Today we're going to answer a fun question: how do you turn a handful of isolated text messages to and from the same party into a true conversation?
The problem here? SMS is a stateless protocol. You're going to need some way to remember state between each exchanged message.
Luckily for us, building traditional web applications has this same hurdle, as HTTP is also a stateless protocol. This problem has been solved for web applications through the use of HTTP cookies and, rather than reinvent that old wheel, the Twilio Messaging API uses the same solution.
This guide will show you how to use Programmable Messaging to do this. We'll use Python, the Flask framework, and the Twilio Python SDK.
Twilio Conversations, a more recent product offering, is an omni-channel messaging platform that allows you to build engaging conversational, two-way messaging experiences. Be sure to check out our Conversations product to see if it's a better fit for your needs.
Just like in web applications, a cookie is a small file that your application can store on Twilio's servers to keep track of information, such as a username or account.
For Twilio Programmable Messaging, cookies are scoped to the "conversation" between two parties — you can have a unique cookie for each To/From phone number pair. For example, you can store a unique cookie for any messages sent between 415-555-2222 and 415-555-1111, which will be different than the cookie used between 415-555-3333 and 415-555-1111.
Why Should You Use a Cookie?
In a web app, you write a cookie to keep "statefulness" between separate requests from the same browser.
Similarly, SMS messages are independent communications between two parties, so Twilio allows you to tie them together as a logical session via cookies. This means you can use server-side sessions to keep track of application state between requests. (How cool is that?)
Twilio will expire the cookies for that conversation after four hours of inactivity, as if the user "closed the browser."
Storing Conversation Data
The cookies let you share state across multiple messages allowing you to treat separate messages as a conversation, and store data about the conversation in the cookies for future reference.
You can store the data directly in a cookie, or you can use a session state management framework. We'll use Flask's built-in session library for this example.
Track SMS Conversations Using a Session
Now let's try using session counters to see if a particular user has messaged us before. If they're a new sender, we'll thank them for the new message. If they've sent us messages before, we'll specify how many messages we've gotten from them.
Tracking SMS Conversations using Cookies
Python
_45
from flask import Flask, request, session
_45
from twilio.twiml.messaging_response import MessagingResponse
_45
_45
# The session object makes use of a secret key.
_45
SECRET_KEY = 'a secret key'
_45
app = Flask(__name__)
_45
app.config.from_object(__name__)
_45
_45
# Try adding your own number to this list!
_45
callers = {
_45
"+14158675308": "Rey",
_45
"+12349013030": "Finn",
_45
"+12348134522": "Chewy",
_45
}
_45
_45
_45
@app.route("/", methods=['GET', 'POST'])
_45
def hello():
_45
"""Respond with the number of text messages sent between two parties."""