Question

我制作了一个脚本，它读取给定的输入文件（ csv ），以某种方式操纵数据并写入输出文件（ csv ）。

就我而言，我的给定输入文件如下所示：

| sku | article_name |
| 1   | MyArticle    |

对于我的输出文件，我需要重新安排这些列（还有更多，但我想我可以解决它，当有人给我指路时）

我的输出文件应如下所示：

| article_name | another_column | sku |
| MyArticle    |                | 1   |

请注意，这是一个新列，不在源csv文件中，但必须打印它（顺序也很重要）

这是我到目前为止所做的：

#!/usr/bin/env python
# -*- coding: latin_1 -*-

import csv
import argparse
import sys


header_mappings = {'attr_artikel_bezeichnung1': 'ARTICLE LABEL',
                   'sku': 'ARTICLE NUMBER',
                   'Article label locale': 'Article label locale',
                   'attr_purchaseprice': 'EK-Preis',
                   'attr_salesPrice': 'EuroNettoPreis',
                   'attr_salesunit': 'Einheit',
                   'attr_salesvatcode': 'MwSt.-Satz',
                   'attr_suppliercode': 'Lieferantennummer',
                   'attr_suppliersitemcode': 'Artikelnummer Lieferant',
                   'attr_isbatchitem': 'SNWarenausgang'}

row_mapping = {'Einheit': {'pc': 'St.'},
               'MwSt.-Satz': {'3': '19'}}


def remap_header(header):
    for h_map in header_mappings:
        if h_map in header:
            yield header_mappings.get(h_map), header.get(h_map)


def map_header(header):
    for elem in header:
        yield elem, header.index(elem)


def read_csv(filename):
    with open(filename, 'rb') as incsv:
        csv_reader = csv.reader(incsv, delimiter=';')
        for r in csv_reader:
            yield r


def add_header(header, fields=()):
    for f in fields:
        header.append(f)

    return header


def duplicate(csv_row, header_name, fields):
    csv_row[new_csv_header.index(fields)] = csv_row[new_csv_header.index(header_name)]
    return csv_row


def do_new_row(csv_row):
    for header_name in new_csv_header:
        for r_map in row_mapping:
            row_content = csv_row[mapped_header.get(r_map)]
            if row_content in row_mapping.get(r_map):
                csv_row[mapped_header.get(r_map)] = row_mapping.get(r_map).get(row_content)
        try:
            yield csv_row[mapped_header.get(header_name)]
        except TypeError:
            continue


if __name__ == '__main__':

    parser = argparse.ArgumentParser()
    parser.add_argument('-i', '--infile', metavar='CSV')
    parser.add_argument('-o', '--outfile', metavar='CSV')

    args = parser.parse_args()
    arguments = vars(args)
    if len(sys.argv[1:]) == 0:
        parser.print_usage()
        sys.exit(0)

    # print arguments
    # parse_csv(**arguments)
    """
    """
    csv_reader_iter = read_csv(arguments.get('infile'))

    # neuer csv header
    new_csv_header = list()
    csv_header = next(csv_reader_iter)
    for h in csv_header:
        if h in header_mappings:
            new_csv_header.append(header_mappings.get(h))

    # print new_csv_header
    new_csv_header = add_header(new_csv_header, ('Article label locale', 'Nummer'))
    mapped_header = dict(remap_header(dict(map_header(csv_header))))
    # print mapped_header

    with open(arguments.get('outfile'), 'wb') as outcsv:
        csv_writer = csv.writer(outcsv, delimiter=';')
        csv_writer.writerow(new_csv_header)
        for row in csv_reader_iter:
            row = list(do_new_row(row))
            delta = len(new_csv_header) - len(row)
            if delta > 0:
                row = row + (delta * [''])

            # duplicate(row, 'SNWarenausgang', 'SNWareneingang')
            # duplicate(row, 'SNWarenausgang', 'SNWareneingang')
            csv_writer.writerow(row)


    print "Done."
    """
    print new_csv_header
    for row in csv_reader_iter:
        row = list(do_new_row(row))
        delta = len(new_csv_header) - len(row)
        if delta > 0:
            row = row + (delta * [''])

        duplicate(row, 'Herstellernummer', 'Nummer')
        duplicate(row, 'SNWarenausgang', 'SNWareneingang')
        print row
    """

现在，即使它说＆＃34; ARTICLE LABEL＆＃34;首先，首先打印sku。我的猜测：这是由于csv文件的顺序，因为sku是第一个字段......对吗？

Answer 1

如果您使用DictWriter lib中的csv，则可以指定列的顺序。使用DictReader从文件中读取行作为dicts。然后，您只需在创建DictWriter时明确指定键的顺序。

https://docs.python.org/2/library/csv.html#csv.DictReader

Answer 2

正如riotburn已经建议的那样，您可以使用DictWriter及其fieldnames参数来调整新文件中列的顺序。

重新排序文件可能如下所示：

def read_csv (filename):
    with open(filename) as incsv:
       reader = csv.DictReader(incsv, delimiter=';')
       for r in reader:
           yield r

columns = ['article_name', 'another_column', 'sku']

with open('newfile.csv', 'w+') as f:
    writer = csv.DictWriter(f, columns, delimiter=';')
    writer.writeheader()

    for row in read_csv('oldfile.csv'):
        # add a property
        row['another_column'] = 'foo'

        # write row (using the order specified in columns)
        writer.writerow(row)

更改csv（python）中列的顺序

2 个答案: