vcf2gpx.py

import vobject
import requests
import time
import sys
import re
import datetime
import colorama
from colorama import Fore, Style

colorama.init(autoreset=True)

def clean_address_part(part):
    if not part:
        return ''
    part = part.replace('\\n', ', ').replace('^n', ', ')
    return part.strip()

def parse_geo(geo_value):
    geo_value = geo_value.strip()
    if geo_value.lower().startswith('geo:'):
        geo_value = geo_value[4:]
    geo_value = geo_value.replace('\\,', ',').replace(';', ',').replace(' ', '')
    m = re.match(r'^(-?\d+(?:\.\d+)?),(-?\d+(?:\.\d+)?)$', geo_value)
    if m:
        return float(m.group(1)), float(m.group(2))
    return None, None

def extract_geo_from_vcard(v):
    for prop in v.getChildren():
        if prop.name.upper() == 'GEO':
            return parse_geo(prop.value)
    if 'geo' in v.contents:
        geo_val = v.contents['geo'][0].value
        return parse_geo(geo_val)
    if 'GEO' in v.contents:
        geo_val = v.contents['GEO'][0].value
        return parse_geo(geo_val)
    return None, None

def geocode_mapbox(address, api_key, debug=False):
    if not api_key:
        return None, None
    url = f'https://api.mapbox.com/geocoding/v5/mapbox.places/{requests.utils.quote(address)}.json'
    params = {
        'access_token': api_key,
        'limit': 1,
        'language': 'en',
        'types': 'address'
    }
    if debug:
        req = requests.Request('GET', url, params=params).prepare()
        print(f"DEBUG: Mapbox URL: {req.url}" + Style.RESET_ALL)
    r = requests.get(url, params=params)
    data = r.json()
    if r.ok and data.get('features'):
        feature = data['features'][0]
        lon, lat = feature['center']
        return lat, lon
    else:
        if debug:
            print(f"DEBUG: Address not found by Mapbox: {address}" + Style.RESET_ALL)
            if 'message' in data:
                print(f"DEBUG: Mapbox message: {data['message']}" + Style.RESET_ALL)
    return None, None

def geocode_nominatim(address, debug=False):
    url = "https://nominatim.openstreetmap.org/search"
    params = {
        "q": address,
        "format": "json",
        "limit": 1,
        "addressdetails": 0,
    }
    headers = {
        "User-Agent": "vcf2gpx/1.0"
    }
    if debug:
        req = requests.Request('GET', url, params=params, headers=headers).prepare()
        print(f"DEBUG: Nominatim URL: {req.url}" + Style.RESET_ALL)
    r = requests.get(url, params=params, headers=headers)
    if r.ok:
        results = r.json()
        if results:
            lat = float(results[0]["lat"])
            lon = float(results[0]["lon"])
            return lat, lon
    return None, None

def vcf2gpx(vcf_path, gpx_path, api_key, debug=False):
    count_geo = 0
    count_mapbox = 0
    count_nominatim = 0
    count_notfound = 0
    count_noprecise = 0
    total_geocoded = 0

    with open(vcf_path, 'r', encoding='utf-8') as f:
        vcards = vobject.readComponents(f)
        points = []
        for v in vcards:
            name = v.fn.value if hasattr(v, 'fn') else "No name"
            lat, lon = extract_geo_from_vcard(v)
            if lat is not None and lon is not None:
                print(Fore.GREEN + f"Geocoded (GEO): {name} -> {lat}, {lon}" + Style.RESET_ALL)
                count_geo += 1
                total_geocoded += 1
            elif not hasattr(v, 'adr') or not v.adr or not getattr(v.adr, 'value', None):
                print(Fore.RED + f"Address not specified in source file: {name}" + Style.RESET_ALL, file=sys.stderr)
                count_noprecise += 1
                continue
            else:
                adr = v.adr.value
                if hasattr(adr, 'label') and adr.label:
                    address = clean_address_part(adr.label)
                else:
                    parts = [
                        clean_address_part(adr.street),
                        clean_address_part(adr.code),
                        clean_address_part(adr.city),
                        clean_address_part(adr.region),
                        clean_address_part(adr.country)
                    ]
                    address = ', '.join([p for p in parts if p])
                if not address or len(address.split(',')) < 2:
                    if debug:
                        print(f"DEBUG: Address too incomplete, ignored: {address}" + Style.RESET_ALL)
                    print(Fore.RED + f"Not found: {name} ({address})" + Style.RESET_ALL, file=sys.stderr)
                    count_notfound += 1
                    continue
                lat, lon = None, None
                if api_key:
                    lat, lon = geocode_mapbox(address, api_key, debug=debug)
                    if lat and lon:
                        print(Fore.GREEN + f"Geocoded (Mapbox): {name} -> {lat}, {lon}" + Style.RESET_ALL)
                        count_mapbox += 1
                        total_geocoded += 1
                    else:
                        lat, lon = geocode_nominatim(address, debug=debug)
                        if lat and lon:
                            print(Fore.GREEN + f"Geocoded (Nominatim): {name} -> {lat}, {lon}" + Style.RESET_ALL)
                            count_nominatim += 1
                            total_geocoded += 1
                        else:
                            print(Fore.RED + f"Not found: {name} ({address})" + Style.RESET_ALL, file=sys.stderr)
                            count_notfound += 1
                else:
                    lat, lon = geocode_nominatim(address, debug=debug)
                    if lat and lon:
                        print(Fore.GREEN + f"Geocoded (Nominatim): {name} -> {lat}, {lon}" + Style.RESET_ALL)
                        count_nominatim += 1
                        total_geocoded += 1
                    else:
                        print(Fore.RED + f"Not found: {name} ({address})" + Style.RESET_ALL, file=sys.stderr)
                        count_notfound += 1
                time.sleep(0.2)
            if lat is not None and lon is not None:
                points.append((name, lat, lon))

        # Add the metadata block with description in English
        geocoding_date = datetime.datetime.now().strftime('%A, %d %B %Y')
        with open(gpx_path, 'w', encoding='utf-8') as gpx:
            gpx.write('<?xml version="1.0" encoding="UTF-8"?>\n')
            gpx.write('<gpx version="1.1" creator="VCF to GPX">\n')
            gpx.write('  <metadata>\n')
            gpx.write('    <name>Geocoded contacts</name>\n')
            gpx.write(f'    <desc>Waypoints geocoded from exported addresses on {geocoding_date}, by VCF to GPX script.</desc>\n')
            gpx.write('  </metadata>\n')
            for name, lat, lon in points:
                safe_name = name.replace('&', '_').replace('<', '_').replace('>', '_').replace('"', '_')
                gpx.write(f'  <wpt lat="{lat}" lon="{lon}"><name>{safe_name}</name></wpt>\n')
            gpx.write('</gpx>\n')

    print()
    print(Fore.GREEN + f"{total_geocoded} addresses geocoded (including {count_mapbox} with Mapbox, {count_nominatim} with Nominatim, and {count_geo} with coordinates included in vCard)." + Style.RESET_ALL)
    print(Fore.RED + f"{count_notfound} addresses not found." + Style.RESET_ALL)
    print(Fore.RED + f"{count_noprecise} addresses not specified in source file." + Style.RESET_ALL)

if __name__ == "__main__":
    debug = False
    args = sys.argv[1:]
    if '--debug' in args:
        debug = True
        args.remove('--debug')
    if len(args) == 2:
        vcf2gpx(args[0], args[1], None, debug=debug)
    elif len(args) == 3:
        vcf2gpx(args[0], args[1], args[2], debug=debug)
    else:
        print("Usage: python3 vcf2gpx.py file.vcf file.gpx [YOUR_MAPBOX_TOKEN] [--debug]")
        sys.exit(1)