This program uses the Python geocoder library (https://geocoder.readthedocs.io/)and the OpenStreetMap/Nominatim database to geocode a plain text CSV (comma-separated values)file containing street addresses. The pandas library is used for data wrangling.

geocode.py

# geocode.py
# Richard A. Lent, Thursday, September 19, 2019 at 11:43 AM.
# Inspired by https://janakiev.com/blog/geocoding-in-python/ .

# Tuesday, March 10, 2020. Fixed display of progress dots.

"""Python geocoder

This program uses the Python geocoder library (https://geocoder.readthedocs.io/)
and the OpenStreetMap/Nominatim database to geocode a plain text CSV (comma-separated values)
file containing street addresses. The pandas library is used for data wrangling.

Run the program from the command line like this:

    python geocode.py input.csv output.csv

The input CSV file should look like this (but you can have additional variables):

    label,address
    "Holy Cross","1 College Street, Worcester, Massachusetts"
    "Apple","One Infinite Loop, Cupertino, CA 95014"
    "Google","1600 Amphitheatre Parkway, Mountain View, CA 94043"
    "O'Hare Airport","ORD, Chicago, IL 60666"
    "Paris","Centre Pompidou, Paris, FR"
    "Times Square","Times Square, NY"

Be sure to use the name "address" for the address string.
Enclose label and address strings in double quotes to deal with
embedded commas, apostrophes, and other oddities.

The same data contained in the input file will be written to the
output file, in CSV format, with the addition of the latitude and longitude.

If an address cannot be geocoded the latitude and longitude fields
will be blank (NULL).

The geocoding loop has a built-in time delay of 1 second between each API
call to deal with Nominatim's allowed maximum of 1 request per second.

The code was tested with Anaconda Python 3.7.3. Your mileage may vary.
"""

import geocoder
import time
import sys
import pandas as pd
import datetime

if len(sys.argv) < 3:
    print(__doc__)
    print("Insufficient command line arguments specified,\n")
    print("e.g., python geocode.py input.csv output.csv\n")
    sys.exit("\nProgram terminated. Bad command line arguments.")

addresses = pd.read_csv(sys.argv[1])
addresses['latitude'] = 0
addresses['longitude'] = 0

print("Start: ", datetime.datetime.now())

for i in range(len(addresses)):
    g = geocoder.osm(addresses.loc[i, 'address'])
    addresses.loc[i, 'latitude'] = g.lat
    addresses.loc[i, 'longitude'] = g.lng
    # Sleep to ensure to be below 1 request per second.
    # See https://operations.osmfoundation.org/policies/nominatim/.
    time.sleep(1)
    sys.stdout.write(".")
    sys.stdout.flush()
    
print("\nDone: ", datetime.datetime.now())

addresses.to_csv(sys.argv[2], index=None, header=True)