ben/scrape-giant
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Finalize post-refactor layout and remove old pipeline files

Browse Source
This commit is contained in:
ben
2026-03-24 17:09:57 -04:00
parent cdb7a15739
commit 09829b2b9d
17 changed files with 59 additions and 1154 deletions
Download Patch File Download Diff File Expand all files Collapse all files

71
tests/test_costco_pipeline.py
View File

@@ -7,7 +7,6 @@ from unittest import mock
import enrich_costco
import scrape_costco
import validate_cross_retailer_flow
class CostcoPipelineTests(unittest.TestCase):
@@ -423,76 +422,6 @@ class CostcoPipelineTests(unittest.TestCase):
self.assertIn("matched_discount=4873222", purchase_row["parse_notes"])
self.assertIn("matched_to_item=4873222", discount_row["parse_notes"])
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"])
def test_main_writes_summary_request_metadata(self):
with tempfile.TemporaryDirectory() as tmpdir:
outdir = Path(tmpdir) / "costco_output"