summaryrefslogtreecommitdiffstats
path: root/activities/models/post.py
blob: 378de9183ca984eeed20647943ffaffb40332ed6 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
import re
from typing import Dict, Optional

import httpx
import urlman
from django.db import models, transaction
from django.template.defaultfilters import linebreaks_filter
from django.utils import timezone
from django.utils.safestring import mark_safe

from activities.models.fan_out import FanOut
from activities.models.timeline_event import TimelineEvent
from core.html import sanitize_post, strip_html
from core.ld import canonicalise, format_ld_date, get_list, parse_ld_date
from stator.models import State, StateField, StateGraph, StatorModel
from users.models.follow import Follow
from users.models.identity import Identity


class PostStates(StateGraph):
    new = State(try_interval=300)
    fanned_out = State()

    new.transitions_to(fanned_out)

    @classmethod
    async def handle_new(cls, instance: "Post"):
        """
        Creates all needed fan-out objects for a new Post.
        """
        post = await instance.afetch_full()
        # Non-local posts should not be here
        # TODO: This seems to keep happening. Work out how?
        if not post.local:
            print(f"Trying to run handle_new on a non-local post {post.pk}!")
            return cls.fanned_out
        # Build list of targets - mentions always included
        targets = set()
        async for mention in post.mentions.all():
            targets.add(mention)
        # Then, if it's not mentions only, also deliver to followers
        if post.visibility != Post.Visibilities.mentioned:
            async for follower in post.author.inbound_follows.select_related("source"):
                targets.add(follower.source)
        # Fan out to each one
        for follow in targets:
            await FanOut.objects.acreate(
                identity=follow,
                type=FanOut.Types.post,
                subject_post=post,
            )
        # And one for themselves if they're local
        if post.author.local:
            await FanOut.objects.acreate(
                identity_id=post.author_id,
                type=FanOut.Types.post,
                subject_post=post,
            )
        return cls.fanned_out


class Post(StatorModel):
    """
    A post (status, toot) that is either local or remote.
    """

    class Visibilities(models.IntegerChoices):
        public = 0
        unlisted = 1
        followers = 2
        mentioned = 3

    # The author (attributedTo) of the post
    author = models.ForeignKey(
        "users.Identity",
        on_delete=models.CASCADE,
        related_name="posts",
    )

    # The state the post is in
    state = StateField(PostStates)

    # If it is our post or not
    local = models.BooleanField()

    # The canonical object ID
    object_uri = models.CharField(max_length=500, blank=True, null=True, unique=True)

    # Who should be able to see this Post
    visibility = models.IntegerField(
        choices=Visibilities.choices,
        default=Visibilities.public,
    )

    # The main (HTML) content
    content = models.TextField()

    # If the contents of the post are sensitive, and the summary (content
    # warning) to show if it is
    sensitive = models.BooleanField(default=False)
    summary = models.TextField(blank=True, null=True)

    # The public, web URL of this Post on the original server
    url = models.CharField(max_length=500, blank=True, null=True)

    # The Post it is replying to as an AP ID URI
    # (as otherwise we'd have to pull entire threads to use IDs)
    in_reply_to = models.CharField(max_length=500, blank=True, null=True)

    # The identities the post is directly to (who can see it if not public)
    to = models.ManyToManyField(
        "users.Identity",
        related_name="posts_to",
        blank=True,
    )

    # The identities mentioned in the post
    mentions = models.ManyToManyField(
        "users.Identity",
        related_name="posts_mentioning",
        blank=True,
    )

    # Hashtags in the post
    hashtags = models.JSONField(blank=True, null=True)

    # When the post was originally created (as opposed to when we received it)
    published = models.DateTimeField(default=timezone.now)

    # If the post has been edited after initial publication
    edited = models.DateTimeField(blank=True, null=True)

    created = models.DateTimeField(auto_now_add=True)
    updated = models.DateTimeField(auto_now=True)

    class urls(urlman.Urls):
        view = "{self.author.urls.view}posts/{self.id}/"
        object_uri = "{self.author.actor_uri}posts/{self.id}/"
        action_like = "{view}like/"
        action_unlike = "{view}unlike/"
        action_boost = "{view}boost/"
        action_unboost = "{view}unboost/"

        def get_scheme(self, url):
            return "https"

        def get_hostname(self, url):
            return self.instance.author.domain.uri_domain

    def __str__(self):
        return f"{self.author} #{self.id}"

    def get_absolute_url(self):
        return self.urls.view

    def absolute_object_uri(self):
        """
        Returns an object URI that is always absolute, for sending out to
        other servers.
        """
        if self.local:
            return self.author.absolute_profile_uri() + f"posts/{self.id}/"
        else:
            return self.object_uri

    ### Content cleanup and extraction ###

    mention_regex = re.compile(
        r"(^|[^\w\d\-_])@([\w\d\-_]+(?:@[\w\d\-_]+\.[\w\d\-_\.]+)?)"
    )

    def linkify_mentions(self, content, local=False):
        """
        Links mentions _in the context of the post_ - as in, using the mentions
        property as the only source (as we might be doing this without other
        DB access allowed)
        """

        possible_matches = {}
        for mention in self.mentions.all():
            if local:
                url = str(mention.urls.view)
            else:
                url = mention.absolute_profile_uri()
            possible_matches[mention.username] = url
            possible_matches[f"{mention.username}@{mention.domain_id}"] = url

        def replacer(match):
            precursor = match.group(1)
            handle = match.group(2)
            if handle in possible_matches:
                return f'{precursor}<a href="{possible_matches[handle]}">@{handle}</a>'
            else:
                return match.group()

        return mark_safe(self.mention_regex.sub(replacer, content))

    def safe_content_local(self):
        """
        Returns the content formatted for local display
        """
        return self.linkify_mentions(sanitize_post(self.content), local=True)

    def safe_content_remote(self):
        """
        Returns the content formatted for remote consumption
        """
        return self.linkify_mentions(sanitize_post(self.content))

    ### Async helpers ###

    async def afetch_full(self):
        """
        Returns a version of the object with all relations pre-loaded
        """
        return (
            await Post.objects.select_related("author", "author__domain")
            .prefetch_related("mentions", "mentions__domain")
            .aget(pk=self.pk)
        )

    ### Local creation ###

    @classmethod
    def create_local(
        cls,
        author: Identity,
        content: str,
        summary: Optional[str] = None,
        visibility: int = Visibilities.public,
    ) -> "Post":
        with transaction.atomic():
            # Find mentions in this post
            mention_hits = cls.mention_regex.findall(content)
            mentions = set()
            for precursor, handle in mention_hits:
                if "@" in handle:
                    username, domain = handle.split("@", 1)
                else:
                    username = handle
                    domain = author.domain_id
                identity = Identity.by_username_and_domain(
                    username=username,
                    domain=domain,
                    fetch=True,
                )
                if identity is not None:
                    mentions.add(identity)
            # Strip all HTML and apply linebreaks filter
            content = linebreaks_filter(strip_html(content))
            # Make the Post object
            post = cls.objects.create(
                author=author,
                content=content,
                summary=summary or None,
                sensitive=bool(summary),
                local=True,
                visibility=visibility,
            )
            post.object_uri = post.urls.object_uri
            post.url = post.absolute_object_uri()
            post.mentions.set(mentions)
            post.save()
        return post

    ### ActivityPub (outbound) ###

    def to_ap(self) -> Dict:
        """
        Returns the AP JSON for this object
        """
        value = {
            "to": "as:Public",
            "cc": [],
            "type": "Note",
            "id": self.object_uri,
            "published": format_ld_date(self.published),
            "attributedTo": self.author.actor_uri,
            "content": self.safe_content_remote(),
            "as:sensitive": self.sensitive,
            "url": self.absolute_object_uri(),
            "tag": [],
        }
        if self.summary:
            value["summary"] = self.summary
        # Mentions
        for mention in self.mentions.all():
            value["tag"].append(
                {
                    "href": mention.actor_uri,
                    "name": "@" + mention.handle,
                    "type": "Mention",
                }
            )
            value["cc"].append(mention.actor_uri)
        # Remove tag and cc if they're empty
        if not value["cc"]:
            del value["cc"]
        if not value["tag"]:
            del value["tag"]
        return value

    def to_create_ap(self):
        """
        Returns the AP JSON to create this object
        """
        object = self.to_ap()
        return {
            "to": object["to"],
            "cc": object.get("cc", []),
            "type": "Create",
            "id": self.object_uri + "#create",
            "actor": self.author.actor_uri,
            "object": object,
        }

    ### ActivityPub (inbound) ###

    @classmethod
    def by_ap(cls, data, create=False, update=False) -> "Post":
        """
        Retrieves a Post instance by its ActivityPub JSON object.

        Optionally creates one if it's not present.
        Raises KeyError if it's not found and create is False.
        """
        # Do we have one with the right ID?
        created = False
        try:
            post = cls.objects.get(object_uri=data["id"])
        except cls.DoesNotExist:
            if create:
                # Resolve the author
                author = Identity.by_actor_uri(data["attributedTo"], create=create)
                post = cls.objects.create(
                    object_uri=data["id"],
                    author=author,
                    content=data["content"],
                    local=False,
                )
                created = True
            else:
                raise KeyError(f"No post with ID {data['id']}", data)
        if update or created:
            post.content = data["content"]
            post.summary = data.get("summary")
            post.sensitive = data.get("as:sensitive", False)
            post.url = data.get("url")
            post.published = parse_ld_date(data.get("published"))
            post.edited = parse_ld_date(data.get("updated"))
            post.in_reply_to = data.get("inReplyTo")
            # Mentions and hashtags
            post.hashtags = []
            for tag in get_list(data, "tag"):
                if tag["type"].lower() == "mention":
                    mention_identity = Identity.by_actor_uri(tag["href"], create=True)
                    post.mentions.add(mention_identity)
                elif tag["type"].lower() == "as:hashtag":
                    post.hashtags.append(tag["name"].lstrip("#"))
                elif tag["type"].lower() == "http://joinmastodon.org/ns#emoji":
                    # TODO: Handle incoming emoji
                    pass
                else:
                    raise ValueError(f"Unknown tag type {tag['type']}")
            # Visibility and to
            # (a post is public if it's ever to/cc as:Public, otherwise we
            # regard it as unlisted for now)
            targets = get_list(data, "to") + get_list(data, "cc")
            post.visibility = Post.Visibilities.unlisted
            for target in targets:
                if target.lower() == "as:public":
                    post.visibility = Post.Visibilities.public
            # Attachments
            # These have no IDs, so we have to wipe them each time
            post.attachments.all().delete()
            for attachment in get_list(data, "attachment"):
                if "http://joinmastodon.org/ns#focalPoint" in attachment:
                    focal_x, focal_y = attachment[
                        "http://joinmastodon.org/ns#focalPoint"
                    ]["@list"]
                else:
                    focal_x, focal_y = None, None
                post.attachments.create(
                    remote_url=attachment["url"],
                    mimetype=attachment["mediaType"],
                    name=attachment.get("name"),
                    width=attachment.get("width"),
                    height=attachment.get("height"),
                    blurhash=attachment.get("http://joinmastodon.org/ns#blurhash"),
                    focal_x=focal_x,
                    focal_y=focal_y,
                )
            post.save()
        return post

    @classmethod
    def by_object_uri(cls, object_uri, fetch=False):
        """
        Gets the post by URI - either looking up locally, or fetching
        from the other end if it's not here.
        """
        try:
            return cls.objects.get(object_uri=object_uri)
        except cls.DoesNotExist:
            if fetch:
                # Go grab the data from the URI
                response = httpx.get(
                    object_uri,
                    headers={"Accept": "application/json"},
                    follow_redirects=True,
                )
                if 200 <= response.status_code < 300:
                    return cls.by_ap(
                        canonicalise(response.json(), include_security=True),
                        create=True,
                        update=True,
                    )
            raise cls.DoesNotExist(f"Cannot find Post with URI {object_uri}")

    @classmethod
    def handle_create_ap(cls, data):
        """
        Handles an incoming create request
        """
        with transaction.atomic():
            # Ensure the Create actor is the Post's attributedTo
            if data["actor"] != data["object"]["attributedTo"]:
                raise ValueError("Create actor does not match its Post object", data)
            # Create it
            post = cls.by_ap(data["object"], create=True, update=True)
            # Make timeline events for followers if it's not a reply
            # TODO: _do_ show replies to people we follow somehow
            if not post.in_reply_to:
                for follow in Follow.objects.filter(
                    target=post.author, source__local=True
                ):
                    TimelineEvent.add_post(follow.source, post)
            # Make timeline events for mentions if they're local
            for mention in post.mentions.all():
                if mention.local:
                    TimelineEvent.add_mentioned(mention, post)
            # Force it into fanned_out as it's not ours
            post.transition_perform(PostStates.fanned_out)

    @classmethod
    def handle_update_ap(cls, data):
        """
        Handles an incoming update request
        """
        with transaction.atomic():
            # Ensure the Create actor is the Post's attributedTo
            if data["actor"] != data["object"]["attributedTo"]:
                raise ValueError("Create actor does not match its Post object", data)
            # Find it and update it
            cls.by_ap(data["object"], create=False, update=True)

    @classmethod
    def handle_delete_ap(cls, data):
        """
        Handles an incoming create request
        """
        with transaction.atomic():
            # Find our post by ID if we have one
            try:
                post = cls.by_object_uri(data["object"]["id"])
            except cls.DoesNotExist:
                # It's already been deleted
                return
            # Ensure the actor on the request authored the post
            if not post.author.actor_uri == data["actor"]:
                raise ValueError("Actor on delete does not match object")
            post.delete()

    def debug_fetch(self):
        """
        Fetches the Post from its original URL again and updates us with it
        """
        response = httpx.get(
            self.object_uri,
            headers={"Accept": "application/json"},
            follow_redirects=True,
        )
        if 200 <= response.status_code < 300:
            return self.by_ap(
                canonicalise(response.json(), include_security=True),
                update=True,
            )