前処理に使えそうなライブラリ

URL

概要(or Nothing)

サンプルコード(Need)

日本語関連

code:sample.py

import neologdn

neologdn.normalize("ﾊﾝｶｸｶﾅ")

# => 'ハンカクカナ'

neologdn.normalize("全角記号！？＠＃")

# => '全角記号!?@#'

convert 半角 to 全角

code:sample.py

>> import mojimoji

>> print mojimoji.zen_to_han(u'アイウａｂｃ０１２')

ｱｲｳabc012

convert 和暦(str) to 西暦(int)

code:sample.py

from jeraconv import jeraconv

# Create J2W class instance

j2w = jeraconv.J2W()

# ex.1 : General usage

print(j2w.convert('文治6年'))

# result (int) 1190

# ex.2 : Full-width numbers are also available

print(j2w.convert('平成３１年'))

# result (int) 2019

# ex.3 : It is also possible to write "1年" as "元年"

print(j2w.convert('令和元年'))

# result (int) 2019

convert 漢数字(str) to 数字(int)

code:sample.py

from kanjize import int2kanji, kanji2int, Number

print(int2kanji(58076099))

# 五千八百七万六千九十九

print(kanji2int("五千八百七万六千九十九"))

# 58076099

convert 和暦(str) to datetime

janera.strptime の formatが作りこんであってよい

code:sample.py

from japanera import Japanera

janera = Japanera()

janera.strptime('令和元年05月01日', "%-E%-O年%m月%d日")

janera.strptime('令和01年05月01日', "%-E%-o年%m月%d日")

import pandas as pd

# get part of it from list

# 0 2019-05-01

# 1 2020-05-01