Convert array of dictionaries into 2 dimensional array with potentially non-matching fieldnames

Question

I'm using csv.Dictreader to read in a csv file because accessing items by fieldnames is very convenient, but the the csv.Dictwriter class is very finicky about how it deals with fieldnames. I get lots of "dict contains fields not in fieldnames" exceptions, when I specifically don't want the dict to contain all the fieldnames I provided. I also want to be able to provide a list of fieldnames with keys that might not appear in every row of dictionary list.

So I've created a way to convert a list of dictionaries to a 2d array I can use with the csv.writer.writerow function.

Question:

What I'm wondering is if my method is good, bad, or ugly. Is there a better/more pythonic way of converting a list of dictionaries with arbitrary fieldnames to a 2d array? Am I missing something obvious with csv.DictWriter?

Code:

What it does is:

The output will skip fieldnames that you don't provide, but will also just put a blank space if you provide a fieldname that doesn't appear in every (or any) row, but will still include it in the header at the top of the csv file.

def csvdict_to_array(dictlist, fieldnames):
    # Start with header row
    csv_array = [fieldnames]

    for row in dictlist:
        csv_array.append(dictlist_row_to_list(row, fieldnames))

    return csv_array

def dictlist_row_to_list(dictlist_row, fieldnames):
    csv_row = []

    for field in fieldnames:
        if field not in dictlist_row:
            csv_row.append('')
        else:
            csv_row.append(dictlist_row[field])

    return csv_row

Sample input/output:

fieldnames = ["one", "three", "ten"]
dictlist = [{"one": "bob", "two": "bill", "three":"cat"},
            {"one": "john", "two": "jack", "ten":"dog"}]

Output:
one,three,ten
bob,cat,
john,,dog

Thanks for your time

Answer 1

This produces your output:

fieldnames = ["one", "three", "ten"]
dictlist = [{"one": "bob", "two": "bill", "three":"cat"},
            {"one": "john", "two": "jack", "ten":"dog"}]

res = [[item.get(key, '') for key in fieldnames] for item in dictlist]
res.insert(0, fieldnames)
print(res)

Result:

[['one', 'three', 'ten'], ['bob', 'cat', ''], ['john', '', 'dog']]