Add Costco acquisition and enrich flow

This commit is contained in:
ben
2026-03-16 09:17:46 -04:00
parent 9497565978
commit da00288f10
4 changed files with 1090 additions and 0 deletions

View File

@@ -0,0 +1,201 @@
import csv
import json
import tempfile
import unittest
from pathlib import Path
import enrich_costco
import scrape_costco
import validate_cross_retailer_flow
class CostcoPipelineTests(unittest.TestCase):
def test_flatten_costco_data_preserves_discount_rows(self):
summary_payload = {
"data": {
"receiptsWithCounts": {
"receipts": [
{
"transactionBarcode": "abc",
"tenderArray": [{"tenderDescription": "VISA"}],
"couponArray": [{"upcnumberCoupon": "2100003746641"}],
}
]
}
}
}
detail_payloads = [
{
"data": {
"receiptsWithCounts": {
"receipts": [
{
"transactionBarcode": "abc",
"transactionDate": "2026-03-12",
"receiptType": "In-Warehouse",
"total": 10.0,
"totalItemCount": 2,
"instantSavings": 5.0,
"warehouseName": "MT VERNON",
"warehouseNumber": 1115,
"warehouseAddress1": "7940 RICHMOND HWY",
"warehouseCity": "ALEXANDRIA",
"warehouseState": "VA",
"warehousePostalCode": "22306",
"itemArray": [
{
"itemNumber": "4873222",
"itemDescription01": "ALL F&C",
"itemDescription02": "200OZ 160LOADS P104",
"itemDepartmentNumber": 14,
"transDepartmentNumber": 14,
"unit": 1,
"itemIdentifier": "E",
"amount": 19.99,
"itemUnitPriceAmount": 19.99,
},
{
"itemNumber": "374664",
"itemDescription01": "/ 4873222",
"itemDescription02": None,
"itemDepartmentNumber": 14,
"transDepartmentNumber": 14,
"unit": -1,
"itemIdentifier": None,
"amount": -5,
"itemUnitPriceAmount": 0,
},
],
}
]
}
}
}
]
orders, items = scrape_costco.flatten_costco_data(
summary_payload, detail_payloads, Path("costco_output/raw")
)
self.assertEqual(1, len(orders))
self.assertEqual(2, len(items))
self.assertEqual("false", items[0]["is_discount_line"])
self.assertEqual("true", items[1]["is_discount_line"])
self.assertEqual("true", items[1]["is_coupon_line"])
def test_costco_enricher_parses_size_pack_and_discount(self):
row = enrich_costco.parse_costco_item(
order_id="abc",
order_date="2026-03-12",
raw_path=Path("costco_output/raw/abc.json"),
line_no=1,
item={
"itemNumber": "60357",
"itemDescription01": "MIXED PEPPER",
"itemDescription02": "6-PACK",
"itemDepartmentNumber": 65,
"transDepartmentNumber": 65,
"unit": 1,
"itemIdentifier": "E",
"amount": 7.49,
"itemUnitPriceAmount": 7.49,
},
)
self.assertEqual("60357", row["retailer_item_id"])
self.assertEqual("MIXED PEPPER", row["item_name_norm"])
self.assertEqual("6", row["pack_qty"])
self.assertEqual("count", row["measure_type"])
discount = enrich_costco.parse_costco_item(
order_id="abc",
order_date="2026-03-12",
raw_path=Path("costco_output/raw/abc.json"),
line_no=2,
item={
"itemNumber": "374664",
"itemDescription01": "/ 4873222",
"itemDescription02": None,
"itemDepartmentNumber": 14,
"transDepartmentNumber": 14,
"unit": -1,
"itemIdentifier": None,
"amount": -5,
"itemUnitPriceAmount": 0,
},
)
self.assertEqual("true", discount["is_discount_line"])
self.assertEqual("true", discount["is_coupon_line"])
def test_cross_retailer_validation_writes_proof_example(self):
with tempfile.TemporaryDirectory() as tmpdir:
giant_csv = Path(tmpdir) / "giant_items_enriched.csv"
costco_csv = Path(tmpdir) / "costco_items_enriched.csv"
outdir = Path(tmpdir) / "combined"
fieldnames = enrich_costco.OUTPUT_FIELDS
giant_row = {field: "" for field in fieldnames}
giant_row.update(
{
"retailer": "giant",
"order_id": "g1",
"line_no": "1",
"order_date": "2026-03-01",
"retailer_item_id": "100",
"item_name": "FRESH BANANA",
"item_name_norm": "BANANA",
"upc": "4011",
"measure_type": "weight",
"is_store_brand": "false",
"is_fee": "false",
"is_discount_line": "false",
"is_coupon_line": "false",
"line_total": "1.29",
}
)
costco_row = {field: "" for field in fieldnames}
costco_row.update(
{
"retailer": "costco",
"order_id": "c1",
"line_no": "1",
"order_date": "2026-03-12",
"retailer_item_id": "30669",
"item_name": "BANANAS 3 LB / 1.36 KG",
"item_name_norm": "BANANA",
"upc": "",
"size_value": "3",
"size_unit": "lb",
"measure_type": "weight",
"is_store_brand": "false",
"is_fee": "false",
"is_discount_line": "false",
"is_coupon_line": "false",
"line_total": "2.98",
}
)
with giant_csv.open("w", newline="", encoding="utf-8") as handle:
writer = csv.DictWriter(handle, fieldnames=fieldnames)
writer.writeheader()
writer.writerow(giant_row)
with costco_csv.open("w", newline="", encoding="utf-8") as handle:
writer = csv.DictWriter(handle, fieldnames=fieldnames)
writer.writeheader()
writer.writerow(costco_row)
validate_cross_retailer_flow.main.callback(
giant_items_enriched_csv=str(giant_csv),
costco_items_enriched_csv=str(costco_csv),
outdir=str(outdir),
)
proof_path = outdir / "proof_examples.csv"
self.assertTrue(proof_path.exists())
with proof_path.open(newline="", encoding="utf-8") as handle:
rows = list(csv.DictReader(handle))
self.assertEqual(1, len(rows))
self.assertEqual("banana", rows[0]["proof_name"])
if __name__ == "__main__":
unittest.main()