climperor/_cf_traffic.py

"""Pull traffic for dota2.refining.dev via Cloudflare GraphQL Analytics API.

Credentials from env (keyzoo inject). Read-only; prints a short summary.
"""

from __future__ import annotations

import json
import os
import urllib.error
import urllib.request
from datetime import date, timedelta

HOST = "dota2.refining.dev"
ZONE_NAME = "refining.dev"
DAYS = 14


def headers() -> dict[str, str]:
    email = os.environ.get("CLOUDFLARE_EMAIL") or os.environ.get(
        "KEYZOO_ASSET_META_USERNAME"
    )
    key = os.environ.get("CLOUDFLARE_API_KEY") or os.environ.get(
        "KEYZOO_ASSET_SECRET_GLOBAL_API_KEY"
    )
    if not email or not key:
        raise SystemExit("missing credentials in env")
    return {
        "X-Auth-Email": email,
        "X-Auth-Key": key,
        "Content-Type": "application/json",
    }


def api(path: str) -> dict:
    req = urllib.request.Request(
        "https://api.cloudflare.com/client/v4" + path,
        method="GET",
        headers=headers(),
    )
    with urllib.request.urlopen(req) as r:
        return json.loads(r.read().decode())


def gql(query: str, variables: dict) -> dict:
    req = urllib.request.Request(
        "https://api.cloudflare.com/client/v4/graphql",
        data=json.dumps({"query": query, "variables": variables}).encode(),
        method="POST",
        headers=headers(),
    )
    try:
        with urllib.request.urlopen(req) as r:
            return json.loads(r.read().decode())
    except urllib.error.HTTPError as e:
        body = e.read().decode()
        raise SystemExit(f"HTTP {e.code}: {body[:2000]}") from e


def fmt_bytes(n: int | float | None) -> str:
    if n is None:
        return "-"
    n = float(n)
    for unit in ("B", "KB", "MB", "GB", "TB"):
        if abs(n) < 1024:
            return f"{n:.1f} {unit}"
        n /= 1024
    return f"{n:.1f} PB"


def main() -> None:
    zd = api(f"/zones?name={ZONE_NAME}")
    zone = (zd.get("result") or [None])[0]
    if not zone:
        raise SystemExit(f"zone not found: {zd.get('errors')}")
    zid = zone["id"]

    end = date.today()
    start = end - timedelta(days=DAYS)

    # Free plan: httpRequests1dGroups works for multi-day zone totals.
    # Adaptive (host/path filter) is limited to a 1-day window.
    zone_q = """
    query($zoneTag: string!, $start: Date!, $end: Date!) {
      viewer {
        zones(filter: {zoneTag: $zoneTag}) {
          httpRequests1dGroups(
            limit: 20
            filter: {date_geq: $start, date_leq: $end}
            orderBy: [date_ASC]
          ) {
            dimensions { date }
            sum { requests bytes cachedRequests threats }
            uniq { uniques }
          }
        }
      }
    }
    """
    data = gql(
        zone_q,
        {"zoneTag": zid, "start": start.isoformat(), "end": end.isoformat()},
    )
    if data.get("errors"):
        raise SystemExit(json.dumps(data["errors"], indent=2)[:4000])
    rows = data["data"]["viewer"]["zones"][0]["httpRequests1dGroups"]
    print(f"\n{ZONE_NAME} zone (all hosts)  {start} .. {end}")
    print(f"{'date':<12} {'requests':>10} {'uniques':>10} {'bytes':>12} {'threats':>8}")
    tot_r = tot_u = tot_b = tot_t = 0
    for row in rows:
        d = row["dimensions"]["date"]
        s = row["sum"]
        u = row["uniq"]["uniques"]
        print(
            f"{d:<12} {s['requests']:>10} {u:>10} "
            f"{fmt_bytes(s['bytes']):>12} {s['threats']:>8}"
        )
        tot_r += s["requests"]
        tot_u += u
        tot_b += s["bytes"]
        tot_t += s["threats"]
    print(
        f"{'TOTAL':<12} {tot_r:>10} {tot_u:>10} "
        f"{fmt_bytes(tot_b):>12} {tot_t:>8}"
    )

    # Host-scoped detail: one day at a time (Free quota).
    host_q = """
    query($zoneTag: string!, $day: Date!, $host: string!) {
      viewer {
        zones(filter: {zoneTag: $zoneTag}) {
          total: httpRequestsAdaptiveGroups(
            limit: 1
            filter: {date: $day, clientRequestHTTPHost: $host}
          ) {
            count
            sum { edgeResponseBytes visits }
          }
          byPath: httpRequestsAdaptiveGroups(
            limit: 15
            filter: {date: $day, clientRequestHTTPHost: $host}
            orderBy: [count_DESC]
          ) {
            dimensions { clientRequestPath }
            count
            sum { visits }
          }
          byCountry: httpRequestsAdaptiveGroups(
            limit: 8
            filter: {date: $day, clientRequestHTTPHost: $host}
            orderBy: [count_DESC]
          ) {
            dimensions { clientCountryName }
            count
            sum { visits }
          }
          byStatus: httpRequestsAdaptiveGroups(
            limit: 8
            filter: {date: $day, clientRequestHTTPHost: $host}
            orderBy: [count_DESC]
          ) {
            dimensions { edgeResponseStatus }
            count
          }
        }
      }
    }
    """
    print(f"\n{HOST} host detail (last {min(DAYS, 7)} days, Free=1d/query)")
    print(f"{'date':<12} {'requests':>10} {'visits':>10} {'bytes':>12}")
    host_days = []
    for i in range(min(DAYS, 7), -1, -1):
        day = end - timedelta(days=i)
        hd = gql(
            host_q,
            {"zoneTag": zid, "day": day.isoformat(), "host": HOST},
        )
        if hd.get("errors"):
            print(f"{day.isoformat():<12}  error: {hd['errors'][0].get('message')}")
            continue
        z = hd["data"]["viewer"]["zones"][0]
        tot = (z["total"] or [{}])[0]
        c = tot.get("count") or 0
        v = (tot.get("sum") or {}).get("visits") or 0
        b = (tot.get("sum") or {}).get("edgeResponseBytes") or 0
        print(f"{day.isoformat():<12} {c:>10} {v:>10} {fmt_bytes(b):>12}")
        host_days.append((day, z, c, v, b))

    # Detail for the busiest recent day.
    pick = max(host_days, key=lambda x: x[2], default=None)
    if pick and pick[2] > 0:
        day, z, _, _, _ = pick
        print(f"\nDetail for busiest day {day.isoformat()} ({HOST})")
        print("Top paths:")
        for row in z["byPath"][:15]:
            path = row["dimensions"].get("clientRequestPath") or "/"
            v = (row.get("sum") or {}).get("visits") or 0
            print(f"  {row['count']:>8} req  {v:>6} visits  {path}")
        print("Top countries:")
        for row in z["byCountry"]:
            name = row["dimensions"].get("clientCountryName") or "?"
            v = (row.get("sum") or {}).get("visits") or 0
            print(f"  {row['count']:>8} req  {v:>6} visits  {name}")
        print("Status codes:")
        for row in z["byStatus"]:
            st = row["dimensions"].get("edgeResponseStatus")
            print(f"  {st}: {row['count']}")

    print(
        "\nNote: edge requests/visits via Cloudflare proxy (not RUM pageviews). "
        "Includes assets/bots. Zone totals mix all refining.dev hosts; "
        "host table is dota2-only."
    )


if __name__ == "__main__":
    main()