README.md

metadata

license: odbl
language:
  - id
tags:
  - sports
  - location
  - lantitute
  - longtitude

This is The initial dataset we scraped from open maps

this dataset has not been cleaned yet be aware!

# requirements 
!pip install requests

script

import csv
import time
import requests
from urllib.parse import quote

OUT_CSV = "jabodetabek_sports_osm.csv"

BBOX = (-6.80, 106.30, -5.90, 107.20)

OVERPASS_URL = "https://overpass-api.de/api/interpreter"
WIKIDATA_ENTITY_URL = "https://www.wikidata.org/wiki/Special:EntityData/{qid}.json"

FETCH_WIKIDATA_IMAGES = True

HEADERS = {"User-Agent": "jabodetabek-sports-scraper/1.0 (contact: yourname@example.com)"}

def osm_browse_link(osm_type: str, osm_id: int) -> str:
    return f"https://www.openstreetmap.org/{osm_type}/{osm_id}"

def commons_file_url(filename: str, width: int = 1600) -> str:

    fn = filename.strip()
    if fn.lower().startswith("file:"):
        fn = fn.split(":", 1)[1]
    return f"https://commons.wikimedia.org/wiki/Special:FilePath/{quote(fn)}?width={width}"

def extract_image_link(tags: dict) -> str:

    img = tags.get("image")
    if img:
        if img.startswith("http"):
            return img
        return commons_file_url(img)

    wm = tags.get("wikimedia_commons")
    if wm:
        return commons_file_url(wm)

    qid = tags.get("wikidata")
    if FETCH_WIKIDATA_IMAGES and qid and qid.upper().startswith("Q"):
        try:
            r = requests.get(WIKIDATA_ENTITY_URL.format(qid=qid), headers=HEADERS, timeout=30)
            if r.status_code == 200:
                data = r.json()
                ent = data.get("entities", {}).get(qid.upper(), {})
                claims = ent.get("claims", {})
                p18 = claims.get("P18", [])
                if p18:
                    filename = p18[0]["mainsnak"]["datavalue"]["value"]
                    return commons_file_url(filename)
        except Exception:
            pass

    return ""

def compose_address(tags: dict) -> str:

    if "addr:full" in tags:
        return tags["addr:full"]

    parts = []

    street = tags.get("addr:street")
    houseno = tags.get("addr:housenumber")
    if street and houseno:
        parts.append(f"{street} {houseno}")
    elif street:
        parts.append(street)

    for k in ("addr:neighbourhood", "addr:suburb", "addr:village"):
        if tags.get(k):
            parts.append(tags[k])

    for k in ("addr:city", "addr:municipality", "addr:county"):
        if tags.get(k):
            parts.append(tags[k])

    for k in ("addr:province", "addr:state"):
        if tags.get(k):
            parts.append(tags[k])

    if tags.get("addr:postcode"):
        parts.append(tags["addr:postcode"])

    return ", ".join(parts)

def build_types(tags: dict) -> str:
    bits = []
    if "leisure" in tags:
        bits.append(f"leisure:{tags['leisure']}")
    if "amenity" in tags:
        bits.append(f"amenity:{tags['amenity']}")
    if "sport" in tags:
        bits.append(f"sport:{tags['sport']}")
    return ", ".join(bits)

def fetch_overpass(bbox):
    s, w, n, e = bbox

    leisure_regex = "^(sports_centre|fitness_centre|stadium|pitch|swimming_pool|track)$"

    query = f"""
    [out:json][timeout:180];
    (
      node["leisure"~"{leisure_regex}"]({s},{w},{n},{e});
      way["leisure"~"{leisure_regex}"]({s},{w},{n},{e});
      relation["leisure"~"{leisure_regex}"]({s},{w},{n},{e});

      // Any feature explicitly tagged with sport=*, but avoid retail shops
      node["sport"]["shop"!~".*"]({s},{w},{n},{e});
      way["sport"]["shop"!~".*"]({s},{w},{n},{e});
      relation["sport"]["shop"!~".*"]({s},{w},{n},{e});
    );
    out center tags;
    """
    r = requests.post(OVERPASS_URL, data={"data": query}, headers=HEADERS, timeout=180)
    r.raise_for_status()
    return r.json().get("elements", [])

def element_coords(el) -> tuple[float, float]:
    if el["type"] == "node":
        return el.get("lat"), el.get("lon")

    c = el.get("center") or {}
    return c.get("lat"), c.get("lon")

def main():
    elements = fetch_overpass(BBOX)
    seen = set()
    rows = []

    for el in elements:
        el_type = el.get("type")            

        el_id = el.get("id")
        tags = el.get("tags", {}) or {}

        key = (el_type, el_id)
        if key in seen:
            continue
        seen.add(key)

        lat, lon = element_coords(el)
        if lat is None or lon is None:
            continue

        name = tags.get("name") or "(Unnamed)"
        addr = compose_address(tags)
        types = build_types(tags)
        osm_link = osm_browse_link(el_type, el_id)
        image_link = extract_image_link(tags)

        likely_sporty = (
            "leisure" in tags and tags["leisure"] in
            {"sports_centre", "fitness_centre", "stadium", "pitch", "swimming_pool", "track"}
        ) or ("sport" in tags)

        if not likely_sporty:
            continue

        rows.append({
            "name": name,
            "address": addr,
            "lat": lat,
            "lng": lon,
            "types": types,
            "osm_link": osm_link,
            "image_link": image_link,
            "osm_type": el_type,
            "osm_id": el_id,
        })

    fieldnames = ["name", "address", "lat", "lng", "types", "osm_link", "image_link", "osm_type", "osm_id"]
    with open(OUT_CSV, "w", newline="", encoding="utf-8") as f:
        w = csv.DictWriter(f, fieldnames=fieldnames)
        w.writeheader()
        for row in rows:
            w.writerow(row)

    print(f"Saved {len(rows)} places to {OUT_CSV}")

if __name__ == "__main__":
    main()