Inofficial interface to the private API of the Jodel App. Not affiliated with The Jodel Venture GmbH.
Using pip:
pip install jodel_api
or using setup.py:
git clone https://github.com/nborrmann/jodel_api.git cd jodel_api python setup.py install
Calling the bare constructor creates a new account:
>>> import jodel_api
>>> lat, lng, city = 48.148434, 11.567867, "Munich"
>>> j = jodel_api.JodelAccount(lat=lat, lng=lng, city=city)
Creating new account.
get_account_data()
returns all data associated with this account
(censored by me):
>>> j.get_account_data()
{'access_token': 'xxx', 'expiration_date': 1472660000, 'refresh_token': 'xxx', 'distinct_id': 'xxx', 'device_uid': 'xxx'}
Save this data to reuse the account later on, feed it to the JodelAccount() constructor to reinitiate the account. This constructor issues one request to update the location of the account.
>>> j = jodel_api.JodelAccount(lat=lat, lng=lng, city=city, access_token='xxx', expiration_date='xxx',
refresh_token='xxx', distinct_id='xxx', device_uid='xxx')
(204, '')
Add update_location=False
to suppress this behaviour. The
constructor will only instantiate an object, without making any remote
calls:
>>> j = jodel_api.JodelAccount(lat=lat, lng=lng, city=city, update_location=False, **account_data)
For some functionality (look out for error 478) accounts need to be verified by entering a captcha:
>>> j.verify_account()
https://s3-eu-west-1.amazonaws.com/jodel-image-captcha/xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx.png
Open the url above in a browser and enter the images containing a racoon (left to right, starting with 0) separated by spaces: 3 5
Verification failed. Retrying ...
https://s3-eu-west-1.amazonaws.com/jodel-image-captcha/xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx.png
Open the url above in a browser and enter the images containing a racoon (left to right, starting with 0) separated by spaces: 0 3 7
Account successfully verified.
>>> j.verify_acccount()
Account is already verified.
After expiration_date
has passed, call refresh_access_tokens()
to re-authenticate. If refresh_access_token
fails, use
refresh_all_tokens
instead (this is akin to creating a new account,
but preserves the account's data (karma, etc)):
>>> j.refresh_access_token()
(200, {'token_type': 'bearer', 'access_token': 'xxx', 'expires_in': 604800, 'expiration_date': xxx})
>>> j.refresh_all_tokens()
(200, {'expires_in': 604800, 'access_token': 'xxx', 'token_type': 'bearer', 'returning': True,
'refresh_token': 'xxx', 'expiration_date': 1472600000, 'distinct_id': 'xxx'})
All remote API calls return a tuple of HTTP status_code and the response (if possible a dict, parsed from the API response), but might also be a string (error message).
The following API calls are supported (presented without their respective responses):
>>> j.set_location(lat, lng, city, country=None, name=None) # country and name appear to have no effect
>>> j.create_post(message=None, imgpath=None, b64img=None, color=None, ancestor=None, channel="")
>>> j.get_post_details(post_id)
# This api endpoint implements paging and returns at most 50 replies, use the skip parameter to page through the thread.
>>> j.get_post_details_v3(post_id, skip=0)
>>> j.upvote(post_id)
>>> j.downvote(post_id)
>>> j.pin(post_id)
>>> j.unpin(post_id)
>>> j.enable_notifications(post_id)
>>> j.disable_notifications(post_id)
>>> j.give_thanks(post_id)
>>> j.get_share_url(post_id)
>>> j.delete_post(post_id) # Only works on your own posts ಠ_ಠ
>>> j.get_notifications()
>>> j.get_notifications_new()
>>> j.notification_read(post_id=None, notification_id=None)
>>> j.get_recommended_channels()
>>> j.get_channel_meta(channel)
>>> j.follow_channel(channel)
>>> j.unfollow_channel(channel)
>>> j.get_user_config()
>>> j.get_karma()
>>> j.get_captcha()
>>> j.submit_captcha(key, answer)
The following calls can be used to read posts. The parameters skip
,
limit
and after
implement paging. While skip
and limit
are integers, after
is a post_id
parameter and will return all
jodels that follow that one. The former two paramters seem to be
deprecated in favor of the latter, however after
doesn't work
on all /mine/
endpoints (ie. mine=True
or get_my_x_posts
).
The arguments mine
(boolean), hashtag
, channel
(both strings) are exclusive. If
mine
evaluates to true
, the other two arguments are discarded,
if hashtag
evaluates true
, channel
is discarded.
get_newsfeed()
is a new endpoint (as of March 17) that isn't yet
available through the app. It returns all popular (upvotes/comments)
Jodels from a larger timeframe than the usual get_posts()
methods
can access.
>>> j.get_posts_recent(skip=0, limit=60, after="", mine=False, hashtag="", channel="")
>>> j.get_posts_popular(skip=0, limit=60, after="", mine=False, hashtag="", channel="")
>>> j.get_posts_discussed(skip=0, limit=60, after="", mine=False, hashtag="", channel="")
>>> j.get_my_pinned_posts(skip=0, limit=60, after="")
>>> j.get_my_replied_posts(skip=0, limit=60, after="")
>>> j.get_my_voted_posts(skip=0, limit=60, after="")
>>> j.get_newsfeed(after="")
You can pass additional arguments (such as proxies and timeouts) to all
API calls through the **xargs
argument that will be passed to the
requests.request()
function:
>>> j.upvote(post_id, timeout=5, proxies={'https': '127.0.0.1:5000'})
- 429 "Too Many Requests": Your IP is rate-limited.
- 477 "Signed Request Expected": This library should handle request
signing. If this happens on the latest version of
jodel_api
there is a bug in this library. Please open an issue. - 478 "Account not verified": Solve the captcha challenge (eg.
through
verify_account()
). - 502 "Bad Gateway": Something went wrong server-side. This happens
pretty randomly.
jodel_api
automatically retries two times when it sees this error. If you encounter this status, the jodel servers are probably having issues. Try again later.
Nearly all tests in jodel_api_test.py
are integration tests, which
actually hit the Jodel servers. These can fail for any number of reasons
(eg. connectivity issues), which does not necessarily imply there is
something wrong with this library. As this library tries to make few
assumptions about the content of the json responses they test mostly for
status codes, not the contents of the responses (ie. they test whether
the API endpoints are still valid).
For the tests in
class TestUnverifiedAccount
a new account is created on every run and they test read-only functions for which the account does not need to be verified.Tests in
class TestVerifiedAccount
need an already verified account to test voting and creating posts (posts are deleted after creation). To run these tests you need to verify an account by solving the captcha and save itsdevice_uid
in the environment variableJODEL_ACCOUNT
. Runj.get_account_data()['device_uid']
to get the value.Linux:
export JODEL_ACCOUNT=a8aa02[...]dba
Windows (you need to restart the cmd/shell for this to take effect, or set it through gui):
setx JODEL_ACCOUNT a8aa02[...]dba
If this variable is not present, these tests will be skipped.
Clone the directory, install the library and run the tests with
python setup.py test
The Jodel API appears to have the following (IP-based) rate-limits
- max of 60 new account registrations from one IP per half hour
- max of 200 (?) votes (possibly also post creations?) in an unknown time frame
They also hand out perma-bans if you overdo it.