#1LIikR_U (Python) [ptt.cc] https://www.ptt.cc/bbs/Python/M.1430965147.A.FDE.html

birds.py

#!/usr/bin/env python3
#-*- coding: utf-8 -*-

import csv
import re
import collections


def main():
    #data = parse_csv('bird0-1.csv')
    data = parse_csv('birdsfinall98.csv')
    count_data = count_birds(data)
    print_count(count_data)
    output_loc_count_csv('bird-loc_count.csv', count_data)


# Dealing with raw csv data
def parse_csv(filename):
    data = []

    with open(filename) as csvfile:
        # only DictReader have fieldnames
        # https://docs.python.org/3.4/library/csv.html#csv.DictReader
        records = csv.DictReader(csvfile, delimiter=',')
        fieldnames = records.fieldnames
        fieldnames[0] = fieldnames[0][1:]  # remove BOM characters (<U+FEFF>)
        # https://en.wikipedia.org/wiki/Byte_order_mark

        for row in records:
            d = {}

            for fieldname in fieldnames[:5]:
                # 'date', 'time', 'location', 'lat', 'lng',
                d[fieldname] = row[fieldname].strip()

            d['birds'] = []
            # ('count', 'birdName'), ('count.1', 'birdName.1'), ...
            for count, bird in zip(fieldnames[5::2], fieldnames[6::2]):
                try:
                    p = (row[bird].strip(), row[count].strip())
                except:
                    continue

                if p[0] not in ('', ' ', '0', '0.0'):
                    d['birds'].append(p)

            data.append(d)
    return data


def count_birds(data):
    count_data = []

    def _grep_num(count):
        try:
            return int(re.search('\d', count).group(0))
        except:
            return 0

    for record in data:
        n = sum(_grep_num(count) for _, count in record['birds'])
        count_data.append((record['location'], n))
    return count_data


def print_count(count_data):
    n_birds = 0
    d = collections.defaultdict(int)
    for i, r in enumerate(count_data, 1):
        # r => (location, count_of_r)
        n_birds += r[1]
        d[r[0]] += r[1]
        print('第 {} 筆資料，位於『{}』，共有 {} 隻。'.format(i, r[0], r[1]))
    print('總計: {} 隻'.format(n_birds))


def output_loc_count_csv(filename, count_data):
    d = collections.defaultdict(int)
    for r in count_data:
        d[r[0]] += r[1]

    with open(filename, 'w') as csvfile:
        fieldnames = ['location', 'count']
        # https://docs.python.org/3.4/library/csv.html#csv.DictWriter
        writer = csv.DictWriter(
                csvfile,
                fieldnames=fieldnames,
                delimiter=','
        )
        writer.writeheader()
        for loc, count in d.items():
            writer.writerow({fieldnames[0]: loc, fieldnames[1]: count})


if __name__ == '__main__':
    main()