python批量翻譯推特數據集

最近在做中文的情感分析，國內沒有比較好的數據集（我指的不是評論數據集，而是類似推文的數據集）。然後在kaggle競賽平臺看到了一個推特推文的數據集。然後就想把它翻譯爲中文，然後再來進行訓練。

直接貼上代碼分析

from importlib import reload

import http.client
import hashlib
import urllib
import urllib.parse
import time
import random
import json
import sys
reload(sys)

#
def transf(q):
    appid = '20190930000338551' # 百度翻譯的appid和sk
    secretKey = 'Xf5oc6ENptZOCEeK6goa'

    httpClient = None
    myurl = '/api/trans/vip/translate'
    fromLang = 'en' # 要翻譯的語言
    toLang = 'zh' # 要得到的語言
    str1 = q.split("\t")
    salt = random.randint(32768, 65536) # 鹽值
    sign = appid + str1[1] + str(salt) + secretKey # 簽證
    m1 = hashlib.md5() # 使用hashlib的md5來加密（單獨的md5模塊在py3中已被廢棄）
    m1.update(sign.encode('utf-8'))
    sign = m1.hexdigest()
    #拼接請求鏈接
    myurl = myurl + '?appid=' + appid + '&q=' + urllib.parse.quote(
        str1[1]) + '&from=' + fromLang + '&to=' + toLang + '&salt=' + str(salt) + '&sign=' + sign
    try:
        httpClient = http.client.HTTPConnection('api.fanyi.baidu.com')
        httpClient.request('GET', myurl)
        # response是HTTPResponse對象
        response = httpClient.getresponse()
        resultJason =  response.read().decode('utf-8')
        js = None
        try:
            js = json.loads(resultJason)  # 將json格式的結果轉換成Python的字典結構
        except Exception as e:
            print
            'loads Json error.'
            print
            e

        key = u"trans_result"
        if key in js:
            dst = js["trans_result"][0]["dst"]  # 取得翻譯後的文本結果
            outStr = str1[0] + "\t" + dst
        else:
            outStr = q
        # 追加寫入文件中
        file_object = open('training.txt', 'a+')
        file_object.write(outStr + '\n')
    except Exception as e:
        print
        e
    finally:
        if httpClient:
            httpClient.close()

# 打開要翻譯的文件
f = open('E:\\deepLearning\\si650winter11\\training.txt','r', encoding='UTF-8')

# 逐行讀取整個文件
lines = f.readlines()
for line in lines:
    transf(line)
    #這裏設置了每秒調用一下，太頻繁，，，api會報錯。
    time.sleep(1)
print
f.tell()
# print f.readline()
f.close()