#!/bin/sh
set -eu

SCRIPT_DIR=$(CDPATH= cd -- "$(dirname -- "$0")" && pwd)
TMP_FILE="$SCRIPT_DIR/data.json.tmp"
OUTPUT_FILE="$SCRIPT_DIR/data.json"

python3 - "$TMP_FILE" <<'PY'
import csv
import json
import re
import sys
import urllib.request
from collections import defaultdict
from datetime import datetime, timedelta
from zoneinfo import ZoneInfo

CSV_URL = (
    "https://docs.google.com/spreadsheets/d/"
    "1KaNOwaZ3DqjTXRK8XhF1OGF9qUmOh0NLXOgxVrh14h0/"
    "export?format=csv&gid=1445753035"
)
TIMEZONE = ZoneInfo("Asia/Shanghai")
DAYS = 14
BLOCKS = [
    {"country": "KZ", "date": 2, "geo": 3, "installs": 5, "spend": 8, "cac": 10},
    {"country": "UZ", "date": 14, "geo": 15, "installs": 17, "spend": 18, "cac": 20},
    {"country": "KG", "date": 25, "geo": 26, "installs": 28, "spend": 29, "cac": 31},
    {"country": "AM", "date": 35, "geo": 36, "installs": 38, "spend": 39, "cac": 41},
]


def parse_date(value):
    text = str(value or "").strip()
    for fmt in ("%Y/%m/%d", "%Y-%m-%d", "%Y%m%d", "%m/%d/%Y"):
        try:
            return datetime.strptime(text, fmt).date()
        except ValueError:
            pass
    match = re.fullmatch(r"(\d{4})[/.-](\d{1,2})[/.-](\d{1,2})", text)
    if not match:
        return None
    try:
        return datetime(*map(int, match.groups())).date()
    except ValueError:
        return None


def parse_number(value):
    text = str(value or "").strip().replace("\xa0", "").replace(",", "")
    if not text or text in {"-", "—"}:
        return None
    negative = text.startswith("(") and text.endswith(")")
    cleaned = re.sub(r"[^0-9.\-]", "", text)
    try:
        number = float(cleaned)
        return -number if negative else number
    except ValueError:
        return None


def cell(row, index):
    return row[index] if index < len(row) else ""


request = urllib.request.Request(CSV_URL, headers={"User-Agent": "AgentStage-Ozon-Report/1.0"})
with urllib.request.urlopen(request, timeout=90) as response:
    text = response.read().decode("utf-8-sig")
rows = list(csv.reader(text.splitlines()))

now = datetime.now(TIMEZONE)
end_date = now.date() - timedelta(days=1)
start_date = end_date - timedelta(days=DAYS - 1)
aggregated = defaultdict(lambda: {"installs": 0.0, "spend": 0.0, "cac_values": []})

for row in rows:
    for block in BLOCKS:
        day = parse_date(cell(row, block["date"]))
        if day is None or not (start_date <= day <= end_date):
            continue
        country = str(cell(row, block["geo"])).strip().upper() or block["country"]
        if country != block["country"]:
            continue
        key = (country, day.isoformat())
        installs = parse_number(cell(row, block["installs"]))
        spend = parse_number(cell(row, block["spend"]))
        cac = parse_number(cell(row, block["cac"]))
        aggregated[key]["installs"] += installs or 0
        aggregated[key]["spend"] += spend or 0
        if cac is not None:
            aggregated[key]["cac_values"].append(cac)

records = []
for country in [item["country"] for item in BLOCKS]:
    for offset in range(DAYS):
        day = (start_date + timedelta(days=offset)).isoformat()
        values = aggregated.get((country, day), {"installs": 0, "spend": 0, "cac_values": []})
        cac_values = values["cac_values"]
        records.append({
            "date": day,
            "country": country,
            "installs": round(values["installs"], 2),
            "spend": round(values["spend"], 4),
            "cac": round(sum(cac_values) / len(cac_values), 4) if cac_values else None,
        })

summary = []
for country in [item["country"] for item in BLOCKS]:
    country_rows = [row for row in records if row["country"] == country]
    cac_values = [row["cac"] for row in country_rows if row["cac"] is not None]
    summary.append({
        "country": country,
        "installs": round(sum(row["installs"] for row in country_rows), 2),
        "spend": round(sum(row["spend"] for row in country_rows), 4),
        "avg_cac": round(sum(cac_values) / len(cac_values), 4) if cac_values else None,
        "cac_days": len(cac_values),
    })

payload = {
    "title": "Ozon · 过去14天国家投放表现",
    "generated_at": now.isoformat(timespec="seconds"),
    "timezone": "Asia/Shanghai",
    "period": {"start": start_date.isoformat(), "end": end_date.isoformat(), "days": DAYS},
    "countries": [item["country"] for item in BLOCKS],
    "summary": summary,
    "records": records,
    "definitions": {
        "installs": "Google Sheet 各国家区块的 Install",
        "spend": "Google Sheet 各国家区块的 Daily Spend（USD）",
        "cac": "Google Sheet 各国家区块的 CAC from client report；汇总值为有数据日期的日均 CAC",
    },
}

with open(sys.argv[1], "w", encoding="utf-8") as handle:
    json.dump(payload, handle, ensure_ascii=False, indent=2)
PY

mv "$TMP_FILE" "$OUTPUT_FILE"
echo "Updated $OUTPUT_FILE"