sciyoshi / pyfacebook Goto Github PK
View Code? Open in Web Editor NEWPyFacebook
PyFacebook
== PyFacebook == PyFacebook is a Python client library for the Facebook API. Samuel Cormier-Iijima ([email protected]) Niran Babalola ([email protected]) Shannon -jj Behrens ([email protected]) David B. Edelstein ([email protected]) Max Battcher ([email protected]) Rohan Deshpande ([email protected]) Matthew Stevens ([email protected]) Sandro Turriate ([email protected]) Benjamin Zimmerman ([email protected]) Gisle Aas ([email protected]) Rand Bradley ([email protected]) Luke Worth ([email protected]) Andreas Cederström ([email protected]) Samuel Hoffstaetter ([email protected]) Andreas Ehn ([email protected]) Lee Li ([email protected]) http://github.com/sciyoshi/pyfacebook/ == Usage == To use this in your own projects, do the standard: python setup.py install == Documentation == Have a look at the examples/ directory. Most of the stuff should be self-explanatory. There is also an example Django app in examples/facebook, which should have enough to get you started. See the developer wiki for more information. == License == Copyright (c) 2008, Samuel Cormier-Iijima All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. * Neither the name of the author nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission. THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
According to http://bugs.developers.facebook.com/show_bug.cgi?id=3961 , it is possible to send pictures by the same mechanism as photos.upload, as multipart data together with the api call.
This is already documented:
http://wiki.developers.facebook.com/index.php/Events.create
The PHP api checks for an extra parameter and calls a different API if present:
public function events_create($event_info, $file = null) {
if ($file) {
return $this->call_upload_method('facebook.events.create',
array('event_info' => $event_info),
$file,
Facebook::get_facebook_url('api-photo') . '/restserver.php');
} else {
return $this->call_method('facebook.events.create',
array('event_info' => $event_info));
}
}
I was hoping to develop some code for manipulating Facebook notes, and after working my way through the tutorial I was disappointed to find that pyfacebook has no support for the Notes API (e.g., Notes.get, http://wiki.developers.facebook.com/index.php/Notes.get). Hope we can see this in a future release.
Any plan for PyFacebook to support OAuth & the new Graph API? The current available python-sdk by Facebook is not good enough to support Facebook Platform App (Canvas app).
When tried to connect through a django app the following error occurs.
Exception in request:
Traceback (most recent call last):
File "/base/data/home/apps/app-name/1.339079629847560090/common/zip-packages/django-1.1.zip/django/core/handlers/base.py", line 92, in get_response
response = callback(request, _callback_args, *_callback_kwargs)
File "/base/data/home/apps/app-name/1.339079629847560090/facebook/djangofb/init.py", line 87, in newview
if not fb.check_session(request):
File "/base/data/home/apps/app-name/1.339079629847560090/facebook/init.py", line 1293, in check_session
self.session_key_expires = int(params['expires'])
ValueError: invalid literal for int() with base 10: 'None'
The error could be solved by replacing line 1292 by this
if params.get('expires'):
if params['expires'] == 'None':
params['expires'] = 0
self.session_key_expires = int(params['expires'])
in facebook/init.py line 218 (fql methods) insert this:
'multiquery': [('queries', json, []),],
then try this:
fbc = Facebook("your_api_key", "")
# assign session_key and secret_key
data = fbc.fql.multiquery({"query1":"select post_id, actor_id, target_id, message from stream where filter_key in (select filter_key from stream_filter where uid={your_uid} AND type='newsfeed')", "query2":"select name from user where uid in (select actor_id from #query1)",})
http://wiki.developers.facebook.com/index.php/Fql.multiquery
users.isAppUser has to be use instead:
http://wiki.developers.facebook.com/index.php/Users.isAppAdded
Seems that the method takes the same params. Replacing isAppAdded by isAppUser worked for me.
Regards,
Xav
Hi,
I cannot get a session with an web application:
{u'error_code': 100, u'error_msg': u'Invalid parameter', u'request_args': [{u'value': u'JSON', u'key': u'format'}, {u'value': u'...', u'key': u'auth_token'}, {u'value': u'...', u'key': u'sig'}, {u'value': u'1.0', u'key': u'v'}, {u'value': u'...', u'key': u'api_key'}, {u'value': u'facebook.auth.getSession', u'key': u'method'}]}
I always get a error 100... what could be the possible errors?
I do something as simple as :
self.api.auth_token = auth_datas['auth_token']
# getSession sets the session_key and uid
self.api.auth.getSession()
self.authenticated = True
Right now, we have facebook.require_login() that forcefully make the user add the application.
Can we have a check where it just makes sure that the user has to be logged into facebook but not necessary has authorize the application?
I want to let all users who logged into facebook to interact with my form because the only thing i care about is their facebook uid. But lock out the users who are not logged into facebook.
Thanks.
I am getting following problem when I ran the tests\test.py
Traceback (most recent call last):
File "C:/source/pyfacebook/tests\tests.py", line 227, in test10
import Image
ImportError: No module named Image
Running pyfacebook by itself causes an error to be generated.
Traceback (most recent call last):
File "/Users/harrytruman/projects/d51/givingnetwork/parts/django/django/core/servers/basehttp.py", line 279, in run
self.result = application(self.environ, self.start_response)
File "/Users/harrytruman/projects/d51/givingnetwork/parts/django/django/core/servers/basehttp.py", line 651, in __call__
return self.application(environ, start_response)
File "/Users/harrytruman/projects/d51/givingnetwork/parts/django/django/core/handlers/wsgi.py", line 245, in __call__
response = middleware_method(request, response)
File "/Users/harrytruman/projects/d51/givingnetwork/parts/pyfacebook/facebook/djangofb/__init__.py", line 240, in process_response
expire_time = datetime.utcfromtimestamp(fb.session_key_expires)
TypeError: a float is required
This method popped up quite high in some profiling results. Below is a patch that removes the redundant string concatenations, dictionary lookups and join() call - the CPU time for me went from .074 to .000.
Index: facebook/__init__.py
===================================================================
--- facebook/__init__.py (rev XXXX)
+++ facebook/__init__.py (working copy)
@@ -1343,23 +1343,27 @@
"""
Validate parameters passed by cookies, namely facebookconnect or js api.
"""
- if not self.api_key in cookies.keys():
+
+ api_key = self.api_key
+ if api_key not in cookies:
return None
- sigkeys = []
- params = dict()
- for k in sorted(cookies.keys()):
- if k.startswith(self.api_key+"_"):
- sigkeys.append(k)
- params[k.replace(self.api_key+"_","")] = cookies[k]
-
-
- vals = ''.join(['%s=%s' % (x.replace(self.api_key+"_",""), cookies[x]) for x in sigkeys])
+ prefix = api_key + "_"
+
+ params = {}
+ vals = ''
+ for k in sorted(cookies):
+ if k.startswith(prefix):
+ key = k.replace(prefix,"")
+ value = cookies[k]
+ params[key] = value
+ vals += '%s=%s' % (key, value)
+
hasher = hashlib.md5(vals)
hasher.update(self.secret_key)
digest = hasher.hexdigest()
- if digest == cookies[self.api_key]:
+ if digest == cookies[api_key]:
return params
else:
return False
can you guys add support for the dashboard api?
thanks
Using pyfacebook with pylons has issues with redirects nesting inside the application iframe. It appears this issue was already addressed for Django with commit http://github.com/sciyoshi/pyfacebook/commit/2a6b442c0e2d7674f9c347a340045286d3ece8f2. Additionally, the transition from Paster to WebOb HTTP exceptions on an earlier commit needed some updates as some variable names differ.
Not sure of the best way to paste in this patch (all in wsgi.py), so here goes:
Top of the file:
import re
Replace CanvasRedirect definition with:
class CanvasRedirect(HTTPOk):
"""This is for redirects within the canvas using FBML."""
try:
from string import Template
except ImportError:
from webob.util.stringtemplate import Template
html_template_obj = Template('<html><head></head><body>${body}</body></html>')
body_template_obj = Template('<fb:redirect url="${location}" />')
def __init__(self, location):
HTTPOk.__init__(self, headers = { 'location': location } )
class TopLevelRedirect(HTTPOk):
"""This is for redirects to top-level Facebook pages."""
try:
from string import Template
except ImportError:
from webob.util.stringtemplate import Template
html_template_obj = Template('<html><head></head><body>${body}</body></html>')
body_template_obj = Template('<script type="text/javascript">\ntop.location.href = "${location}";\n</script>')
def __init__(self, location):
HTTPOk.__init__(self, headers = { 'location': location } )
Change definition of redirect_to:
def redirect_to(self, url):
"""Wrap Pylons' redirect_to function so that it works in_canvas.
By the way, this won't work until after you call
check_session().
"""
if self.in_canvas:
raise CanvasRedirect(url)
elif re.search("^https?:\/\/([^\/]*\.)?facebook\.com(:\d+)?", url.lower()):
raise TopLevelRedirect(url)
else:
pylons_redirect_to(url)
fb.application.getPublicInfo(application_canvas_name="name")
DEBUG: {'api_key': '000000000000my_key00000000000000', 'format': 'XML', 'application_canvas_name ': 'name', 'method': 'facebook.application.getPublicInfo', 'v': '1.0'}
The DEBUG line is just a printout of the args argument to _hash_args(self, args, secret=None)
Notice that for some reason there is a ' ' space character after the application_canvas_name. This is causing signatures to fail.
The issue is on line 121 of facebook/init.py, http://github.com/sciyoshi/pyfacebook/blob/95fa6c33f002fb0b550df8ecb8619f9491848ec0/facebook/__init__.py#L121
The method definition for data.getCookies is as follows:
'data': {
'getCookies': [
('uid', int, []),
('string', str, []),
],
The second argument 'string' is actually the name of the desired cookie and optional, as per http://wiki.developers.facebook.com/index.php/Data.getCookies so the method definition should actually read
'data': {
'getCookies': [
('uid', int, []),
('string', str, ['optional']),
],
This is important because to get all cookies you must not specify a cookie name, so not having that parameter as optional undermines the entire utility of the method.
I've used pyFacebook for weeks without any problem, but since this week every time I call
facebook.auth.getSession()
I receive this error:
Traceback (most recent call last):
File "../file.py", line 27, in nameFile
facebook.auth.getSession()
File "../pyFacebook/Facebook.py", line 669, in getSession
result = self._client('%s.getSession' % self._name, args)
File "../pyFacebook/Facebook.py", line 1110, in call
return self._parse_response(response, method)
File "../pyFacebook/Facebook.py", line 1047, in _parse_response
self._check_error(result)
File "../pyFacebook/Facebook.py", line 998, in _check_error
raise FacebookError(response['error_code'], response['error_msg'], response['request_args'])
pyFacebook.Facebook.FacebookError: Error 100: Invalid parameter
Anyone can help me?
Thank you
Small change to allow an app built on pyfacebook to detect if it is being run inside an FB iframe easily.
Patch:
diff --git a/facebook/init.py b/facebook/init.py
index 3813142..a9a5ef0 100644
--- a/facebook/init.py
+++ b/facebook/init.py
@@ -856,6 +856,10 @@ class Facebook(object):
in_canvas
True if the current request is for a canvas page.
True if the current request is for an HTML page to embed in Facebook
inside an iframe.
@@ -1264,6 +1268,9 @@ class Facebook(object):
if params.get('in_canvas') == '1':
self.in_canvas = True
if params.get('in_iframe') == '1':
self.in_iframe = True
Hello,
i want to use Stream.publish, http://wiki.developers.facebook.com/index.php/Stream.publish
but this feature is not avaible on pyfacebook ?
In pyfacebook/facebook/djangofb/init.py, because of the "from datetime import datetime" on line 4, it should only be datetime.fromtimestamp on line 226, not datetime.datetime.fromtimestamp.
It appears that pyfacebook does not support the required_permissions argument to require_login, unlike Facebook's php API: http://wiki.developers.facebook.com/index.php/Extended_permissions#Requiring_Extended_Permissions. Is this deliberate? Is there any plan to add support for this? Also, is there a workaround that would automatically prompt users adding an FBML app and "logging into" the app for the first time for extended permissions?
I've posted the same question to stackoverflow.com: http://stackoverflow.com/questions/2440327/require-extended-permissions-in-fbml-pyfacebook-app
It's my fault. Please change url_to to url_for in wsgi.py.
See http://code.google.com/p/pyfacebook/issues/detail?id=117 and the comments at the end of http://code.google.com/p/pyfacebook/issues/detail?id=91
didn't see this issue tracking system -> repost from google code site:
_parse_response(response, method)
python version 2.4.3
What steps will reproduce the problem?
got latest init.py, renamed to facebook.py and dropped in same
directory as my simple gt.py script.
script looks as follows:
import sys
from facebook import Facebook
sys.stderr = sys.stdout
print "Content-Type: text/xml" # HTML is following
print # blank line, end of headers
print "Hello World: running version ",sys.version
fb = Facebook('', '')
fb.auth.createToken()
fb.login()
fb.auth.getSession()
fb.users.getInfo([fb.uid], ['name', 'birthday'])
print ""
view script via browser (Chrome v1)
What is the expected output? What do you see instead?
expected: html page displaying "hello world python 2.4.3"
instead error with stacktrace printed:
Traceback (most recent call last): File "gt.py", line 14, in ?
fb.auth.createToken()
File "/home/clientsi/public_html/northshore/gt/facebook.py", line 602, in
createToken token = self._client('%s.createToken' % self._name)
File "/home/clientsi/public_html/northshore/gt/facebook.py", line 1002, in
call return self._parse_response(response, method)
File "/home/clientsi/public_html/northshore/gt/facebook.py", line 941, in
_parse_response dom = minidom.parseString(response)
File "/usr/lib/python2.4/xml/dom/minidom.py", line 1924, in parseString
File "/usr/lib/python2.4/xml/dom/expatbuilder.py", line 32, in ?
File "/usr/lib/python2.4/xml/parsers/expat.py", line 4, in ? ImportError:
/usr/local/lib/python2.4/lib-dynload/pyexpat.so: undefined symbol:
PyUnicodeUCS4_Decode
What version of the product are you using? On what operating system?
Not sure? Can't find version info in any file on site... running on python
v2.4.3
OS & Kernel version: Linux 2.6.28.5-grsec-sg2
Seems the test10 mock httplib.HTTP, however the code is using httplib.HTTPConnection.
Index: tests/test.py =================================================================== --- tests/test.py (revision 4720) +++ tests/test.py (working copy) @@ -232,20 +232,19 @@ fb = facebook.Facebook(my_api_key, my_secret_key) fb.login = self.login - facebook.httplib.HTTP = Mock('httplib.HTTP') + facebook.httplib.HTTPConnection = Mock('httplib.HTTPConnection') http_connection = Mock('http_connection') - facebook.httplib.HTTP.mock_returns = http_connection + facebook.httplib.HTTPConnection.mock_returns = http_connection http_connection.send.mock_returns_func = self.send - def _http_passes(): - return [200,] - http_connection.getreply.mock_returns_func = _http_passes - def read(): response = {"stuff":"stuff"} response_str = simplejson.dumps(response) return response_str - http_connection.file.read.mock_returns_func = read - + mock_response = Mock('httplib.HTTPResponse') + mock_response.status = 200 + mock_response.read.mock_returns_func = read + http_connection.getresponse.mock_returns = mock_response + response = {"session_key":"key","uid":"my_uid","secret":"my_secret","expires":"my_expires"} response_str = simplejson.dumps(response) res = fb.auth.getSession()
Here's the patch:
diff --git a/facebook/__init__.py b/facebook/__init__.py
index 70c9f12..4809f0c 100644
--- a/facebook/__init__.py
+++ b/facebook/__init__.py
@@ -203,6 +203,7 @@ METHODS = {
('template_data', json, ['optional']),
('target_ids', list, ['optional']),
('body_general', str, ['optional']),
+ ('story_size', int, ['optional']),
],
},
I provided this patch previously in Issue 93 on the Google Code site.
Traceback (most recent call last):
File "C:\Python26\lib\site-packages\django\core\servers\basehttp.py", line 279, in run
self.result = application(self.environ, self.start_response)
File "C:\Python26\lib\site-packages\django\core\servers\basehttp.py", line 651, in call
return self.application(environ, start_response)
File "C:\Python26\lib\site-packages\django\core\handlers\wsgi.py", line 241, in call
response = self.get_response(request)
File "C:\Python26\lib\site-packages\django\core\handlers\base.py", line 73, in get_response
response = middleware_method(request)
File "build\bdist.win32\egg\socialregistration\middleware.py", line 13, in process_request
request.facebook.check_session(request)
File "C:\Python26\lib\site-packages\facebook__init__.py", line 1293, in check_session
self.session_key_expires = int(params['expires'])
ValueError: invalid literal for int() with base 10: 'None'
FacebookError
does not call Exception.__init__
, in addition it abuses self.args
, which is used by Exception
to keep the original arguments used when instantiating the exception.
This is an example triggering the bug:
>>> import pickle
>>> from facebook import FacebookError
>>> x = FacebookError("code", "msg", ("arg1", "arg2", "arg3"))
>>> pickle.loads(pickle.dumps(x))
FacebookError('a', 'r', 'g', '3')
Fix for the issue is at: http://github.com/ask/pyfacebook/commit/ac4925452b1a691bc7cde9452947042d96e45de3
This fix has a side-effect in that it renames the previous .args
attribute to
.extra_args
, which is a backwards incompatible change.
I've created an iframe XFBML app where auth_token is a GET parameter from facebook. In check_session, if auth.getSession succeeds[1], the rest of the params aren't parsed. I think they should be, shouldn't they?
Just ran into an issue where my various FB cookies were set to have an "expires" value of "None". This causes an issue in check_session()
which assumes it's going to get something that can be run through int()
.
The exception message is: invalid literal for int() with base 10: 'None'
and it's being thrown from line 1293 of facebook/__init__.py
.
File "/home/lee/test_pyf/t/models.py", line 28, in User
language = models.CharField(maxlength=64, default='Python')
TypeError: init() got an unexpected keyword argument 'maxlength'
for >= django 1.0, change maxlength
to max_length
could fix this error.
When the user removes the application within facebook they post this back to the application. The 'check_session' method fails to validate this condition. I added the following two lines near the end (line 1310) to address this case by triggering up the 'uninstall' parameter:
elif 'canvas_user' in params:
self.uid = params['canvas_user']
elif 'uninstall' in params: << new
self.uid = params['user'] << new
else:
return False
I was using the require_login decorator and noticed that when people arrived at my app for the first time by clicking a link from a request, it wasn't asking them to allow access, and various pyfacebook calls were failing. I tracked this down to the part of check_session:
elif 'canvas_user' in params:
self.uid = params['canvas_user']
else:
return False
return True
In this case there is a canvas_user but we don't have a valid session_key. Shouldn't that mean that check_session returns False, and thus the require_login decorator will redirect the user to allow access?
I am using python v. 2.6 and I get the following:
C:\Python26\lib\site-packages\facebook\djangofb__init__.py:185:
DeprecationWarning: Calling the deprecated function 'new_wrapper'
Downgrade to decorator 2.3 if you want to use this functionality
return decorator.new_wrapper(updated, f)
This patch adds the ability to pass in a callback to the upload method for displaying progress in an application.
--- __init__.py.orig 2009-04-20 00:18:04.000000000 +0100
+++ __init__.py 2009-05-06 13:01:04.000000000 +0100
@@ -635,7 +635,7 @@
class PhotosProxy(PhotosProxy):
"""Special proxy for facebook.photos."""
- def upload(self, image, aid=None, caption=None, size=(604, 1024), filename=None):
+ def upload(self, image, aid=None, caption=None, size=(604, 1024), filename=None, callback=None):
"""Facebook API call. See http://developers.facebook.com/documentation.php?v=1.0&method=photos.upload
size -- an optional size (width, height) to resize the image to before uploading. Resizes by default
@@ -678,21 +678,38 @@
content_type, body = self.__encode_multipart_formdata(list(args.iteritems()), [(image, data)])
urlinfo = urlparse.urlsplit(self._client.facebook_url)
try:
- h = httplib.HTTP(urlinfo[1])
+ content_length = len(body)
+ chunk_size = 4096
+
+ h = httplib.HTTPConnection(urlinfo[1])
h.putrequest('POST', urlinfo[2])
h.putheader('Content-Type', content_type)
- h.putheader('Content-Length', str(len(body)))
+ h.putheader('Content-Length', str(content_length))
h.putheader('MIME-Version', '1.0')
h.putheader('User-Agent', 'PyFacebook Client Library')
h.endheaders()
- h.send(body)
-
- reply = h.getreply()
-
- if reply[0] != 200:
- raise Exception('Error uploading photo: Facebook returned HTTP %s (%s)' % (reply[0], reply[1]))
-
- response = h.file.read()
+
+ if callback:
+ count = 0
+ while len(body) > 0:
+ if len(body) < chunk_size:
+ data = body
+ body = ''
+ else:
+ data = body[0:chunk_size]
+ body = body[chunk_size:]
+
+ h.send(data)
+ count += 1
+ callback(count, chunk_size, content_length)
+ else:
+ h.send(body)
+
+ response = h.getresponse()
+
+ if response.status != 200:
+ raise Exception('Error uploading photo: Facebook returned HTTP %s (%s)' % (response.status, response.reason))
+ response = response.read()
except:
# sending the photo failed, perhaps we are using GAE
try:
I am sorry if am not following the correct procedures, i am not familiar with neither github nor pyfacebook development practices.
According to
http://code.google.com/p/pyfacebook/issues/detail?id=96
and
http://forum.developers.facebook.com/viewtopic.php?id=24188&p=3
pyfacebook will only properly work in Canvas Iframe mode if the parameters are always passed in
the URI.
This means that if the user clicks a link in the app, pyfacebook will ask facebook for credentials, losing whatever action was requested.
The following patch was based on facebook.php, creates Javascript cookies that are compatible with the already existing validate_cookie_signature() method.
I have included a P3P HTTP Header so that it will work on IE, but there are still issues with safari, common to all facebook Iframe apps.
[PATCH] * Creates signed cookies from GET/POST tokens so that app properly works in Iframe canvas mode.
---
facebook/__init__.py | 8 ++++++++
facebook/djangofb/__init__.py | 21 +++++++++++++++++++++
2 files changed, 29 insertions(+), 0 deletions(-)
diff --git a/facebook/__init__.py b/facebook/__init__.py
index 3813142..b498045 100644
--- a/facebook/__init__.py
+++ b/facebook/__init__.py
@@ -856,6 +856,9 @@ class Facebook(object):
in_canvas
True if the current request is for a canvas page.
+ is_session_from_cookie
+ True if the current request session comes from a session cookie.
+
in_profile_tab
True if the current request is for a user's tab for your application.
@@ -918,6 +921,7 @@ class Facebook(object):
self.uid = None
self.page_id = None
self.in_canvas = False
+ self.is_session_from_cookie = False
self.in_profile_tab = False
self.added = False
self.app_name = app_name
@@ -1227,6 +1231,7 @@ class Facebook(object):
if self.session_key and (self.uid or self.page_id):
return True
+
if request.method == 'POST':
params = self.validate_signature(request.POST)
else:
@@ -1253,10 +1258,12 @@ class Facebook(object):
# first check if we are in django - to check cookies
if hasattr(request, 'COOKIES'):
params = self.validate_cookie_signature(request.COOKIES)
+ self.is_session_from_cookie = True
else:
# if not, then we might be on GoogleAppEngine, check their request object cookies
if hasattr(request,'cookies'):
params = self.validate_cookie_signature(request.cookies)
+ self.is_session_from_cookie = True
if not params:
return False
@@ -1358,6 +1365,7 @@ class Facebook(object):
hasher.update(self.secret_key)
digest = hasher.hexdigest()
if digest == cookies[self.api_key]:
+ params['is_session_from_cookie'] = True
return params
else:
return False
diff --git a/facebook/djangofb/__init__.py b/facebook/djangofb/__init__.py
index 7d6e68e..987271b 100644
--- a/facebook/djangofb/__init__.py
+++ b/facebook/djangofb/__init__.py
@@ -5,6 +5,7 @@ import facebook
from django.http import HttpResponse, HttpResponseRedirect
from django.core.exceptions import ImproperlyConfigured
from django.conf import settings
+from datetime import datetime
try:
from threading import local
@@ -213,4 +214,24 @@ class FacebookMiddleware(object):
if not self.internal and request.facebook.session_key and request.facebook.uid:
request.session['facebook_session_key'] = request.facebook.session_key
request.session['facebook_user_id'] = request.facebook.uid
+
+ try:
+ fb = request.facebook
+ except:
+ return response
+
+ if not fb.is_session_from_cookie:
+ # Make sure the browser accepts our session cookies inside an Iframe
+ response['P3P'] = 'CP="NOI DSP COR NID ADMa OPTa OUR NOR"'
+ fb_cookies = {
+ 'expires': fb.session_key_expires,
+ 'session_key': fb.session_key,
+ 'user': fb.uid,
+ }
+
+ expire_time = datetime.utcfromtimestamp(fb.session_key_expires)
+
+ for k in fb_cookies:
+ response.set_cookie(self.api_key + '_' + k, fb_cookies[k], expires=expire_time )
+ response.set_cookie(self.api_key , fb._hash_args(fb_cookies), expires=expire_time )
return response
--
1.6.0.4
diff --git a/facebook/__init__.py b/facebook/__init__.py
index 3813142..3e9a74d 100644
--- a/facebook/__init__.py
+++ b/facebook/__init__.py
@@ -127,6 +127,9 @@ METHODS = {
'getAllocation': [
('integration_point_name', str, []),
],
+ 'banUsers': [
+ ('uids', list, []),
+ ],
},
# auth methods
@@ -225,6 +228,7 @@ METHODS = {
],
'get': [
+ ('uid', int, ['optional']),
('flid', int, ['optional']),
],
I'm having problems with the "added" instance variable. It seems to be
sticking at False even though indeed the user has authorized the
application.
Here's the relevant part of my code, I'm working in Google App Engine:
class MainPage(webapp.RequestHandler):
def get(self):
fb = facebook.Facebook(_FbApiKey, _FbSecret, app_name=_FbAppName)
if fb.check_session(self.request) and fb.added:
pass
else:
url = fb.get_add_url(next=fb.get_app_url())
self.response.out.write('<script language="javascript">top.location.href="' + url + '"</script>')
return
self.response.out.write("<html><body>You are now logged in and you added the app!</body></html>")
It's getting into an infinite redirect cycle, because fb.added is
always False. The first time, check_session returns False, but after
logging in and authorizing the application, it returns True. In any
case, fb.added is always False.
Hi,
Using this code:
result = request.facebook.notifications.sendEmail( userList, "Hello", "World" )
where userList is a list of user ids.
When this code executes, result hasn't any value and users doen't receive any email. I think that it should return an error code as in :
http://wiki.developers.facebook.com/index.php/Notifications.sendEmail
auth.createToken() fails after latest commit. Bug was introduced at the latest version of the code. Going back one revision makes it work again.
Hi, I have been trying to solve this myself, but I cannot make any sense of it.
When I enable XFBML, my app will only log in the user once maybe twice, and then they get sent to the frontpage with a new auth_token attached to the url.
I have a thread about it here and the suggestion from XaeroDegreaz seems to be a bit off, but I am not sure.
THREADLINK: http://forum.developers.facebook.com/viewtopic.php?pid=218519
Maybe this is not affection pyFacebook directly but it is causing it to deliver a cookie/session to the XFBML parser that is then changed for some reason and therefor nothing works.
Could someone with a greater understanding of pyFacebook please take a look in my thread and see if there is anything that can be done on this end, or I will have to continue hitting the facebook dev in the head till they stop breaking things for no reason.
Many thanks for any hepl you might be able to give. If you need any more info, just say the word and I will try to produce it as fast as I can.
the 'is_session_from_cookie' flag is not being set at all in the facebook django middleware, nor are properties like uid, session_key and session_key_expires being properly loaded. I believe this normally occurs in the 'checkSession' method, however it is not being called in the middleware at all.
The effect is that in webkit-based, and possibly other browsers (firefox doesn't seem to honor the cookie settings in the response header for some reasons), the fbconnect cookies are getting overwritten with null data, effectively logging the user out of fbconnect and making the session invalid.
Hi all, it's not really an issue, but we face to a situation where we have many forks and a dead upstream (sciyoshi). Maybe I've not understand the whole power of github, but I never know which fork to merge to find the function I want. I'm never sure that I implement something new or lacking.
Is there any solution to this, or shall we define a new fork as upstream and update the doc consequently ?
For information, I've merged Mrbouyou's fork.
Hi, don't know if this is the right place for this, but I couldn't find anywhere else so, here goes.
I have been trying to use pyfacebook on Google-App-Engine. The facebook app is running in iframe mode and I keep getting errors.
I have a thread about it on SO, which also includes my rather hackish solution. But maybe this will be enough for you to find the error.
http://stackoverflow.com/questions/2039366/app-engine-patch-and-pyfacebook-not-working
I will be monitoring both this issue and the thread on SO, so if you have any more questions just put them where you find it most convenient.
I use pyfacebook as a session proxy for iPhone applications using Facebook Connect. In order for Facebook to return a session secret, you must pass "generate_session_secret=1" as a GET variable in the getSession() method.
This is easily patched by adding generate_session_secret as a named paramater in init() and adding generate_session_secret to args{} in auth.getSession()
New permissions dialog seems to required additional parameters.
You can try the new dialog in your Canvas application settings in Facebook (Developer App -> Your App -> Settings -> Migrations -> "New Data Permissions" -> Enable)
and then catch error:
API Error Code: 100
API Error Description: Invalid parameter
Error Message: Requires valid next URL.
Guess all of us may hit the issue in June because "(Migration ends on: June 1, 2010)"
In the situation where you have GAE accessible on your system's path, but it is not actually being used (e.g. you're currently developing a Pylons project that doesn't use GAE) there is a problem when using the urlread function, which is defined near the top of pyfacebook.py.
Because GAE is on the system, line 79 ("from google.appengine.api import urlfetch") does not raise an import error, so the GAE version of the urlread function is used. However, when it is invoked in this situation it raises an AssertionError with this explanation: 'No api proxy found for service "urlfetch"'. This is probably caused by the fact that GAE isn't actually running in this case.
The quickest fix is just to except this error and use the default behavior:
91c91,92
< result = urlfetch.fetch(url, method=method,
---
> try:
> result = urlfetch.fetch(url, method=method,
92a94,96
> except AssertionError:
> res = urllib2.urlopen(url, data=data)
> return res.read()
There's probably a better, long-term fix for this problem though.
Notes API do not support,hope it can support immediate
I just tried to go through the Desktop application tutorial (http://wiki.developers.facebook.com/index.php/User:PyFacebook_Tutorial), and it works until I try the fb.auth.getSession() which returns the following error:
Traceback (most recent call last):
File "", line 1, in ?
File "/Users/tramdas/repos/pyfacebook/facebook/init.py", line 670, in getSession
result = self._client('%s.getSession' % self._name, args)
File "/Users/tramdas/repos/pyfacebook/facebook/init.py", line 1123, in call
return self._parse_response(response, method)
File "/Users/tramdas/repos/pyfacebook/facebook/init.py", line 1056, in _parse_response
self._check_error(result)
File "/Users/tramdas/repos/pyfacebook/facebook/init.py", line 1007, in _check_error
raise FacebookError(response['error_code'], response['error_msg'], response['request_args'])
facebook.FacebookError: Error 100: Invalid parameter
This was with Python 2.4.6 on OS X, with PyFacebook checked out a couple of days ago.
This could be related to issue 6 (http://github.com/sciyoshi/pyfacebook/issues#issue/6). Seems like this problem has been coming up from time to time, e.g. see here: http://code.google.com/p/pyfacebook/issues/detail?id=50
Is this a known issue? If it is, maybe someone should make a note of it on the wiki tutorial (http://wiki.developers.facebook.com/index.php/User:PyFacebook_Tutorial) since I'm sure many other newbies like me have lost a fair bit of time trying to get it to work :-)
The method template for pages.getInfo is as follows:
'pages': {
'getInfo': [
('page_ids', list, ['optional']),
('uid', int, ['optional']),
],
This causes an Error 100 if one tries to use the getInfo method in a manner similar to users.getInfo. I think the parameters should be:
'pages': {
'getInfo': [
('fields', list, [('default', ['name'])]),
('page_ids', list, ['optional']),
],
to conform with the API. I could be wrong, however; I've jsut started using PyFacebook.
I got integrity error as explained in http://forum.developers.facebook.com/viewtopic.php?pid=152522#p152522 and it's caused by a limit of pyfacebook lib or database model.
I moved id from int(11) to varchar(20) and moved from
user, created = self.get_or_create(id=int(facebook.uid))
to
user, created = self.get_or_create(id=facebook.uid)
now it works and in mysql I get 100000000130682 as a user id (that caused the error, because it was changed to 2147483647 when saving).
it seems a very strange facebook id, I think it's a problem of the library...
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.