chore: import existing codebase as baseline

scripts/inspect_source.py

@@ -0,0 +1,49 @@
+"""Read-only source reconciliation, without patient-level output."""
+
+import json
+import sys
+from collections import Counter, defaultdict
+from decimal import Decimal
+
+from openpyxl import load_workbook
+
+wb = load_workbook(sys.argv[1], read_only=True, data_only=True)
+rates = {
+    str(row[3]): Decimal(str(row[4]))
+    for row in wb["产品毛利率"].iter_rows(min_row=6, values_only=True)
+    if row[3] and row[4] is not None
+}
+totals = defaultdict(lambda: {"rows": 0, "revenue": Decimal(0), "gross": Decimal(0), "customers": set()})
+precision = Counter()
+quality = Counter()
+for row in wb["明细"].iter_rows(min_row=3, values_only=True):
+    period = f"{int(row[0])}-{int(str(row[1]).removesuffix('月')):02d}"
+    v = Decimal(str(row[13]))
+    t = totals[period]
+    t["rows"] += 1
+    t["revenue"] += v
+    t["gross"] += v * rates.get(row[14], Decimal(0))
+    t["customers"].add(row[3])
+    for i in [5, 11, 12, 13]:
+        if row[i] is not None:
+            n = Decimal(str(row[i]))
+            if n != n.quantize(Decimal("0.0001")):
+                precision[i] += 1
+    quality["negative_revenue"] += v < 0
+    quality["g_l_mismatch"] += row[6] != row[11]
+    quality["i_n_mismatch"] += row[8] != row[13]
+    quality["missing_margin"] += row[14] not in rates
+print(
+    json.dumps(
+        {
+            "totals": {p: {**v, "customers": len(v["customers"])} for p, v in totals.items()},
+            "precision_exceeds_four": dict(precision),
+            "quality": dict(quality),
+            "rates": rates,
+        },
+        default=str,
+        ensure_ascii=False,
+        indent=2,
+    )
+)
+wb.close()