python-intercom

Python bindings for the Intercom API (https://api.intercom.io).

Upgrading information

Version 2 of python-intercom is not backwards compatible with previous versions.

One change you will need to make as part of the upgrade is to set Intercom.app_api_key and not set Intercom.api_key.

Installation

pip install python-intercom

Basic Usage

Configure your access credentials

Intercom.app_id = "my_app_id"
Intercom.app_api_key = "my-super-crazy-api-key"

Resources

Resources this API supports:

https://api.intercom.io/users
https://api.intercom.io/companies
https://api.intercom.io/tags
https://api.intercom.io/notes
https://api.intercom.io/segments
https://api.intercom.io/events
https://api.intercom.io/conversations
https://api.intercom.io/messages
https://api.intercom.io/counts
https://api.intercom.io/subscriptions

Additionally, the library can handle incoming webhooks from Intercom and convert to intercom models.

Examples

Users

from intercom import User
# Find user by email
user = User.find(email="[email protected]")
# Find user by user_id
user = User.find(user_id="1")
# Find user by id
user = User.find(id="1")
# Create a user
user = User.create(email="[email protected]", name="Bob Smith")
# Delete a user
deleted_user = User.find(id="1").delete()
# Update custom_attributes for a user
user.custom_attributes["average_monthly_spend"] = 1234.56
user.save()
# Perform incrementing
user.increment('karma')
user.save()
# Iterate over all users
for user in User.all():
    ...

Admins

from intercom import Admin
# Iterate over all admins
for admin in Admin.all():
    ...

Companies

from intercom import Company
from intercom import User
# Add a user to one or more companies
user = User.find(email="[email protected]")
user.companies = [
    {"company_id": 6, "name": "Intercom"},
    {"company_id": 9, "name": "Test Company"}
]
user.save()
# You can also pass custom attributes within a company as you do this
user.companies = [
    {
        "id": 6,
        "name": "Intercom",
        "custom_attributes": {
            "referral_source": "Google"
        }
    }
]
user.save()
# Find a company by company_id
company = Company.find(company_id="44")
# Find a company by name
company = Company.find(name="Some company")
# Find a company by id
company = Company.find(id="41e66f0313708347cb0000d0")
# Update a company
company.name = 'Updated company name'
company.save()
# Iterate over all companies
for company in Company.all():
    ...
# Get a list of users in a company
company.users

Segments

from intercom import Segment
# Find a segment
segment = Segment.find(id=segment_id)
# Update a segment
segment.name = 'Updated name'
segment.save()
# Iterate over all segments
for segment in Segment.all():
    ...

Notes

# Find a note by id
note = Note.find(id=note)
# Create a note for a user
note = Note.create(
    body="<p>Text for the note</p>",
    email='[email protected]')
# Iterate over all notes for a user via their email address
for note in Note.find_all(email='[email protected]'):
    ...
# Iterate over all notes for a user via their user_id
for note in Note.find_all(user_id='123'):
    ...

Conversations

from intercom import Conversation
# FINDING CONVERSATIONS FOR AN ADMIN
# Iterate over all conversations (open and closed) assigned to an admin
for convo in Conversation.find_all(type='admin', id='7'):
    ...
# Iterate over all open conversations assigned to an admin
for convo Conversation.find_all(type='admin', id=7, open=True):
    ...
# Iterate over closed conversations assigned to an admin
for convo Conversation.find_all(type='admin', id=7, open=False):
    ...
# Iterate over closed conversations for assigned an admin, before a certain
# moment in time
for convo in Conversation.find_all(
        type='admin', id= 7, open= False, before=1374844930):
    ...

# FINDING CONVERSATIONS FOR A USER
# Iterate over all conversations (read + unread, correct) with a user based on
# the users email
for convo in Conversation.find_all(email='[email protected]',type='user'):
    ...
# Iterate over through all conversations (read + unread) with a user based on
# the users email
for convo in Conversation.find_all(
        email='[email protected]', type='user', unread=False):
    ...
# Iterate over all unread conversations with a user based on the users email
for convo in Conversation.find_all(
        email='[email protected]', type='user', unread=true):
    ...

# FINDING A SINGLE CONVERSATION
conversation = Conversation.find(id='1')

# INTERACTING WITH THE PARTS OF A CONVERSATION
# Getting the subject of a part (only applies to email-based conversations)
conversation.rendered_message.subject
# Get the part_type of the first part
conversation.conversation_parts[0].part_type
# Get the body of the second part
conversation.conversation_parts[1].body

# REPLYING TO CONVERSATIONS
# User (identified by email) replies with a comment
conversation.reply(
    type='user', email='[email protected]',
    message_type= comment', body='foo')
# Admin (identified by email) replies with a comment
conversation.reply(
    type='admin', email='[email protected]',
    message_type='comment', body='bar')
# Admin (identified by id) opens a conversation
conversation.open_conversation(admin_id=7)
# Admin (identified by id) closes a conversation
conversation.close_conversation(admin_id=7)
# Admin (identified by id) assigns a conversation to an assignee
conversation.assign(assignee_id=8, admin_id=7)

# MARKING A CONVERSATION AS READ
conversation.read = True
conversation.save()

Counts

from intercom import Count
# Get Conversation per Admin
conversation_counts_for_each_admin = Count.conversation_counts_for_each_admin()
for count in conversation_counts_for_each_admin:
    print "Admin: %s (id: %s) Open: %s Closed: %s" % (
        count.name, count.id, count.open, count.closed)
# Get User Tag Count Object
Count.user_counts_for_each_tag()
# Get User Segment Count Object
Count.user_counts_for_each_segment()
# Get Company Segment Count Object
Count.company_counts_for_each_segment()
# Get Company Tag Count Object
Count.company_counts_for_each_tag()
# Get Company User Count Object
Count.company_counts_for_each_user()
# Get total count of companies, users, segments or tags across app
Company.count()
User.count()
Segment.count()
Tag.count()

Full loading of and embedded entity

    # Given a converation with a partial user, load the full user. This can be done for any entity
    conversation.user.load()

Sending messages

# InApp message from admin to user
Message.create(**{
    "message_type": "inapp",
    "body": "What's up :)",
    "from": {
        "type": "admin",
        "id": "1234"
    },
    "to": {
        "type": "user",
        "id": "5678"
    }
})

# Email message from admin to user
Message.create(**{
    "message_type": "email",
    "subject": "Hey there",
    "body": "What's up :)",
    "template": "plain", # or "personal",
    "from": {
        "type": "admin",
        "id": "1234"
    },
    "to": {
        "type": "user",
        "id": "536e564f316c83104c000020"
    }
})

# Message from a user
Message.create(**{
    "from": {
        "type": "user",
        "id": "536e564f316c83104c000020"
    },
    "body": "halp"
})

Events

from intercom import Event
Event.create(
    event_name="invited-friend",
    created_at=time.mktime(),
    email=user.email,
    metadata={
        "invitee_email": "[email protected]",
        "invite_code": "ADDAFRIEND",
        "found_date": 12909364407
    }
)

Metadata Objects support a few simple types that Intercom can present on your behalf

Event.create(
    event_name="placed-order",
    email=current_user.email,
    created_at=1403001013
    metadata={
        "order_date": time.mktime(),
        "stripe_invoice": 'inv_3434343434',
        "order_number": {
            "value": '3434-3434',
            "url": 'https://example.org/orders/3434-3434'
        },
        "price": {
            "currency": 'usd',
            "amount": 2999
        }
    }
)

The metadata key values in the example are treated as follows-

order_date: a Date (key ends with '_date').
stripe_invoice: The identifier of the Stripe invoice (has a 'stripe_invoice' key)
order_number: a Rich Link (value contains 'url' and 'value' keys)
price: An Amount in US Dollars (value contains 'amount' and 'currency' keys)

Subscriptions

Subscribe to events in Intercom to receive webhooks.

from intercom import Subscription
# create a subscription
Subscription.create(url="http://example.com", topics=["user.created"])

# fetch a subscription
Subscription.find(id="nsub_123456789")

# list subscriptions
Subscription.all():

Webhooks

from intercom import Notification
# create a payload from the notification hash (from json).
payload = Intercom::Notification.new(notification_hash)

payload.type
# 'user.created'

payload.model_type
# User

user = payload.model
# Instance of User

Note that models generated from webhook notifications might differ slightly from models directly acquired via the API. If this presents a problem, calling payload.load will load the model from the API using the id field.

Errors

You do not need to deal with the HTTP response from an API call directly. If there is an unsuccessful response then an error that is a subclass of intercom.Error will be raised. If desired, you can get at the http_code of an Error via it's http_code method.

The list of different error subclasses are listed below. As they all inherit off IntercomError you can choose to except IntercomError or the more specific error subclass:

AuthenticationError
ServerError
ServiceUnavailableError
ResourceNotFound
BadGatewayError
BadRequestError
RateLimitExceeded
MultipleMatchingUsersError
HttpError
UnexpectedError

Rate Limiting

Calling Intercom.rate_limit_details returns a dict that contains details about your app's current rate limit.

Intercom.rate_limit_details
# {'limit': 500, 'reset_at': datetime.datetime(2015, 3, 28, 13, 22), 'remaining': 497}

Running the Tests

Unit tests:

nosetests tests/unit

Integration tests:

INTERCOM_APP_ID=xxx INTERCOM_APP_API_KEY=xxx nosetests tests/integration

bdziendzik / python-intercom Goto Github PK

python-intercom's Introduction