読者です 読者をやめる 読者になる 読者になる

日に日に分からんことが増えていく…

プログラマー初心者としての日々を学んでいることを記録していく。

日本語のcsvを取り扱う際便利(python2)

Python2

csvunicode変換は手間取るけど、良さげな変換をクラスにしてまとめている人がいたので拝借した。

import csv
import sys
import tempfile
import shutil
import os

calendar_dict = {
    "January": "1",
    "February": "2",
    "March": "3",
    "April": "4",
    "May": "5",
    "June": "6",
    "July": "7",
    "August": "8",
    "September": "9",
    "October": "10",
    "November": "11",
    "December": "12",
}


class UnicodeCsvReader:
    """csv.reader wrapper which decodes each value with designated encoding"""
    def __init__(self, iterable, dialect='excel', encoding="utf-8", *args, **kwds):
        self.reader = csv.reader(iterable, dialect=dialect, *args, **kwds)
        self.encoding = encoding
        self.dialect = self.reader.dialect
        self.line_num = 0

    def __iter__(self):
        return self

    def decode(self, value):
        return value and value.decode(self.encoding) or value

    def next(self):
        # csv.reader.next returns a list of values of next row
        cols = [self.decode(x) for x in self.reader.next()]
        self.line_num = self.reader.line_num
        return cols


class UnicodeCsvDictReader(csv.DictReader):
    def __init__(self, f, fieldnames=None, restkey=None, restval=None,
                 dialect="excel", encoding="utf-8", *args, **kwds):
        csv.DictReader.__init__(
            self, f, fieldnames, restkey, restval, dialect, *args, **kwds)
        self.encoding = encoding
        self.reader = UnicodeCsvReader(f, encoding=encoding)


def reader_csv(input_file, e="utf-8"):
    with open(input_file, 'r') as f:
        reader = UnicodeCsvReader(f)
        for row in reader:
            yield row


# ここの変換処理にてすべてOK
def writer_tsv(row):
    try:
        date = row[0].split()
        if date[0] in calendar_dict.keys():
            row[0] = "-".join([date[2], calendar_dict.get(date[0], ''), date[1]])
        row[0] = row[0].split(",")[0]
        row[1] = row[1].replace(u' ', u' ').replace(u'+', u' ')
    except Exception as e1:
        print e1, row
    finally:
        print row
        return u'\t'.join(row) + u'\n'


def write_str_into_file(iterable, output_filename,):
    with tempfile.NamedTemporaryFile(delete=False, dir='/var/tmp',) as f:
        for row in iterable:
            f.write(writer_tsv(row).encode('utf-8'))
        shutil.move(f.name, output_filename)
    if os.path.exists(f.name):
        os.remove(f.name)


def main():
    input_file = sys.sys.argv[1]
    write_str_into_file(
        reader_csv(os.path.abspath(os.path.expanduser(input_file))),
        "/tmp/result.tsv")


if __name__ == '__main__':
    main()

msiz.hatenablog.jp