"""Fetch REAL ALPR camera locations from OpenStreetMap (Overpass API). These are the camera

positions crowdsourced largely by the DeFlock project (deflock.me) and tagged

man_made=surveillance + surveillance:type=ALPR. Flock does not publish its camera locations;

this is the closest publicly-available, verifiable dataset.

We query per state (by ISO3166-2 boundary) so every camera has a known state, capping per state

to keep the dataset map-renderable. Output is cached to sample_data/real_cameras.json.

NOTE: only camera *locations/metadata* are public. Plate-read data is NOT public; Sightline

generates realistic synthetic reads against these real camera positions for the POC.

"""

from __future__ import annotations

import json

import os

import time

import httpx

UA = {"User-Agent": "Sightline-ALPR-POC/1.0 (security research; portfolio)"}

ENDPOINTS = [

    "https://overpass-api.de/api/interpreter",

    "https://overpass.kumi.systems/api/interpreter",

    "https://maps.mail.ru/osm/tools/overpass/api/interpreter",

]

OUT = os.path.join(os.path.dirname(__file__), "sample_data", "real_cameras.json")

PER_STATE_CAP = 220

STATES = ["AL","AK","AZ","AR","CA","CO","CT","DE","FL","GA","HI","ID","IL","IN","IA","KS","KY",

          "LA","ME","MD","MA","MI","MN","MS","MO","MT","NE","NV","NH","NJ","NM","NY","NC","ND",

          "OH","OK","OR","PA","RI","SC","SD","TN","TX","UT","VT","VA","WA","WV","WI","WY"]

def query_state(st: str) -> list[dict]:

    q = (f'[out:json][timeout:60];area["ISO3166-2"="US-{st}"]->.s;'

         f'(node(area.s)["man_made"="surveillance"]["surveillance:type"="ALPR"];);'

         f'out body {PER_STATE_CAP};')

    for ep in ENDPOINTS:

        for attempt in range(2):

            try:

                r = httpx.post(ep, data={"data": q}, headers=UA, timeout=90)

                if r.status_code == 200:

                    return r.json().get("elements", [])

                if r.status_code in (429, 504):

                    time.sleep(5)

            except Exception:

                time.sleep(3)

    return []

def main():

    cams: list[dict] = []

    for i, st in enumerate(STATES):

        els = query_state(st)

        kept = 0

        for e in els:

            t = e.get("tags", {})

            op = t.get("operator") or t.get("brand") or t.get("manufacturer") or "Unknown"

            cams.append({

                "osm_id": e.get("id"), "state": st,

                "lat": e.get("lat"), "lon": e.get("lon"),

                "operator": op,

                "manufacturer": t.get("manufacturer") or t.get("brand"),

                "direction": t.get("direction") or t.get("camera:direction"),

                "zone": t.get("surveillance:zone"),

                "mount": t.get("camera:mount") or t.get("support"),

                "ref": t.get("ref"),

            })

            kept += 1

        print(f"[{i+1}/50] {st}: {kept} cameras (running total {len(cams)})", flush=True)

        time.sleep(1.0)

    os.makedirs(os.path.dirname(OUT), exist_ok=True)

    with open(OUT, "w") as fh:

        json.dump(cams, fh)

    ops = {}

    for c in cams:

        ops[c["operator"]] = ops.get(c["operator"], 0) + 1

    top = sorted(ops.items(), key=lambda kv: kv[1], reverse=True)[:8]

    print(f"\nDONE: {len(cams)} real ALPR cameras across {len(STATES)} states -> {OUT}")

    print("operators:", top)

if __name__ == "__main__":

    main()