Update DSP data through August 30

update.sh

@@ -30,7 +30,9 @@ import csv
 import datetime as dt
 import io
 import json
+import os
 import pathlib
+import re
 import sys
 
 source_path = pathlib.Path(sys.argv[1])
@@ -72,7 +74,9 @@ def number(value):
 campaign_daily_map = {}
 for row in reader:
     pkg = pick(row, "维度1", "pkg", "Pkg").strip()
-    adping_id = pick(row, "维度2", "adping_id", "adpingid").strip()
+    raw_adping_id = pick(row, "维度2", "adping_id", "adpingid").strip()
+    match = re.match(r"^(\d+)", raw_adping_id)
+    adping_id = match.group(1) if match else raw_adping_id
     campaign = campaign_by_key.get((pkg, adping_id))
     if not campaign:
         continue
@@ -105,6 +109,11 @@ def metrics(item):
     return item
 
 campaign_daily = [metrics(v) for _, v in sorted(campaign_daily_map.items())]
+if os.environ.get("MERGE_EXISTING") == "1" and output_path.exists():
+    existing = json.loads(output_path.read_text(encoding="utf-8"))
+    merged = {(r["date"], r["pkg"], r["adping_id"]): r for r in existing.get("campaign_daily", [])}
+    merged.update({(r["date"], r["pkg"], r["adping_id"]): r for r in campaign_daily})
+    campaign_daily = [r for _, r in sorted(merged.items())]
 daily_map = {}
 for row in campaign_daily:
     key = (row["date"], row["pkg"])