@vmeyet: This looks great! @joshthecoder: Does this look good to you?

Yes looks fine. We can just fix the test locally and push it manually.

@Aaron1011 @joshthecoder, thanks for taking care !

Could you show an example of updating with an object rather than a local file? I am hoping to be able to encode images I scrape and update them to a Twitter account.

Thanks.

@petulla the problem is the _pack_image method which does various check regarding the file size etc.

For my use I made a workaround, by monkey patching _pack_image:

• without the file checks and using 'Content-Transfer-Encoding: base64' the following would work:

import base64
import requests
from cStringIO import StringIO
thefile = base64.b64encode(StringIO(requests.get(url).content)

tweepy.API.update_with_media(thefile, status='my status')

May be it worth opening another issue, and to make a PR: there are some refacto needed with _pack_image, plus check for the url image to be thought.

I had something similar but couldn't get it to work; your fix is also throwing an error for me.

Any ideas?

Here's the code:

image = value[0]
#encoding
image = base64.b64encode(StringIO(requests.get(image).content))
#passing
twitter_api.update_with_media(image, status=statustext)

Error: TypeError: must be string or buffer, not cStringIO.StringI

oups made a typo:
image = base64.b64encode(StringIO(requests.get(image).content).read())

The encoding seems to be working. Still throwing an error, though. Indeed pack_image is unhappy.

'''
92 """ statuses/update_with_media """
93 def update_with_media(self, filename, _args, *_kwargs):
---> 94 headers, post_data = API._pack_image(filename, 3072, form_field='media[]')
95 kwargs.update({'headers': headers, 'post_data': post_data})
96

707 raise TweepError('File is too big, must be less than 700kb.')
708 except os.error:
--> 709 raise TweepError('Unable to access file')
710
711 # image must be gif, jpeg, or png

TweepError: Unable to access file'''

of course, you need to monkey patch the _pack_image method which it seems you have not. (that's one of the check in _pack_image that is used when this is a local file)

https://github.com/vmeyet/tweepy/blob/1c108bba79090b054b84d45d8031019ea3475585/tweepy/api.py#L690-694

Do you mind breaking out how to do that? I have never done that before.

the basic idea is:

import base64
import requests
import tweepy
from cStringIO import StringIO
from tweepy.binder import bind_api


def _pack_image_from_url(url, max_size, form_field="image", file_type='image/png'):

    #
    # Checks for the url validity goes there
    base64_encoded = base64.b64encode(StringIO(requests.get(url).content).read())

    BOUNDARY = 'Tw3ePy Pa7cH3d'
    body = []
    body.append('--' + BOUNDARY)
    body.append('Content-Disposition: form-data; name="%s"; filename="%s"' % (form_field, url))
    body.append('Content-Type: %s' % file_type)
    body.append('Content-Transfer-Encoding: base64')
    body.append('')
    body.append(base64_encoded)
    body.append('--' + BOUNDARY + '--')
    body.append('')
    body = '\r\n'.join(body)

    headers = {
        'Content-Type': 'multipart/form-data; boundary=%s' % BOUNDARY,
        'Content-Length': str(len(body))
    }

    return headers, body

def update_with_media_url(self, url, *args, **kwargs):
    headers, post_data = _pack_image_from_url(url, 3072, form_field='media[]')
    kwargs.update({'headers': headers, 'post_data': post_data})

    return bind_api(
        path='/statuses/update_with_media.json',
        method = 'POST',
        payload_type='status',
        allowed_param = [
            'status', 'possibly_sensitive', 'in_reply_to_status_id', 'lat', 'long',
            'place_id', 'display_coordinates'
        ],
        require_auth=True
    )(self, *args, **kwargs)


def patch_tweepy():
    tweepy.API.update_with_media_url = update_with_media_url


patch_tweepy()

# Now you can do
tweepy.API.update_with_media_url('http://mypic.jpg', status='my status')

But you should definitely open an issue for the tweepy lib to support the feature in the future

@vmeyet: I've added this in d908bf7

@Aaron1011 pretty cool how you are reactive. Thanks for that !

@Aaron1011 @vmeyet

Hi I tried using the patch today but still aren't having any luck. I upgraded Tweepy to 2.3, which I assume has the file improvement.

I'm using the following code. Statustext and twitter_api are already defined. Do I need to pass in the type rather than the filename? I'm a bit confused by your documentation here: #422

#this is our url, assigned to variable image  
image = {an image url}

#get the filename from the url
    imagename = image.split('/')[-1].split('#')[0].split('?')[0]

#convert the url to an encoded image format
    imagere = StringIO(requests.get(image).content).read()
    imagefile = base64.b64encode(imagere)

#here is my argument. Where am I going wrong?
    twitter_api.update_with_media(file=imagefile, filename=imagename, status=statustext)  

@petulla: Sorry, I should have made it more clear.

file should be a file-like object. In this case, you'd want to pass the StringIO instance itself (skip the base64 encoding). Your code should look like this:

#this is our url, assigned to variable image  
image = {an image url}

#get the filename from the url
imagename = image.split('/')[-1].split('#')[0].split('?')[0]

#convert the url to an encoded image format
imagefile = StringIO(requests.get(image).content)

#here is my argument. Where am I going wrong?
twitter_api.update_with_media(file=imagefile, filename=imagename, status=statustext)