solution.py

import json
import unittest
from functools import partial
from operator import itemgetter
from math import sin, cos, sqrt, asin, radians
from cStringIO import StringIO


def get_records(fl):
    for line in fl:
        record = json.loads(line)
        record['coordinates'] = (
            float(record.pop('latitude')),
            float(record.pop('longitude')))
        yield record


def search_records(records, coordinates, within_kms):
    """
    :param records: iterable providing the records of customers
    :param coordinates: tuple providing latitude and longitude
    :param within_kms: float for within distance in kms
    """
    distance_from_coords = partial(distance_between, coordinates)

    for record in records:
        if distance_from_coords(record['coordinates']) <= within_kms:
            yield record


def distance_between(coords1, coords2):
    """
    :param coords1: tuple providing latitude and longitude
    :param coords2: tuple providing latitude and longitude
    :returns: float
    """
    coords1_lat, coords1_long = to_radians(coords1)
    coords2_lat, coords2_long = to_radians(coords2)

    lat_delta = coords2_lat - coords1_lat
    long_delta = coords2_long - coords1_long

    a = sin(
        lat_delta/2)**2 + cos(coords1_lat) * cos(
            coords2_lat) * sin(long_delta/2)**2
    c = 2 * asin(sqrt(a))

    # approximate radius of earth in km
    radius = 6373.0

    return radius * c


def to_radians(coordinates):
    """
    :param coordinates: tuple providing latitude and longitude
    :returns: list of coordinates in radians
    """
    return map(radians, coordinates)


def main():
    dublin_office_coords = (53.339428, -6.257664)
    within_kms = 100.0

    with open('gistfile1.txt', 'r') as f:
        records = get_records(f)
        sorted_records = sorted(
            search_records(records, dublin_office_coords, within_kms),
            key=itemgetter('user_id')
        )
        # Note on sorted algorithm, this sorting is done
        # in memory. If we need to optimize on this  solution or if the
        # input dataset is huge in the given file, we can use bisect
        # module to sort the collection in place when we read the
        # content from the file. Better yet use this abstraction:
        # https://code.activestate.com/recipes/577197-sortedcollection/

        for record in sorted_records:
            print record['user_id'], record['name']


class TestSearchRecords(unittest.TestCase):

    def setUp(self):
        sample_lines = [
            '{"latitude": "52.986375", "user_id": 12, "name": "Christina McArdle", "longitude": "-6.043701"}',
            '{"latitude": "51.8856167", "user_id": 2, "name": "Ian McArdle", "longitude": "-10.4240951"}',
            '{"latitude": "52.966", "user_id": 15, "name": "Michael Ahearn", "longitude": "-6.463"}'
        ]
        self.sample_data = StringIO('\n'.join(sample_lines))

    def test_search_records(self):
        self.assertEqual(
            len(list(search_records(
                get_records(self.sample_data),
                (53.339428, -6.257664),
                100.0
            ))),
            2
        )

    def test_search_non_existance(self):
        self.assertEqual(
            len(list(search_records(
                get_records(self.sample_data),
                (53.339428, -6.257664),
                1.0
            ))),
            0
        )


if __name__ == "__main__":
    main()