Python File I/O

Just for reminding.

pickle

The easiest way is to use pickle, it can support all kinds of data formats, list & dict.

import pickle
test = [1, 2, 3, 4]
# write
with open('test.txt', 'wb') as fp:
    pickle.dump(test, fp)
# read
with open('test.txt', 'rb') as fp:
    read_test = pickle.load(fp)
print(read_test)
# [1, 2, 3, 4]

The problem with pickle is that the output file needs to be opened in binary mode. The following code will output an error. So the file is not readable using text edit tool.

with open('test.txt', 'w') as fp:
    pickle.dump(test, fp)

json

While, the json library can do the same work with readable output file.

import json
data = {"a": 1, "b": 2, "c": 3}
# write
with open('test.txt', 'w') as fp:
    json.dump(data, fp)
# read
with open('test.txt', 'r') as fp:
    data = json.load(fp)
print(data)
# {'a': 1, 'b': 2, 'c': 3}

Write and read list to txt file

# write one item a line
with open('test.txt', 'w') as fp:
    for item in test:
        fp.write('%s\n' % item)
# read
with open('test.txt', 'r') as fp:
    read_test = fp.read().splitlines()
print(read_test)
# ['1', '2', '3', '4']

Note that the element in read_test is a string type, you may need to convert it back to Integer or other formats.

read_test = list(map(int, read_test))

write and read list of list to csv file

import csv
test = [[1, 2, 3], [4, 5, 6]]
# write
with open('test.csv', 'w') as fp:
    wr = csv.writer(fp)
    wr.writerows(test)
# read
with open('test.csv', 'r') as fp:
    read_test = [rec for rec in csv.reader(fp)]
print(read_test)
# [['1', '2', '3'], ['4', '5', '6']]

Written on March 21, 2018