|
||||
if not items:
|
||||
goods_failed += 1
|
||||
for g in items:
|
||||
try:
|
||||
goods_inputs.append(map_riin_goods(g, order_no, created_at))
|
||||
except Exception as err: # noqa: BLE001
|
||||
print(f"[RIIN] 映射商品失败 order={order_no}: {err}")
|
||||
|
||||
stats = db.sync_orders_and_goods(order_inputs, goods_inputs)
|
||||
stats["orderCount"] = len(order_inputs)
|
||||
stats["goodsCount"] = len(goods_inputs)
|
||||
stats["goodsFetchFailed"] = goods_failed
|
||||
return stats
|
||||
@@ -0,0 +1,59 @@
|
||||
"""RIIN 指定单号同步:通过订单号列表(逗号/换行分隔)拉取指定订单。
|
||||
|
||||
对应 riin-by-orders。
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import re
|
||||
|
||||
from prefect import flow, task
|
||||
|
||||
from api.riin.models import OrderPageRequest
|
||||
from flows.riin._pipeline import sync_riin_records
|
||||
from flows.shared import SyncDB, create_riin_client
|
||||
|
||||
|
||||
def _split_order_nos(text: str) -> list[str]:
|
||||
seen: dict[str, None] = {}
|
||||
for part in re.split(r"[\n\r,,\s]+", text or ""):
|
||||
part = part.strip()
|
||||
if part and part not in seen:
|
||||
seen[part] = None
|
||||
return list(seen.keys())
|
||||
|
||||
|
||||
@task(retries=3, retry_delay_seconds=[3, 6, 9])
|
||||
def fetch_by_orders(order_nos: list[str]) -> list:
|
||||
client = create_riin_client()
|
||||
try:
|
||||
resp = client.get_orders(OrderPageRequest(platform_oid_list=",".join(order_nos), all=True))
|
||||
return list(resp.data.records)
|
||||
finally:
|
||||
client.close()
|
||||
|
||||
|
||||
@task
|
||||
def upsert_records(raw_orders: list) -> dict:
|
||||
client = create_riin_client()
|
||||
try:
|
||||
return sync_riin_records(raw_orders, client, SyncDB())
|
||||
finally:
|
||||
client.close()
|
||||
|
||||
|
||||
@flow(name="riin-by-orders-sync", log_prints=True)
|
||||
def riin_by_orders_sync(order_nos: str) -> dict:
|
||||
nos = _split_order_nos(order_nos)
|
||||
print(f"RIIN 指定单号同步 输入 {len(nos)} 个")
|
||||
if not nos:
|
||||
return {"inputCount": 0, "matchedCount": 0}
|
||||
|
||||
raw = fetch_by_orders(nos)
|
||||
stats = upsert_records(raw)
|
||||
print(f"完成: 匹配 {len(raw)} | {stats}")
|
||||
return {"inputCount": len(nos), "matchedCount": len(raw), **stats}
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
riin_by_orders_sync.serve(name="riin-by-orders-sync")
|
||||
@@ -0,0 +1,60 @@
|
||||
"""RIIN 增量同步:从数据库最近下单时间增量拉取(无参,全自动)。
|
||||
|
||||
对应 riin-incremental-sync。from = floorToHour(MAX(orderedAt)),to = now。
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from datetime import datetime
|
||||
|
||||
from prefect import flow, task
|
||||
|
||||
from flows.riin._pipeline import sync_riin_records
|
||||
from flows.shared import (
|
||||
SyncDB, create_riin_client, fetch_riin_all_pages, floor_to_hour, format_date,
|
||||
)
|
||||
|
||||
|
||||
@task(retries=3, retry_delay_seconds=[3, 6, 9])
|
||||
def extract_orders(params: dict) -> list:
|
||||
client = create_riin_client()
|
||||
try:
|
||||
return fetch_riin_all_pages(client, params)
|
||||
finally:
|
||||
client.close()
|
||||
|
||||
|
||||
@task
|
||||
def upsert_records(raw_orders: list) -> dict:
|
||||
client = create_riin_client()
|
||||
try:
|
||||
return sync_riin_records(raw_orders, client, SyncDB())
|
||||
finally:
|
||||
client.close()
|
||||
|
||||
|
||||
@flow(name="riin-incremental-sync", log_prints=True)
|
||||
def riin_incremental_sync() -> dict:
|
||||
db = SyncDB()
|
||||
latest = db.get_latest_ordered_at("RIIN")
|
||||
now = datetime.now()
|
||||
from_dt = floor_to_hour(latest) if latest else floor_to_hour(now)
|
||||
to_dt = now
|
||||
if from_dt >= to_dt:
|
||||
from_dt = floor_to_hour(now)
|
||||
|
||||
params = {
|
||||
"time_type": 3,
|
||||
"start_time": format_date(from_dt),
|
||||
"end_time": format_date(to_dt),
|
||||
"page_size": 200,
|
||||
}
|
||||
print(f"RIIN 增量同步 {params['start_time']} ~ {params['end_time']}")
|
||||
raw = extract_orders(params)
|
||||
stats = upsert_records(raw)
|
||||
print(f"完成: 订单 {stats['orderCount']} / 商品 {stats['goodsCount']}")
|
||||
return {"syncRange": {"from": params["start_time"], "to": params["end_time"]}, **stats}
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
riin_incremental_sync.serve(name="riin-incremental-sync", cron="0 * * * *")
|
||||
@@ -0,0 +1,77 @@
|
||||
"""RIIN 订单同步:按时间窗口(+可选筛选)拉取订单与商品。
|
||||
|
||||
对应 riin-order-sync。timeType 固定 3(商家推送时间)。
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from typing import Any
|
||||
|
||||
from prefect import flow, task
|
||||
|
||||
from api.riin.models import OrderPageRequest
|
||||
from flows.riin._pipeline import sync_riin_records
|
||||
from flows.shared import SyncDB, create_riin_client, fetch_riin_all_pages, format_date
|
||||
|
||||
|
||||
@task(retries=3, retry_delay_seconds=[3, 6, 9])
|
||||
def extract_orders(params: dict[str, Any]) -> list[Any]:
|
||||
client = create_riin_client()
|
||||
try:
|
||||
return fetch_riin_all_pages(client, params)
|
||||
finally:
|
||||
client.close()
|
||||
|
||||
|
||||
@task
|
||||
def upsert_records(raw_orders: list[Any]) -> dict[str, int]:
|
||||
client = create_riin_client()
|
||||
try:
|
||||
return sync_riin_records(raw_orders, client, SyncDB())
|
||||
finally:
|
||||
client.close()
|
||||
|
||||
|
||||
@flow(name="riin-order-sync", log_prints=True)
|
||||
def riin_order_sync(
|
||||
order_date_begin: str,
|
||||
order_date_end: str,
|
||||
platform_type: int | None = None,
|
||||
order_status: int | None = None,
|
||||
factory_ids: list[str] | None = None,
|
||||
express_source: int | None = None,
|
||||
urgent_order: int | None = None,
|
||||
has_refund: bool | None = None,
|
||||
page_size: int = 200,
|
||||
) -> dict[str, Any]:
|
||||
if not order_date_begin or not order_date_end:
|
||||
raise ValueError("order_date_begin / order_date_end 为必填")
|
||||
|
||||
params: dict[str, Any] = {
|
||||
"time_type": 3,
|
||||
"start_time": order_date_begin,
|
||||
"end_time": order_date_end,
|
||||
"page_size": page_size,
|
||||
}
|
||||
if platform_type is not None:
|
||||
params["platform_type"] = platform_type
|
||||
if order_status is not None:
|
||||
params["order_status"] = order_status
|
||||
if factory_ids:
|
||||
params["factory_ids"] = factory_ids
|
||||
if express_source is not None:
|
||||
params["express_source"] = express_source
|
||||
if urgent_order is not None:
|
||||
params["urgent_order"] = urgent_order
|
||||
if has_refund is not None:
|
||||
params["has_refund"] = has_refund
|
||||
|
||||
print(f"RIIN 订单同步 {order_date_begin} ~ {order_date_end}")
|
||||
raw = extract_orders(params)
|
||||
stats = upsert_records(raw)
|
||||
print(f"完成: 订单 {stats['orderCount']} / 商品 {stats['goodsCount']} | {stats}")
|
||||
return {"syncRange": {"from": order_date_begin, "to": order_date_end}, **stats}
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
riin_order_sync.serve(name="riin-order-sync")
|
||||
@@ -0,0 +1,56 @@
|
||||
"""RIIN 近期同步:拉取最近 N 天订单(默认 7 天,1~90)。
|
||||
|
||||
对应 riin-recent-sync。from = startOfDay(now - N days),to = now。
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from datetime import datetime, timedelta
|
||||
|
||||
from prefect import flow, task
|
||||
|
||||
from flows.riin._pipeline import sync_riin_records
|
||||
from flows.shared import (
|
||||
SyncDB, create_riin_client, fetch_riin_all_pages, format_date, start_of_day,
|
||||
)
|
||||
|
||||
|
||||
@task(retries=3, retry_delay_seconds=[3, 6, 9])
|
||||
def extract_orders(params: dict) -> list:
|
||||
client = create_riin_client()
|
||||
try:
|
||||
return fetch_riin_all_pages(client, params)
|
||||
finally:
|
||||
client.close()
|
||||
|
||||
|
||||
@task
|
||||
def upsert_records(raw_orders: list) -> dict:
|
||||
client = create_riin_client()
|
||||
try:
|
||||
return sync_riin_records(raw_orders, client, SyncDB())
|
||||
finally:
|
||||
client.close()
|
||||
|
||||
|
||||
@flow(name="riin-recent-sync", log_prints=True)
|
||||
def riin_recent_sync(recent_days: int = 7) -> dict:
|
||||
days = min(max(recent_days, 1), 90)
|
||||
now = datetime.now()
|
||||
from_dt = start_of_day(now - timedelta(days=days))
|
||||
|
||||
params = {
|
||||
"time_type": 3,
|
||||
"start_time": format_date(from_dt),
|
||||
"end_time": format_date(now),
|
||||
"page_size": 200,
|
||||
}
|
||||
print(f"RIIN 近期同步 近 {days} 天 {params['start_time']} ~ {params['end_time']}")
|
||||
raw = extract_orders(params)
|
||||
stats = upsert_records(raw)
|
||||
print(f"完成: 订单 {stats['orderCount']} / 商品 {stats['goodsCount']}")
|
||||
return {"syncRange": {"from": params["start_time"], "to": params["end_time"]}, "recentDays": days, **stats}
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
riin_recent_sync.serve(name="riin-recent-sync", cron="30 * * * *")
|
||||
@@ -0,0 +1,58 @@
|
||||
"""RIIN 刷新未完结:查 DB 中未完结订单号,按 platformOidList 批量(500)刷新。
|
||||
|
||||
对应 riin-refresh-pending。终态:已退款 / 已关闭 / 已发货。
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from prefect import flow, task
|
||||
|
||||
from api.riin.models import OrderPageRequest
|
||||
from flows.riin._pipeline import sync_riin_records
|
||||
from flows.shared import SyncDB, create_riin_client
|
||||
|
||||
RIIN_FINISHED_STATUSES = ["已退款", "已关闭", "已发货"]
|
||||
ORDER_BATCH = 500
|
||||
|
||||
|
||||
@task(retries=3, retry_delay_seconds=[3, 6, 9])
|
||||
def fetch_pending(order_nos: list[str]) -> list:
|
||||
client = create_riin_client()
|
||||
try:
|
||||
all_raw: list = []
|
||||
for i in range(0, len(order_nos), ORDER_BATCH):
|
||||
batch = order_nos[i : i + ORDER_BATCH]
|
||||
resp = client.get_orders(
|
||||
OrderPageRequest(platform_oid_list=",".join(batch), all=True)
|
||||
)
|
||||
all_raw.extend(resp.data.records)
|
||||
return all_raw
|
||||
finally:
|
||||
client.close()
|
||||
|
||||
|
||||
@task
|
||||
def upsert_records(raw_orders: list) -> dict:
|
||||
client = create_riin_client()
|
||||
try:
|
||||
return sync_riin_records(raw_orders, client, SyncDB())
|
||||
finally:
|
||||
client.close()
|
||||
|
||||
|
||||
@flow(name="riin-refresh-pending", log_prints=True)
|
||||
def riin_refresh_pending() -> dict:
|
||||
db = SyncDB()
|
||||
order_nos = db.get_pending_orders("RIIN", RIIN_FINISHED_STATUSES)
|
||||
print(f"RIIN 未完结订单 {len(order_nos)} 个")
|
||||
if not order_nos:
|
||||
return {"pendingCount": 0, "matchedCount": 0}
|
||||
|
||||
raw = fetch_pending(order_nos)
|
||||
stats = upsert_records(raw)
|
||||
print(f"完成: 匹配 {len(raw)} | {stats}")
|
||||
return {"pendingCount": len(order_nos), "matchedCount": len(raw), **stats}
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
riin_refresh_pending.serve(name="riin-refresh-pending", cron="0 */3 * * *")
|
||||
@@ -0,0 +1,126 @@
|
||||
"""RIIN 数据核验:逐天比对远程/本地按状态计数,有差异则拉取修复。
|
||||
|
||||
对应 riin-verify。mode: recent(近 N 天,不含今天)/ range(指定区间)。
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from datetime import datetime, timedelta
|
||||
|
||||
from prefect import flow, task
|
||||
|
||||
from api.riin.models import OrderPageRequest
|
||||
from flows.riin._pipeline import sync_riin_records
|
||||
from flows.shared import (
|
||||
SyncDB, create_riin_client, fetch_riin_all_pages, format_date, start_of_day,
|
||||
)
|
||||
|
||||
RIIN_COUNT_STATUS_MAP = {
|
||||
"assignedSuceessed": "已分配",
|
||||
"factoryAuditing": "工厂审核中",
|
||||
"inProducing": "生产中",
|
||||
"shipped": "已发货",
|
||||
"closed": "已关闭",
|
||||
"refunding": "退款中",
|
||||
"refunded": "已退款",
|
||||
"rejecting": "拒审中",
|
||||
"rejectedToShop": "已退回店铺",
|
||||
}
|
||||
|
||||
|
||||
def _parse_days(mode: str, recent_days: int, date_range: str | None) -> list[datetime]:
|
||||
today = start_of_day(datetime.now())
|
||||
if mode == "range" and date_range:
|
||||
parts = [p.strip() for p in date_range.split("~")]
|
||||
from_dt = datetime.fromisoformat(parts[0])
|
||||
to_dt = datetime.fromisoformat(parts[1]) if len(parts) > 1 and parts[1] else from_dt
|
||||
days = []
|
||||
cur = from_dt
|
||||
while cur <= to_dt:
|
||||
days.append(cur)
|
||||
cur += timedelta(days=1)
|
||||
return days
|
||||
days_n = min(max(recent_days, 1), 90)
|
||||
return [today - timedelta(days=i) for i in range(days_n, 0, -1)]
|
||||
|
||||
|
||||
def _count_to_status_map(data: dict) -> dict[str, int]:
|
||||
out: dict[str, int] = {}
|
||||
for key, count in data.items():
|
||||
if key.startswith("dayRange"):
|
||||
continue
|
||||
label = RIIN_COUNT_STATUS_MAP.get(key)
|
||||
if label and count and count > 0:
|
||||
out[label] = count
|
||||
return out
|
||||
|
||||
|
||||
def _diff(remote: dict[str, int], local: dict[str, int]) -> tuple[dict[str, int], bool]:
|
||||
diff: dict[str, int] = {}
|
||||
has_diff = False
|
||||
for status in set(remote) | set(local):
|
||||
r = remote.get(status, 0)
|
||||
l = local.get(status, 0)
|
||||
if r != l:
|
||||
diff[status] = l - r
|
||||
has_diff = True
|
||||
return diff, has_diff
|
||||
|
||||
|
||||
@flow(name="riin-verify", log_prints=True)
|
||||
def riin_verify(mode: str = "recent", recent_days: int = 1, date_range: str | None = None) -> dict:
|
||||
days = _parse_days(mode, recent_days, date_range)
|
||||
db = SyncDB()
|
||||
client = create_riin_client()
|
||||
|
||||
totals = {k: 0 for k in ("ordersAdded", "ordersUpdated", "ordersRemoved",
|
||||
"goodsAdded", "goodsUpdated", "goodsRemoved")}
|
||||
days_with_diff = 0
|
||||
day_reports = []
|
||||
|
||||
try:
|
||||
for day in days:
|
||||
day_str = format_date(day).split(" ")[0]
|
||||
day_start = start_of_day(day)
|
||||
day_end = day_start + timedelta(days=1)
|
||||
|
||||
count_resp = client.get_order_count(
|
||||
OrderPageRequest(time_type=3, start_time=format_date(day_start), end_time=format_date(day_end))
|
||||
)
|
||||
remote_map = _count_to_status_map(count_resp.data.model_dump(by_alias=False))
|
||||
remote_total = sum(remote_map.values())
|
||||
local_counts = db.get_order_count_by_status("RIIN", day_start, day_end)
|
||||
local_goods = db.get_local_goods_count("RIIN", day_start, day_end)
|
||||
diff, has_diff = _diff(remote_map, local_counts)
|
||||
|
||||
print(f"[核验 {day_str}] 远程 {remote_total} 本地 {sum(local_counts.values())}"
|
||||
+ (f" 差异 {diff}" if has_diff else " 一致,跳过"))
|
||||
if not has_diff:
|
||||
day_reports.append({"date": day_str, "diff": diff, "ok": True})
|
||||
continue
|
||||
|
||||
days_with_diff += 1
|
||||
raw_orders = fetch_riin_all_pages(
|
||||
client,
|
||||
{"time_type": 3, "start_time": format_date(day_start), "end_time": format_date(day_end), "page_size": 200},
|
||||
)
|
||||
stats = sync_riin_records(raw_orders, client, db)
|
||||
for k in ("ordersCreated", "ordersUpdated", "goodsCreated", "goodsUpdated"):
|
||||
pass
|
||||
totals["ordersAdded"] += stats["ordersCreated"]
|
||||
totals["ordersUpdated"] += stats["ordersUpdated"]
|
||||
totals["goodsAdded"] += stats["goodsCreated"]
|
||||
totals["goodsUpdated"] += stats["goodsUpdated"]
|
||||
day_reports.append({"date": day_str, "diff": diff, "ok": False,
|
||||
"orders": stats["orderCount"], "goods": stats["goodsCount"]})
|
||||
finally:
|
||||
client.close()
|
||||
|
||||
return {
|
||||
"verifyReport": {"days": day_reports, "summary": {"totalDays": len(days), "daysWithDiff": days_with_diff, **totals}},
|
||||
"syncRange": {"from": format_date(start_of_day(days[0])) if days else "", "to": format_date(start_of_day(days[-1])) if days else ""},
|
||||
}
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
riin_verify.serve(name="riin-verify", cron="0 6 * * *")
|
||||
Reference in New Issue
Block a user