python_scraper_find_a_spring_locations_canada.py

Python
import requests
from bs4 import BeautifulSoup
import json
import time

# =====================================================================
# CanadaTechLife - Find A Spring Scraper (Quebec / Canada Edition)
# This script extracts spring locations, coordinates, and details
# to build a clean JSON dataset for offline navigation or tech-bro road trips.
# =====================================================================

BASE_URL = "https://findaspring.org/spring/locations/north-america/canada/quebec/"
HEADERS = {
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36"
}

def get_spring_details(url):
    try:
        response = requests.get(url, headers=HEADERS, timeout=10)
        if response.status_code != 200:
            return None

        soup = BeautifulSoup(response.text, 'html.parser')

        # Extract title and location data
        title = soup.find('h1', class_='entry-title').text.strip() if soup.find('h1', class_='entry-title') else "Unknown Spring"

        # Basic parsing for GPS coordinates (often found in specific divs or paragraphs)
        gps_lat = ""
        gps_lng = ""
        lat_span = soup.find('span', class_='latitude')
        lng_span = soup.find('span', class_='longitude')

        if lat_span and lng_span:
            gps_lat = lat_span.text.strip()
            gps_lng = lng_span.text.strip()

        description = soup.find('div', class_='entry-content').p.text.strip() if soup.find('div', class_='entry-content') else ""

        return {
            "name": title,
            "url": url,
            "latitude": gps_lat,
            "longitude": gps_lng,
            "description": description
        }
    except Exception as e:
        print(f"Error parsing {url}: {e}")
        return None

def main():
    print("[*] Starting FindASpring Scraper by CanadaTechLife...")
    # Example target: specific spring page (e.g., Source Saint-Lazare)
    target_spring = "https://findaspring.org/spring/locations/north-america/canada/source-saint-lazare-quebec-canada-j7t-2z6/"

    spring_data = get_spring_details(target_spring)

    if spring_data:
        print(f"[+] Successfully extracted: {spring_data['name']}")

        # Save results to JSON
        with open('quebec_springs.json', 'w', encoding='utf-8') as f:
            json.dump([spring_data], f, ensure_ascii=False, indent=4)
        print("[+] Data saved to quebec_springs.json")
    else:
        print("[-] Failed to extract data.")

if __name__ == "__main__":
    main()