source: perl/modules/Facebook/lib/BarnOwl/Module/Facebook/Handle.pm @ d387e0d

Last change on this file since d387e0d was d387e0d, checked in by Edward Z. Yang <ezyang@mit.edu>, 13 years ago
Rename zsig to permalink. Signed-off-by: Edward Z. Yang <ezyang@mit.edu>
  • Property mode set to 100644
File size: 15.6 KB
Line 
1use warnings;
2use strict;
3
4=head1 NAME
5
6BarnOwl::Module::Facebook::Handle
7
8=head1 DESCRIPTION
9
10Contains everything needed to send and receive messages from Facebook
11
12=cut
13
14package BarnOwl::Module::Facebook::Handle;
15
16use Facebook::Graph;
17
18use List::Util qw(reduce);
19
20eval { require Lingua::EN::Keywords; };
21if ($@) {
22    *keywords = sub {
23        # stupidly pick the longest one, and only return one.
24        my $sentence = shift;
25        $sentence =~ s/[[:punct:]]+/ /g;
26        my @words = split(' ', lc($sentence));
27        return () unless @words;
28        return (reduce{ length($a) > length($b) ? $a : $b } @words,);
29    };
30} else {
31    *keywords = \&Lingua::EN::Keywords::keywords;
32}
33
34use JSON;
35use Date::Parse;
36use POSIX;
37use Ouch;
38
39use Scalar::Util qw(weaken);
40
41use BarnOwl;
42use BarnOwl::Message::Facebook;
43
44our $app_id = 235537266461636; # for application 'barnowl'
45
46# Unfortunately, Facebook does not offer a comment stream, in the same
47# way we can get a post stream using the news feed.  This makes it a bit
48# difficult to de-duplicate comments we have already seen.  We use a
49# simple heuristic to fix this: we check if the comment's time is dated
50# from before our last update, and don't re-post if it's dated before.
51# Be somewhat forgiving, since it's better to duplicate a post than to
52# drop one.  Furthermore, we must use Facebook's idea of time, since the
53# server BarnOwl is running on may be desynchronized.  So we need to
54# utilize Facebook's idea of time, not ours.  We do this by looking at
55# all of the timestamps we see while processing an update, and take the
56# latest one and increment it by one second.
57#
58# What properties do we get with this setup?
59#
60#   - We get comment updates only for the latest N posts on a news feed.
61#   Any later ones, you have to use Facebook's usual mechanisms (e.g.
62#   email notifications).
63#
64#   - Processing a poll is relatively expensive, since we have to
65#   iterate over N new posts.  It might be worthwhile polling for new
66#   comments less frequently than polling for new posts.
67
68sub new {
69    my $class = shift;
70    my $cfg = shift;
71
72    my $self = {
73        'cfg'  => $cfg,
74        'facebook' => undef,
75
76        # Ideally this should be done using Facebook realtime updates,
77        # but we can't assume that the BarnOwl lives on a publically
78        # addressable server (XXX maybe we can setup an option for this.)
79        'friend_timer' => undef,
80
81        # Initialized with our 'time', but will be synced to Facebook
82        # soon enough. (Subtractive amount is just to preseed with some
83        # values.) XXX Remove subtraction altogether.
84        'last_poll' => time - 60 * 60,
85        'timer' => undef,
86
87        # Message polling not implemented yet
88        #'last_message_poll' => time,
89        #'message_timer' => undef,
90
91        # yeah yeah, inelegant, I know.  You can try using
92        # $fb->authorize, but at time of writing (1.0300) they didn't support
93        # the response_type parameter.
94        # 'login_url' => 'https://www.facebook.com/dialog/oauth?client_id=235537266461636&scope=read_stream,read_mailbox,publish_stream,offline_access,read_friendlists,rsvp_event,user_events&redirect_uri=http://www.facebook.com/connect/login_success.html&response_type=token',
95        # minified to fit in most terminal windows.
96        # Be careful about updating these values, since BarnOwl will not
97        # notice that it is missing necessary permissions until it
98        # attempt to perform an operation which fails due to lack of
99        # permissions.
100        'login_url' => 'http://goo.gl/rcM9s',
101
102        'logged_in' => 0,
103
104        # would need another hash for topic de-dup
105        'topics' => {},
106
107        # deduplicated map of names to user ids
108        'friends' => {},
109    };
110
111    bless($self, $class);
112
113    $self->{facebook} = Facebook::Graph->new( app_id => $app_id );
114    if ( defined $self->{cfg}->{token} ) {
115        $self->facebook_do_auth;
116    }
117
118    return $self;
119}
120
121=head2 sleep N
122
123Stop polling Facebook for N seconds.
124
125=cut
126
127sub sleep {
128    my $self  = shift;
129    my $delay = shift;
130
131    # prevent reference cycles
132    my $weak = $self;
133    weaken($weak);
134
135    # Stop any existing timers.
136    if (defined $self->{friend_timer}) {
137        $self->{friend_timer}->stop;
138        $self->{friend_timer} = undef;
139    }
140    if (defined $self->{timer}) {
141        $self->{timer}->stop;
142        $self->{timer} = undef;
143    }
144    if (defined $self->{message_timer}) {
145        # XXX doesn't do anything right now
146        $self->{message_timer}->stop;
147        $self->{message_timer} = undef;
148    }
149
150    $self->{friend_timer} = BarnOwl::Timer->new({
151        name     => "Facebook friend poll",
152        after    => $delay,
153        interval => 60 * 60 * 24,
154        cb       => sub { $weak->poll_friends if $weak }
155       });
156    $self->{timer} = BarnOwl::Timer->new({
157        name     => "Facebook poll",
158        after    => $delay,
159        interval => 90,
160        cb       => sub { $weak->poll_facebook if $weak }
161       });
162    # XXX implement message polling
163}
164
165sub check_result {
166    my $self = shift;
167    if (kiss 400) {
168        # Ugh, no easy way of accessing the JSON error type
169        # which is OAuthException.
170        $self->{logged_in} = 0;
171        $self->facebook_do_auth;
172        return 0;
173    } elsif (hug) {
174        my $code = $@->code;
175        warn "Poll failed with $code: $@";
176        return 0;
177    }
178    return 1;
179}
180
181sub poll_friends {
182    my $self = shift;
183
184    return unless BarnOwl::getvar('facebook:poll') eq 'on';
185    return unless $self->{logged_in};
186
187    my $friends = eval { $self->{facebook}->fetch('me/friends'); };
188    return unless $self->check_result;
189
190    $self->{friends} = {};
191
192    for my $friend ( @{$friends->{data}} ) {
193        if (defined $self->{friends}{$friend->{name}}) {
194            # XXX We should try a little harder here, rather than just
195            # tacking on a number.  Ideally, we should be able to
196            # calculate some extra piece of information that the user
197            # needs to disambiguate between the two users.  An old
198            # version of Facebook used to disambiguate with your primary
199            # network (so you might have Edward Yang (MIT) and Edward
200            # Yang (Cambridge), the idea being that users in the same
201            # network would probably have already disambiguated
202            # themselves with middle names or nicknames.  We no longer
203            # get network information, since Facebook axed that
204            # information, but the Education/Work fields may still be
205            # a reasonable approximation (but which one do you pick?!
206            # The most recent one.)  Since getting this information
207            # involves extra queries, there are also caching and
208            # efficiency concerns (though hopefully you don't have too
209            # many friends with the same name).  Furthermore, accessing
210            # this information requires a pretty hefty extra set of
211            # permissions requests, which we don't currently ask for.
212            #   It may just be better to let users specify custom
213            # aliases for Facebook users, which are added into this
214            # hash.  See also username support.
215            warn "Duplicate friend name " . $friend->{name};
216            my $name = $friend->{name};
217            my $i = 2;
218            while (defined $self->{friends}{$friend->{name} . ' ' . $i}) { $i++; }
219            $self->{friends}{$friend->{name} . ' ' . $i} = $friend->{id};
220        } else {
221            $self->{friends}{$friend->{name}} = $friend->{id};
222        }
223    }
224
225    # XXX We should also have support for usernames, and not just real
226    # names. However, since this data is not returned by the friends
227    # query, it would require a rather expensive set of queries. We
228    # might try to preserve old data, but all-in-all it's a bit
229    # complicated.  One possible way of fixing this is to construct a
230    # custom FQL query that joins the friends table and the users table.
231}
232
233sub poll_facebook {
234    my $self = shift;
235
236    return unless BarnOwl::getvar('facebook:poll') eq 'on';
237    return unless $self->{logged_in};
238
239    # XXX Oh no! This blocks the user interface.  Not good.
240    # Ideally, we should have some worker thread for polling facebook.
241    # But BarnOwl is probably not thread-safe >_<
242
243    my $old_topics = $self->{topics};
244    $self->{topics} = {};
245
246    my $updates = eval {
247        $self->{facebook}
248             ->query
249             ->from("my_news")
250             # Not using this, because we want to pick up comment
251             # updates. We need to manually de-duplicate, though.
252             # ->where_since( "@" . $self->{last_poll} )
253             # Facebook doesn't actually give us that many results.
254             # But it can't hurt to ask!
255             ->limit_results( 200 )
256             ->request
257             ->as_hashref
258    };
259    return unless $self->check_result;
260
261    my $new_last_poll = $self->{last_poll};
262    for my $post ( reverse @{$updates->{data}} ) {
263        # No app invites, thanks! (XXX make configurable)
264        if ($post->{type} eq 'link' && $post->{application}) {
265            next;
266        }
267
268        # XXX Filtering out interest groups for now
269        # A more reasonable strategy may be to show their
270        # posts, but not the comments.
271        if (defined $post->{from}{category}) {
272            next;
273        }
274
275        # There can be multiple recipients! Strange! Pick the first one.
276        my $name    = $post->{to}{data}[0]{name} || $post->{from}{name};
277        my $name_id = $post->{to}{data}[0]{id} || $post->{from}{id};
278        my $post_id  = $post->{id};
279
280        my $topic;
281        if (defined $old_topics->{$post_id}) {
282            $topic = $old_topics->{$post_id};
283            $self->{topics}->{$post_id} = $topic;
284        } else {
285            my @keywords = keywords($post->{name} || $post->{message});
286            $topic = $keywords[0] || 'personal';
287            $topic =~ s/ /-/g;
288            $self->{topics}->{$post_id} = $topic;
289        }
290
291        # Only handle post if it's new
292        my $created_time = str2time($post->{created_time});
293        if ($created_time >= $self->{last_poll}) {
294            # XXX indexing is fragile
295            my $msg = BarnOwl::Message->new(
296                type      => 'Facebook',
297                sender    => $post->{from}{name},
298                sender_id => $post->{from}{id},
299                name      => $name,
300                name_id   => $name_id,
301                direction => 'in',
302                body      => $self->format_body($post),
303                post_id   => $post_id,
304                topic     => $topic,
305                time      => asctime(localtime $created_time),
306                # XXX The intent is to get the 'Comment' link, which also
307                # serves as a canonical link to the post.  The {name}
308                # field should equal 'Comment'.
309                permalink => $post->{actions}[0]{link},
310               );
311            BarnOwl::queue_message($msg);
312        }
313
314        # This will interleave times (they'll all be organized by parent
315        # post), but since we don't expect too many updates between
316        # polls this is pretty acceptable.
317        my $updated_time = str2time($post->{updated_time});
318        if ($updated_time >= $self->{last_poll} && defined $post->{comments}{data}) {
319            for my $comment ( @{$post->{comments}{data}} ) {
320                my $comment_time = str2time($comment->{created_time});
321                if ($comment_time < $self->{last_poll}) {
322                    next;
323                }
324                my $msg = BarnOwl::Message->new(
325                    type      => 'Facebook',
326                    sender    => $comment->{from}{name},
327                    sender_id => $comment->{from}{id},
328                    name      => $name,
329                    name_id   => $name_id,
330                    direction => 'in',
331                    body      => $comment->{message},
332                    post_id   => $post_id,
333                    topic     => $topic,
334                    time      => asctime(localtime $comment_time),
335                   );
336                BarnOwl::queue_message($msg);
337            }
338        }
339        if ($updated_time + 1 > $new_last_poll) {
340            $new_last_poll = $updated_time + 1;
341        }
342    }
343    # old_topics gets GC'd
344
345    $self->{last_poll} = $new_last_poll;
346}
347
348sub format_body {
349    my $self = shift;
350
351    my $post = shift;
352
353    # XXX implement optional URL minification
354    if ($post->{type} eq 'status') {
355        return $post->{message};
356    } elsif ($post->{type} eq 'link' || $post->{type} eq 'video' || $post->{type} eq 'photo') {
357        return $post->{name}
358          . ($post->{caption} ? " (" . $post->{caption} . ")\n" : "\n")
359          . $post->{link}
360          . ($post->{description} ? "\n\n" . $post->{description} : "")
361          . ($post->{message} ? "\n\n" . $post->{message} : "");
362    } else {
363        return "(unknown post type " . $post->{type} . ")";
364    }
365}
366
367# Invariant: we don't become logged out between entering text field
368# and actually processing the request.  XXX I don't think this actually
369# holds, but such a case would rarely happen.
370
371sub facebook {
372    my $self = shift;
373
374    my $user = shift;
375    my $msg = shift;
376
377    if (defined $user) {
378        $user = $self->{friends}{$user} || $user;
379        eval { $self->{facebook}->add_post( $user )->set_message( $msg )->publish; };
380        return unless $self->check_result;
381    } else {
382        eval { $self->{facebook}->add_post->set_message( $msg )->publish; };
383        return unless $self->check_result;
384    }
385    $self->sleep(0);
386}
387
388sub facebook_comment {
389    my $self = shift;
390
391    my $post_id = shift;
392    my $msg = shift;
393
394    eval { $self->{facebook}->add_comment( $post_id )->set_message( $msg )->publish; };
395    return unless $self->check_result;
396    $self->sleep(0);
397}
398
399sub facebook_auth {
400    my $self = shift;
401
402    my $url = shift;
403
404    if (!defined $url) {
405        $self->facebook_do_auth;
406        return;
407    }
408
409    # http://www.facebook.com/connect/login_success.html#access_token=TOKEN&expires_in=0
410    $url =~ /access_token=([^&]+)/; # XXX Ew regex
411
412    if (!defined $1) {
413        BarnOwl::message("Invalid URL.");
414        return;
415    }
416
417    $self->{cfg}->{token} = $1;
418    if ($self->facebook_do_auth) {
419        my $raw_cfg = to_json($self->{cfg});
420        BarnOwl::admin_message('Facebook', "Add this as the contents of your ~/.owl/facebook file:\n$raw_cfg");
421    }
422}
423
424sub facebook_do_auth {
425    my $self = shift;
426    if ( ! defined $self->{cfg}->{token} ) {
427        BarnOwl::admin_message('Facebook', "Login to Facebook at ".$self->{login_url}
428            . "\nand run command ':facebook-auth URL' with the URL you are redirected to."
429            . "\n\nWhat does Barnowl use these permissions for?  As a desktop"
430            . "\nmessaging application, we need persistent read/write access to your"
431            . "\nnews feed and your inbox.  Other permissions are for pending"
432            . "\nfeatures: we intend on adding support for event streaming, RSVP,"
433            . "\nand BarnOwl filtering on friend lists."
434        );
435        return 0;
436    }
437    $self->{facebook}->access_token($self->{cfg}->{token});
438    # Do a quick check to see if things are working
439    my $result = eval { $self->{facebook}->query()->find('me')->select_fields('name')->request->as_hashref; };
440    if ($@) {
441        BarnOwl::admin_message('Facebook', "Failed to authenticate! Login to Facebook at ".$self->{login_url}
442            . "\nand run command ':facebook-auth URL' with the URL you are redirected to.");
443        return 0;
444    } else {
445        my $name = $result->{'name'};
446        BarnOwl::admin_message('Facebook', "Successfully logged in to Facebook as $name!");
447        $self->{logged_in} = 1;
448        $self->sleep(0); # start polling
449        return 1;
450    }
451}
452
4531;
Note: See TracBrowser for help on using the repository browser.