UnicodeEncodeError: 'cp949' codec can't encode character '\u20a9' in position 90: illegal multibyte sequence

问题

I'm a python beginner.

I'm trying to crawl google play store and export to csv file. But I got a error message.

UnicodeEncodeError: 'cp949' codec can't encode character '\u20a9' in position 90: illegal multibyte sequence

Here is my source code.

When I command print, it works. But it shows error message when exporting to csv file

please help me

from bs4 import BeautifulSoup
import urllib.request
import urllib.parse
import codecs
import json
import pickle
from datetime import datetime
import sys
import csv
import os


req = 'https://play.google.com/store/search?q=hana&c=apps&num=300'



response = urllib.request.urlopen(req)
the_page = response.read()
soup = BeautifulSoup(the_page)


#app_link  = soup.find('a', {'class' : 'title'})
#app_url = app_link.get('href')





for div in soup.findAll( 'div', {'class' : 'details'} ):
    title = div.find( 'a', {'class':'title'} )
    #print(title.get('href')) 
    app_url = title.get('href')

    app_details={}


    g_app_url = 'https://play.google.com' + app_url

    app_response = urllib.request.urlopen(g_app_url)
    app_page = app_response.read()
    soup = BeautifulSoup(app_page)
    #print(soup)


    #print( g_app_url )
    title_div = soup.find( 'div', {'class':'document-title'} )
    app_details['title'] = title_div.find( 'div' ).get_text().strip()

    subtitle = soup.find( 'a', {'class' : 'document-subtitle primary'} )
    app_details['developer'] = subtitle.get_text().strip()
    app_details['developer_link'] = subtitle.get( 'href' ).strip()

    price_buy_span = soup.find( 'span', {'class' : 'price buy'} )
    price = price_buy_span.find_all( 'span' )[-1].get_text().strip()
    price = price[:-4].strip() if price != 'Install' else 'Free' 
    app_details['price'] = price

    rating_value_meta = soup.find( 'meta', {'itemprop' : 'ratingValue'} )
    app_details['rating'] = rating_value_meta.get( 'content' ).strip()

    reviewers_count_meta = soup.find( 'meta', {'itemprop' : 'ratingCount'} )
    app_details['reviewers'] = reviewers_count_meta.get( 'content' ).strip()

    num_downloads_div = soup.find( 'div', {'itemprop' : 'numDownloads'} )
    if num_downloads_div: app_details['downloads'] = num_downloads_div.get_text().strip()

    date_published_div = soup.find( 'div', {'itemprop' : 'datePublished'} )
    app_details['date_published'] = date_published_div.get_text().strip()

    operating_systems_div = soup.find( 'div', {'itemprop' : 'operatingSystems'} )
    app_details['operating_system'] = operating_systems_div.get_text().strip()

    content_rating_div = soup.find( 'div', {'itemprop' : 'contentRating'} )
    app_details['content_rating'] = content_rating_div.get_text().strip()

    category_span = soup.find( 'span', {'itemprop' : 'genre'} )
    app_details['category'] = category_span.get_text()
    #print(app_details)


    with open('result.csv', 'w') as f:  # Just use 'w' mode in 3.x
        w = csv.DictWriter(f, app_details.keys())
        w.writeheader()
        w.writerow(app_details)

回答1:

Python 3 opens text files in the locale default encoding; if that encoding cannot handle the Unicode values you are trying to write to it, pick a different codec:

with open('result.csv', 'w', encoding='UTF-8', newline='') as f:

That'd encode any unicode strings to UTF-8 instead, an encoding which can handle all of the Unicode standard.

Note that the csv module recommends you open files using newline='' to prevent newline translation.

You also need to open the file just once, outside of the for loop:

with open('result.csv', 'w') as f:  # Just use 'w' mode in 3.x
    fields = ('title', 'developer', 'developer_link', 'price', 'rating', 'reviewers',
              'downloads', 'date_published', 'operating_system', 'content_rating',
              'category')
    w = csv.DictWriter(f, )
    w.writeheader()

    for div in soup.findAll( 'div', {'class' : 'details'} ):
        #
        # build app_details
        #

        w.writerow(app_details)

来源：https://stackoverflow.com/questions/20994352/unicodeencodeerror-cp949-codec-cant-encode-character-u20a9-in-position-90