geocode_csv.py

import requests

from csv import DictReader, DictWriter
from tqdm import tqdm


def add_lat_long(input_file, output_file):
    "Adds latitude and longitude to the primary table"
    reader = DictReader(input_file)
    fieldnames = []
    for f in reader.fieldnames:
        if "Column" not in f:
            fieldnames.append(f)
    fieldnames.extend(["Latitude", "Longitude", "Latitude_2", "Longitude_2"])

    writer = DictWriter(output_file, fieldnames=fieldnames)
    output_list = []

    for row in tqdm(reader):
        address = f"{row['Street Address']}, {row['City']} {row['State']} {row['Zip']}"
        address = address.replace('\r', ' ').replace('\n', ' ')

        results = geocode(address)
        current = {}
        for f in fieldnames:
            current.update({
                f: row.get(f)
            })

        current.update({
            "Latitude": results[0]["geometry"]["location"]["lat"],
            "Longitude": results[0]["geometry"]["location"]["lng"]
        })
        if len(results) > 1:
            current.update({
                "Latitude_2": results[1]["geometry"]["location"]["lat"],
                "Longitude_2": results[1]["geometry"]["location"]["lng"]
            })
        output_list.append(current)

    writer.writeheader()
    writer.writerows(output_list)


def geocode(address):
    "Feed addresses into google's geocoding service, and get list of results"
    payload = {
        "address": address,
        "key": "GOOGLE_API_KEY"
    }
    response = requests.get(
        "https://maps.googleapis.com/maps/api/geocode/json?",
        params=payload
    )
    return response.json()['results']


with open('data/Primary_Info-Table_1.csv') as input:
    with open ('data/formatted_primary.csv', 'w', newline='') as output:
        add_lat_long(input, output)