translatesimple.py

from googletrans import Translator
from bs4 import BeautifulSoup
import requests

"""
Libraries: 
 - requests (http://docs.python-requests.org/en/master/) by Kenneth Reitz for web requests
 - bs4 (https://www.crummy.com/software/BeautifulSoup/) by Leonard Richardson for parsing HTML pages
 - googletrans (https://pypi.org/project/googletrans/) by SuHun Han for translating
"""

"""
I've learned how to use Beautiful Soup by following its documentation:
https://www.crummy.com/software/BeautifulSoup/bs4/doc/

and this tutorial:
https://www.digitalocean.com/community/tutorials/how-to-scrape-web-pages-with-beautiful-soup-and-python-3
by Lisa Tagliaferri

"""

async def translate(client, message):

    def parsePage(url):
        """take url as a string and parse it"""

        ######################################################
        # Based on Lisa Tagliaferri's tutorial

        page = requests.get(url)
        soup = BeautifulSoup(page.content, 'html.parser')

        #####################################################

        return soup


    def getLanguage(lang):
        """take language name e.g. english and change it to en"""
        soup = parsePage('http://www.lingoes.net/en/translator/langcode.htm')

        languages = []

        for i in soup.find_all('td'):
            languages.append(i.text.lower())

        lang.lower()

        num = languages.index(lang)

        goal = languages[num - 1]
        if goal == 'zh':
            goal = goal + '-CN'
        return goal

    fullQuery = message.content[11:]
    queryList = fullQuery.lower().split(' ')
    toLan = getLanguage(queryList[0])

    queryList.remove(queryList[0])

    query = " ".join(queryList)

    ####################################################
    # Based on example from: https://pypi.org/project/googletrans/

    translator = Translator()
    result = translator.translate(query, dest=toLan)

    ####################################################

    await client.send_message(message.channel, 'Translation: ' + result.text)