feat: daily auto-update script (fetch prev-day, incremental merge)

update.sh

@@ -1,7 +1,109 @@
 #!/bin/bash
-# update.sh - 从上传的 Excel 数据同步到 data.json
-# 服务器每天自动执行一次（由 schedule.conf 控制）
-# 数据来源：手动上传的 data.json（通过 git push 更新）
-# 如果后续有 API 数据源，可以在这里用 curl 拉取并写入 data.json
+# Moboost 报告 - 每日自动更新脚本
+# 服务器每天执行一次（schedule.conf=24）。
+# 逻辑：拉取"前一天"的 Mintegral 数据，增量合并进 mtg_data.json（去重，不覆盖历史）。
+# 工作目录 = 仓库根目录（同目录下有 mtg_data.json / campaigns.json）。
 
-echo "$(date '+%Y-%m-%d %H:%M:%S') - data.json is up to date (manual upload mode)"
+set -e
+cd "$(dirname "$0")"
+
+python3 - <<'PYEOF'
+import hashlib, time, json, os, urllib.request, urllib.parse, urllib.error
+from datetime import date, timedelta
+
+# ---- Mintegral 认证（access-key=平台"Success key"，token secret=平台"API key"）----
+API_KEY = "22793ef96b076fbdefa7c1266d4ebccb"
+SKEY    = "c47267b67f9e4231325efe4283044721"
+BASE    = "https://ss-api.mintegral.com/api/v2/reports/data"
+
+def headers():
+    ts = str(int(time.time()))
+    token = hashlib.md5((SKEY + hashlib.md5(ts.encode()).hexdigest()).encode()).hexdigest()
+    return {"access-key": API_KEY, "token": token, "timestamp": ts}
+
+def req(params):
+    url = BASE + "?" + urllib.parse.urlencode(params)
+    for attempt in range(6):
+        try:
+            r = urllib.request.Request(url, headers=headers())
+            with urllib.request.urlopen(r, timeout=90) as resp:
+                return resp.status, resp.headers.get("Content-Type",""), resp.read()
+        except urllib.error.HTTPError as e:
+            return e.code, e.headers.get("Content-Type",""), e.read()
+        except Exception as e:
+            time.sleep(5*(attempt+1))
+    raise RuntimeError("网络请求失败")
+
+# ---- 目标 campid（Mintegral 渠道），优先从 campaigns.json 读，读不到则用内置列表 ----
+FALLBACK_IDS = ["91530535","91530534","91530536","91530547","91530548",
+                "91530549","91530550","91530537","91530538","91551606",
+                "91534913","91533906"]
+try:
+    camps = json.load(open("campaigns.json"))
+    mtg = {c["campid"]: c for c in camps if c.get("channel")=="Mintegral"}
+except Exception:
+    mtg = {}
+if not mtg:
+    mtg = {cid: {"campid":cid,"product":"","pkg":""} for cid in FALLBACK_IDS}
+mtg_ids = list(mtg.keys())
+
+# ---- 目标日期：前一天（UTC+8）----
+target = (date.today() - timedelta(days=1)).isoformat()
+print(f"[update] 拉取 {target} 的 Mintegral 数据 ...")
+
+base = {"timezone":"+8","start_time":target,"end_time":target,
+        "dimension_option":"Offer,Location","time_granularity":"daily"}
+
+# type=1 生成
+for attempt in range(20):
+    p = dict(base); p["type"]="1"
+    s,ct,b = req(p)
+    try: code = json.loads(b.decode("utf-8","replace")).get("code")
+    except: code = None
+    if code == 200: break
+    time.sleep(6)
+else:
+    print("[update] 数据生成超时，退出"); raise SystemExit(0)
+
+# type=2 下载
+p = dict(base); p["type"]="2"
+s,ct,b = req(p)
+if "octet-stream" not in ct:
+    print("[update] 下载失败:", b.decode("utf-8","replace")[:200]); raise SystemExit(0)
+
+lines = b.decode("utf-8","replace").splitlines()
+header = lines[0].split("\t"); col = {n:i for i,n in enumerate(header)}
+new_rows = []
+for line in lines[1:]:
+    f = line.split("\t")
+    if len(f) < len(header): continue
+    offer = f[col["Offer Name"]]
+    matched = next((cid for cid in mtg_ids if cid in offer), None)
+    if not matched: continue
+    rec = mtg[matched]
+    d = f[col["Date"]]
+    new_rows.append({
+        "date": f"{d[:4]}-{d[4:6]}-{d[6:]}",
+        "channel":"Mintegral","campid":matched,"offer_name":offer,
+        "product":rec.get("product",""),"pkg":rec.get("pkg",""),
+        "location":f[col["Location"]],"bid_type":"",
+        "impression":int(f[col["Impression"]]),"clicks":int(f[col["Click"]]),
+        "conversions":int(f[col["Conversion"]]),
+        "ctr":float(f[col["Ctr"]]),"cvr":float(f[col["Cvr"]]),"spend":float(f[col["Spend"]]),
+    })
+print(f"[update] 新增匹配 {len(new_rows)} 条")
+
+# ---- 增量合并（按 date+campid+location 去重；新数据覆盖同键，历史保留）----
+existing = []
+if os.path.exists("mtg_data.json"):
+    try: existing = json.load(open("mtg_data.json"))
+    except: existing = []
+merged = {(r["date"],r["campid"],r["location"]): r for r in existing}
+for r in new_rows:
+    merged[(r["date"],r["campid"],r["location"])] = r
+out = sorted(merged.values(), key=lambda x:(x["date"], -x["impression"]))
+json.dump(out, open("mtg_data.json","w"), ensure_ascii=False, indent=2)
+print(f"[update] 合并后共 {len(out)} 条，已写入 mtg_data.json")
+PYEOF
+
+echo "[update] $(date '+%Y-%m-%d %H:%M:%S') 完成"