How to post file that filename include chinese name

[zming]

Long story short

I want to post a file to jira api, but not get a chinese name response

How to post file that filename include chinese name

Expected behaviour

{...
"filename":"中文.md",
...}

Actual behaviour

{...
"filename":"%E6%9C%AA%E5%91%BD%E5%90%8D.md",
...}

Steps to reproduce

from urllib import parse
import base64
import aiohttp
import asyncio

from aiohttp import FormData

def base64_encode(string):
    encode_string = base64.b64encode(string.encode('utf-8'))
    return str(encode_string,"utf-8")

async def main():
    url = 'http://193.112.46.206:8080/rest/api/2/issue/JIRA-1/attachments'

    data = FormData(charset="UTF-8")
    data.add_field('file',
                   open('中文.md', 'rb'),
                   filename='中文.md',
                   content_type='application/md')
    token = "Basic " + base64_encode("admin:123456")
    headers = {'X-Atlassian-Token': 'no-check',"Authorization":token}
    print(url)
    async with aiohttp.ClientSession() as session:
        async with session.post(url, data=data,headers=headers,) as resp:
            print(resp.status)
            print(await resp.text(encoding="utf-8"))


loop = asyncio.get_event_loop()
loop.run_until_complete(main())

I got the response, the filename incorrect

[{...
"filename":"%E4%B8%AD%E6%96%87.md"
...}]

if I use curl command, I got the Expected result

curl -D- -u admin:123456 -X POST -H "X-Atlassian-Token: no-check" -F "file=@中文.md" http://193.112.46.206:8080/rest/api/2/issue/JIRA-1/attachments
HTTP/1.1 100 

HTTP/1.1 200 
X-AREQUESTID: 1019x735x1
X-XSS-Protection: 1; mode=block
X-Content-Type-Options: nosniff
X-Frame-Options: SAMEORIGIN
Content-Security-Policy: frame-ancestors 'self'
X-ASEN: SEN-L14209471
Set-Cookie: JSESSIONID=4127D49BB16DD99A669CA18B972EE711; Path=/; HttpOnly
X-Seraph-LoginReason: OK
Set-Cookie: atlassian.xsrf.token=BJ1H-386Y-A5GZ-A6XR_e09ef403bffb8cbc3bbcddf074859ac0e419d23f_lin; Path=/
X-ASESSIONID: 1hj5kfz
X-AUSERNAME: admin
Cache-Control: no-cache, no-store, no-transform
Content-Type: application/json;charset=UTF-8
Transfer-Encoding: chunked
Date: Wed, 18 Sep 2019 16:59:13 GMT

[{..."filename":"中文.md",...}]

Your environment

python: 3.6.5
OS: windows 10

[mshandrovskiy]

The decoding looks correct.

>>> import urllib.parse
>>> foo='%E6%9C%AA%E5%91%BD%E5%90%8D.md'
>>> urllib.parse.unquote(foo)
'未命名.md'

[asvetlov]

Is it related to #4012 ?

[zming]

Thanks!
I found a useful way. set function FormData(quote_fields=False) ,like this

    data = FormData(quote_fields=False)
    filename = parse.unquote('中文.md')
    data.add_field('file',
                   open('中文.md', 'rb'),
                   content_type='application/md',
                   filename='中文.md'
                   )

It work

[kohtala]

I think this is related to #4012, but with my patch #4031 the quote_fields=False will still be needed. With the default it'll still perform the file name encoding. What the patch #4031 would help is that quote_fields=False will escape any quotes in the file name so that the result will still parse correctly.

[Dreamsorcerer]

The spec recommends using ASCII only, so the default behaviour here seems correct, with quote_fields available to disable it.