在python文件中写入字典列表

Question

My goal is to write dictionaries into a file

f = open(...)
i = 0
for dic in dicolist:
    w = csv.DictWriter(f, dic.keys(), lineterminator='\n')
    if i == 0:
        w.writeheader()
        i = 1
    w.writerow(dic)

My goal is to write the dictionaries like this in a file:

field1,field2,field3 #name of the rows of the dictionary
0,1,1 #first dictionary
1,1,1 #second dictionary
2,2,2 #third dictionary

I don't mind about the order of the fields, I want the field1 of the first dictionary to be at the same place as the field2 of the second dictionary.

For instance there will be at first the first field of the first dictionary and then on the next line, it will be the third field of the second dictionary.

What should I use to write the dictionaries in a good order?

What I want is:

fieldx, fieldy, fieldz
fieldx of first dictionary, fieldy of first dictionary, fieldz of first dictionary
fieldx of second dictionary, fieldy of second dictionary, fieldz of second dictionary

Answer 1

If your dictionaries all have the same keys, just get the field names once and use a single writer while iterating over your dictionaries:

with open(...) as csvfile:
     FIELD_NAMES = [...] # keys of your dictionary 
     writer = csv.DictWriter(csvfile, fieldnames=FIELD_NAMES, lineterminator='\n')
     writer.writeheader()
     for dic in dicolist:
         writer.writerow(dic)

Answer 2

import csv

dicts = [
    {'foo': 1, 'bar': 2},
    {'foo': 1, 'spam': 3},
    {'spam': 3, 'bar': 2},
]

out_filename = 'dicts.csv'

# we use set to ensure uniqueness of header names and to be sure
# that we have all column names even if some are missing in one dict:
keys = (set(d.keys()) for d in dicts)
field_names = set.union(*keys)

with open(out_filename, 'w') as csvfile:
    writer = csv.DictWriter(csvfile, fieldnames=list(field_names))
    writer.writeheader()
    for dct in dicts:
        writer.writerow(dct)

with open(out_filename) as f:
    print f.read()