CSV字典-Python

Question

I am attempting to print a dictionary to CSV and receive this error when I run the code.

Error: ValueError: dict contains fields not in fieldnames: 'u', 'r', 'l'

from nytimesarticle import articleAPI
import csv
api = articleAPI('API Key')

res = api.search( q = ['Abbott Laboratories'],
    fq = {'source':['The New York Times']},
    begin_date = 20110101, end_date = 20131231,
    facet_field = ['source'], facet_filter = True)

for m in res['response']['docs']:
    dic = {}
    dic['url'] = m['web_url']
    dic['id'] = m['_id']
    print dic

with open('Abbot_test.csv', 'wb') as output_file:
    dict_writer = csv.DictWriter(output_file, ('url', 'id'))
    dict_writer.writeheader()
    dict_writer.writerows(dic)

When I run with

dict_writer = csv.DictWriter(outputfile, ('url', 'id', 'u', 'r', 'l')

Error: AttributeError: 'str' object has no attribute 'get'

As a side note, when I run the block of code above with

dict_writer.writerow(dic)

In the last line, it prints a single url to the CSV with no error.

Any thoughts?

Answer 1

Looks like your dic = {} creates a blank dictionary every time in the loop. if you're going to do that, then you have to have that loop inside of the " with open " block (so the blank dict is created, then written each time). and at that point it would be " writerow "

Answer 2

Well you're trying to write multiple rows, with the writerows method, but you only supply 1 row. You need to supply a list, of dicts, that contain the proper keys. When Python tries to split the list into individual objects when it loops, it fails because it tries to split one object. This is also why writerow (no s) works for you, because you supply only one dict.

In short, supply a list of dictionaries, not one.

dict_writer.writerows([dic])

But seems like you're only creating one and rewriting it - you probably want to write it into a list instead.

dictlist = []
for m in res['response']['docs']:
    dic = {}
    dic['url'] = m['web_url']
    dic['id'] = m['_id']
    dictlist.append(dic)

dict_writer.writerows(dictlist)

Answer 3

The argument of writerows must be an iterable of dicts . Every row in your csv is represented by a single dict . You keep overriding the same dict when you ought to collect them:

dics = []  # desired structure: [{..}, {..}, ...]
for m in res['response']['docs']:
    dic = {}
    dic['url'] = m['web_url']
    dic['id'] = m['_id']
    dics.append(dic)

# ....
dict_writer.writerows(dics)

Writerows iterates through whatever you pass it and, assuming iterating through dicts it calls the get method, which fails if it is actually iterating through the string keys of a single dict .

Answer 4

its actually as easy as the following if you just wanna do it all at once:

listWriter = csv.DictWriter(
   open('FULL_PATH/output.csv', 'wb'),
   fieldnames=itemDict[itemDict.keys()[0]].keys(),
   delimiter=',',
   quotechar='|',
   quoting=csv.QUOTE_MINIMAL
)

where itemDict is your dictionary to convert.