うっかりAmazon 3SにNFD(utf-8-mac)なファイル名をアップロードしたときに修正する方法

gorilla3.py

# coding: utf-8
import re
import unicodedata
from sh import aws

BUCKET = "aws-bucket-name" # 対象のs3 bucket名

s3 = aws.bake("s3")

def _s3rename(path, name, rename):
  namePath = "s3://"+BUCKET+path+name
  renamePath = "s3://"+BUCKET+path+rename
  # print("mv "+namePath+" " + renamePath)
  result = s3("mv",namePath,renamePath)
  print(result)

def s3rename(path):
  for name in s3("ls","s3://"+BUCKET+path):
    name = re.sub(r'\d+\-\d+\-\d+(\s+)\d+\:\d+\:\d+\s+\d+\s+','',name.rstrip());
    rename = unicodedata.normalize("NFC",name)
    if name != rename :
      _s3rename(path,name,rename)
    else:
      print("non NFD:"+name)

## 実行例
s3rename("/hoge/fuga/")

使い方

必要なライブラリを用意

sudo pip install awscli
sudo pip install sh

AWSの設定

aws configure 

コマンドの実行

$ python ./gorilla3.py