instabotnet 1.0.18

Instagram readable yaml rpc api for easy instagram scheduling and promotion

instagram-botnet

Write readable declarative yaml files to control your botnet

TODO

• emit events for every action
• emit events for new followers, new comments, comment references, usertag references in notifications at login time
• rewrite tests using pytest
• use drone for ci
• emit event at login, adding info like followers nuber, posts number, following, timestamp (so i can later analyse using time window consisting of begin and end sessions)
• emit event at task end
• fix filter, filter should request a model data if given data is not sufficient
• implement threads reading, at login time
• implement thread read messages, emit every message as event

direct_v2_inbox() ->

Response:
    viewer: User
    inbox:
        threads: [
            thread_id: Str
            thread_v2_id: Str
            pending: Bool
            read_state: ReadState
            users: [User]
            items: [
                item_id: Int
                item_type: "media_share" | "text" | "link"
                timestamp: Int
                user_id: Int
                device_timestamp?: Int
                media_share?: Media
                text?: Str
                link?:
                    link_url: Str
            ]
            thread_type: private
            has_newer: Bool
            has_older: Bool
            last_seen_at: Any
            newest_cursor: Str
            oldest_cursor: Str
            is_spam: Bool
            last_activity_at: Timestamp
            last_seen_at:
                [user_id]:
                    timestamp: Timestamp
                    item_id: Int
        ]

        next_cursor:
            cursor_thread_v2_id: Int
            cursor_timestamp_seconds: Int
        unseen_count: Int
        unseen_count_ts: Int

    pending_requests_total: int
    status: "ok" | Str

ReadState: 0 | 1

direct_v2_thread(thread_id) # not thread_v2_id ->

Response:
    thread: Thread

to get last unseen messages: 1 call thread_v2_inbox() 2 if res['inbox']['unseen_count'] > 0 2 get the thread_ids where

            ->addParam('persistentBadging', 'true')
            ->addParam('use_unified_inbox', 'true');
    /**
     * Marks an item from given thread as seen.
     *
     * @param string $threadId     Thread ID.
     * @param string $threadItemId Thread item ID.
     *
     * @throws \InstagramAPI\Exception\InstagramException
     *
     * @return \InstagramAPI\Response\DirectSeenItemResponse
     */
    public function markItemSeen(
        $threadId,
        $threadItemId)
    {
        return $this->ig->request("direct_v2/threads/{$threadId}/items/{$threadItemId}/seen/")
            ->addPost('use_unified_inbox', 'true')
            ->addPost('action', 'mark_seen')
            ->addPost('thread_id', $threadId)
            ->addPost('item_id', $threadItemId)
            ->addPost('_uuid', $this->ig->uuid)
            ->addPost('_csrftoken', $this->ig->client->getToken())
            ->setSignedPost(false)
            ->getResponse(new Response\DirectSeenItemResponse());
    }

Shell usage

The main module works on yaml script like these:

name:                     test_comment

bot:
        username:         username
        password:         password

actions:
    -
        name: comment on 10 posts from @kimkardashian
        nodes:
            - kimkardashian
        edges:
            - user_feed:
                amount: 10
            - comment:
                max:      1
                comments:
                    - ["hello {author}!!!"]
                    - ["come stai?", "come va?"]
                    - ["url works too! http://instagram.com"]

To execute the above test_comment.yaml run:

python3 -m instabotnet test_comment.yaml

Python API usage

To use inside python modules:

from instabotnet import execute

execute('test_comment.yaml', {'username': 'user', 'some_variables': 'bo'})

Environment variables are accessible under the env object

The variables passed to execute can be mutated, the execute function returns an object that contains the data collected from he scrape action. The variables are mutated to support the multi script feature, where many templates are chained one after another in a single file and can pass data to the next template mutating the variables. It is also useful to mutate the settings variable to update the cookies and other data for the next bot iteration.

All the code inside {{ }} is evaluated by an eval call, this means that the templates can be filled with conditional and complex behaviors, for example making an http call to get the caption for uploading a photo, or get a random item from an array variable. Inside eval code it is avaliable

• variables
• random module
• all the funcy functions
• variables inside the data object
• urlopen module for making api calls
• json module A limitation of the code inside {{ }} is that it must consist only of one statement and can only return lists or strings.

Every action inside a template will emit an event, all these events are avaliable in the object returned by execute:

"template_name": "test_follow",
"action_name": "test_follow_famous_people",
"type": "follow",
"node": "kimkarkdashian",
"args": {},
"metadata": { 
    "username": "",
    "proxy": ""
}

Other events are emitted, for example when someone followes you:

"type": "got_followed",
"node": "username"