python json转储unicode错误

我试图将字典存储为utf-8编码的json文档，但我似乎做错了什么，不知道是什么。我已经在下面发布了stacktrace和函数

def parse_contents(res_dict, file):

content_payload = res_dict['parse']['wikitext']['*']
sections_payload = res_dict['parse']['sections']
db = {}
#parse_captures = ("Owner", "Description", "Usage", "Examples", "Options", "Misc.")

def now_next_iter(iterable):
    import itertools
    a, b = itertools.tee(sections_payload)
    next(b, None)
    return itertools.izip(a, b)

def remove_tags(text):
    import re
    return re.sub('<[^<]+?>', '', text)

for cur, nxt in now_next_iter(sections_payload):

    if cur['toclevel'] == 2:
        head = cur['line']
        db[head] = {}
    elif cur['toclevel']  == 3:
        line = cur['line']
        ibo = cur['byteoffset']
        fbo = nxt['byteoffset']

        content = remove_tags(content_payload[ibo:fbo])
        db[head][line] = content #.encode('utf-8')

with io.open(file, 'w', encoding='utf8') as json_db:
    s = json.dumps( db, sort_keys=True, indent=4,
                    separators=(',', ': '))
    json_db.write(s.encode('utf-8'))

输出： 这令人困惑，因为我认为s.encode（'utf-8'）应该将其更改为unicode。

---

您可能需要设置json.dumps可选参数'sure_ascii=False'，和/或在json.dumps中设置encoding='UTF8'，而不仅仅是file.open（）调用，这将允许json包使用其选项处理非ascii数据

---

请参阅此处的文档：

根据您的建议更新了问题，没有运气：（nvm，我是个白痴，我在尝试1时取出了s.encode，成功了。谢谢

    with io.open(file, 'w', encoding='utf8') as json_db:
    s = json.dumps( db, sort_keys=True, indent=4,
                    ensure_ascii=False, encoding='UTF8', separators=(',', ': '))
    s = s.encode('utf-8')
    json_db.write(s)