How can I access to raw text the lead to the intent in lex
How can I extract the entire conversation including the user input and the Lex responses
I thought of creating a lambda & API gateway to capture the input and record it before sending it to Lex , and on each intent lambda record the response , but seems like a round about way.
In the event which you receive in Lambda function, you can get raw text from inputTranscript
{'messageVersion': '1.0', 'invocationSource': 'DialogCodeHook', 'userId': '', 'sessionAttributes': {}, 'requestAttributes': {}, 'bot':
{'name': 'bot_name', 'alias': 'bot_alias', 'version': '$LATEST'},
'outputDialogMode': 'Text', 'currentIntent': {'name':
'invoked_intent_name', 'slots': {}, 'slotDetails': {},
'confirmationStatus': 'None'}, 'inputTranscript': 'user message which
triggered the intent'}
For storing the conversation, in the Lambda function, just before you provide the response to the user, you can write a function to store the conversation i.e user query and response from bot.
Hope it helps.
The raw text that matched your intent can be reference as event.inputTranscript The full event format is documented here.
Amazon has a great example of how to build out a transcript of the entire conversation. Take a look at the RideService example code. Pay attention to how appendTranscript is used to build the full transcript of the interaction.
Related
I've created a Whatsapp chat bot that collects various user information but I am unable to collect an image that a user sends. How can I do this? Is it a matter of using the right Field type, which I've tried doing but none of the default field's apply to images? Please help if anyone knows of a solution.
Heyooo. 👋 Twilio Developer Evangelist here.
If a User sends an image via Whatsapp the image URL will be available in the sent webhook. You can have a look at the payload the webhook includes:
body: {
MediaContentType0: 'image/jpeg',
SmsMessageSid: 'MM9...',
NumMedia: '1',
SmsSid: 'MM9...',
SmsStatus: 'received',
Body: '',
To: 'whatsapp:+141...',
NumSegments: '1',
MessageSid: 'MM9bc...',
AccountSid: 'ACa34...',
From: 'whatsapp:+49176...',
MediaUrl0: 'https://api.twilio.com/2010-04-01/Accounts/ACa34bb5d3c305d08ae1308786f4d79b72/Messages/MM9bc3...',
ApiVersion: '2010-04-01'
}
You'll find the NumMedia and MediaUrl0 property which includes the URL of the sent image. You can then download these images and do whatever you like with them.
To retrieve the image after the message and webhook were sent you can have a look at the MediaResource Docs. You can fetch media also programmatically with something along the following lines:
client.messages('MM...')
.media('ME...')
.fetch()
.then(media => console.log(media.contentType));
In case you're using Studio you can have a look at this tutorial which handles Whatsapp Media with a fun use case.
Let me know if that helps. 😊
(It's hard to give more advice because I'm not sure what you're trying to do.)
Background:
I am using the python slack API (slackclient) to build an iterative sequence of data-gathering actions in ephemeral messages.
The core of this works fine. After processing the incoming request that contains the user's interaction with a set of message buttons (or menus), I respond immediately with a JSON body, as described in the "Responding right away" section of the official slack docs.
The problem:
Every response replaces the preceding message+attachments. In many cases, this is what I want, but there are situations where I want to add a response rather than replace the previous message.
Per the slack docs,setting replace_original to false should do just that. But the following code, snipped from my handling of a simple button click (for example), replaces the original button (and the text message to which it was attached):
r = {
'response_type': 'ephemeral',
'text': 'foo',
'replace_original': 'false'
}
log.debug("Returning: {}".format(json.dumps(r)))
resp = Response(response=json.dumps(r),
mimetype="application/json",
status=200)
return resp
I have tried this with and without the delete_original and response_type fields, with no change.
In short, it appears that in this case the replace_original field isn't doing anything at all; behavior is always as if it were set to 'true'.
I feel like I must be missing something here - any help is greatly appreciated.
Simple solution here: the slack API is expecting a boolean, not a string. So 'replace_original': 'false' in the above snippet ends up as {"response_type": "ephemeral", "text": "foo", "replace_original": "false"} after the json.dumps() call, which is invalid.
Instead, setting 'replace_original': False becomes {"response_type": "ephemeral", "text": "foo", "replace_original": false}, which then has the expected behavior
I cannot consistently successfully send form variables that may/may not include special characters e.g. ? & #
Depending on where I try to escape the chars I encounter different errors when reading the data server-side.
I am aware that an update is due for React Native 0.7 to include formdata but wondered if I could safely post objects without needing this.
Someone has already posted a similar issue but no example code was posted to illustrate the POST working:
How to post a form using fetch in react native?
I have tried - amongst other things :
fetch(APIURL, {
method: 'POST',
body: JSON.stringify({
object1: {
param1a: "value 1a",
param1b: "value 1b - with bad chars & # ?",
},
object2:
{
param2a: "value 2a",
param2b: 0,
}
})
})
but it groups the data into a single unnamed parameter (changing the API to accept this is not an option).
also this:
fetch(APIURL, {
method: 'GET',
accessPackage: JSON.stringify({
accessToken: "abc123",
tokenType: 2,
}),
taggData: JSON.stringify({
title: "test",
wishlistID: 0,
anotherVar: "anotherVal"
})
})
I wish to receive the data as two strings that can be parsed as json objects at the other end.
Looking at the the fetch repo https://github.com/github/fetch hasn't helped as this assumes the post with be a full JSON post (which it isn't) or uses FormData which isn't available to React Native yet.
Another solution may be to safely encode/serialise all of the data to URL parameters but this has also proven inconsistent so far especially with the # char.
What's the best way to do this?
"it groups the data into a single unnamed parameter (changing the API
to accept this is not an option)."
It would, because you've set the post body. This is how it's supposed to work.
I wish to receive the data as two strings that can be parsed as json objects at the other end.
You can do whatever you want, but there's no magic happening here. You will receive a single string, the string you set body to. Likewise, the post body can contain anything but shouldn't get confused with "special" characters. Most likely it is your server-side that is causing the problems.
If you want to use FormData then I think it'll be in v0.7.0 which should be out any day now, or you could probably just include the JS file in your own project. You can find it here. Usage examples are in the UIExplorer demo.
I've read the whole Box.com developers api guide and spent hours on the web researching this particular question but I can't seem to find a definitive answer and I don't want to start creating a solution if I'm going down the wrong path. We have a production environment where as once we are finished working with files our production software system zips them up and saves them into a local server directory for archival purposes. This local path cannot be changed. My question is how can I programmatically upload these files to our Box.com account so we can archive these on the cloud? Everything I've read regarding this involves using OAuth2 to gain access to our account which I understand but it also requires the user to login. Since this is an internal process that is NOT exposed to outside users I want to be able to automate this otherwise it would not be feasable for us. I have no issues creating the programs to trigger everytime a new files gets saved all I need is to streamline the Box.com access.
I just went through the exact same set of questions and found out that currently you CANNOT bypass the OAuth process. However, their refresh token is now valid for 60 days which should make any custom setup a bit more sturdy. I still think, though, that having to use OAuth for an Enterprise setup is a very brittle implementation -- for the exact reason you stated: it's not feasible for some middleware application to have to rely on an OAuth authentication process.
My Solution:
Here's what I came up with. The following are the same steps as outlined in various box API docs and videos:
use this URL https://www.box.com/api/oauth2/authorize?response_type=code&client_id=[YOUR_CLIENT_ID]&state=[box-generated_state_security_token]
(go to https://developers.box.com/oauth/ to find the original one)
paste that URL into the browser and GO
authenticate and grant access
grab the resulting URL: http://0.0.0.0/?state=[box-generated_state_security_token]&code=[SOME_CODE]
and note the "code=" value.
open POSTMAN or Fiddler (or some other HTTP sniffer) and enter the following:
URL: https://www.box.com/api/oauth2/token
create URL encoded post data:
grant_type=authorization_code
client_id=[YOUR CLIENT ID]
client_secret=[YOUR CLIENT SECRET]
code= < enter the code from step 4 >
send the request and retrieve the resulting JSON data:
{
"access_token": "[YOUR SHINY NEW ACCESS TOKEN]",
"expires_in": 4255,
"restricted_to": [],
"refresh_token": "[YOUR HELPFUL REFRESH TOKEN]",
"token_type": "bearer"
}
In my application I save both auth token and refresh token in a format where I can easily go and replace them if something goes awry down the road. Then, I check my authentication each time I call into the API. If I get an authorization exception back I refresh my token programmatically, which you can do! Using the BoxApi.V2 .NET SDK this happens like so:
var authenticator = new TokenProvider(_clientId, _clientSecret);
// calling the 'RefreshAccessToken' method in the SDK
var newAuthToken = authenticator.RefreshAccessToken([YOUR EXISTING REFRESH TOKEN]);
// write the new token back to my data store.
Save(newAuthToken);
Hope this helped!
If I understand correctly you want the entire process to be automated so it would not require a user login (i.e run a script and the file is uploaded).
Well, it is possible. I am a rookie developer so excuse me if I'm not using the correct terms.
Anyway, this can be accomplished by using cURL.
First you need to define some variables, your user credentials (username and password), your client id and client secret given by Box (found in your app), your redirect URI and state (used for extra safety if I understand correctly).
The oAuth2.0 is a 4 step authentication process and you're going to need to go through each step individually.
The first step would be setting a curl instance:
curl_setopt_array($curl, array(
CURLOPT_URL => "https://app.box.com/api/oauth2/authorize",
CURLOPT_RETURNTRANSFER => true,
CURLOPT_ENCODING => "content-type: application/x-www-form-urlencoded",
CURLOPT_MAXREDIRS => 10,
CURLOPT_TIMEOUT => 30,
CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
CURLOPT_CUSTOMREQUEST => "POST",
CURLOPT_POSTFIELDS =>
"response_type=code&client_id=".$CLIENT_ID."&state=".$STATE,
));
This will return an html text with a request token, you will need it for the next step so I would save the entire output to a variable and grep the tag with the request token (the tag has a "name" = "request_token" and a "value" which is the actual token).
Next step you will need to send another curl request to the same url, this time the post fields should include the request token, user name and password as follows:
CURLOPT_POSTFIELDS => "response_type=code&client_id=".$CLIENT_ID."&state=".$STATE."&request_token=".$REQ_TOKEN."&login=".$USER_LOGIN."&password=".$PASSWORD
At this point you should also set a cookie file:
CURLOPT_COOKIEFILE => $COOKIE, (where $COOKIE is the path to the cookie file)
This will return another html text output, use the same method to grep the token which has the name "ic".
For the next step you're going to need to send a post request to the same url. It should include the postfields:
response_type=code&client_id=".$CLIENT_ID."&state=".$STATE."&redirect_uri=".$REDIRECT_URI."&doconsent=doconsent&scope=root_readwrite&ic=".$IC
Be sure to set the curl request to use the cookie file you set earlier like this:
CURLOPT_COOKIEFILE => $COOKIE,
and include the header in the request:
CURLOPT_HEADER => true,
At step (if done by browser) you will be redirected to a URL which looks as described above:
http://0.0.0.0(*redirect uri*)/?state=[box-generated_state_security_token]&code=[SOME_CODE] and note the "code=" value.
Grab the value of "code".
Final step!
send a new cur request to https//app.box.com/api/oauth2/token
This should include fields:
CURLOPT_POSTFIELDS => "grant_type=authorization_code&code=".$CODE."&client_id=".$CLIENT_ID."&client_secret=".$CLIENT_SECRET,
This will return a string containing "access token", "Expiration" and "Refresh token".
These are the tokens needed for the upload.
read about the use of them here:
https://box-content.readme.io/reference#upload-a-file
Hope this is somewhat helpful.
P.S,
I separated the https on purpuse (Stackoverflow wont let me post an answer with more than 1 url :D)
this is for PHP cURL. It is also possible to do the same using Bash cURL.
For anyone looking into this recently, the best way to do this is to create a Limited Access App in Box.
This will let you create an access token which you can use for server to server communication. It's simple to then upload a file (example in NodeJS):
import box from "box-node-sdk";
import fs from "fs";
(async function (){
const client = box.getBasicClient(YOUR_ACCESS_TOKEN);
await client.files.uploadFile(BOX_FOLDER_ID, FILE_NAME, fs.createReadStream(LOCAL_FILE_PATH));
})();
Have you thought about creating a box 'integration' user for this particular purpose. It seems like uploads have to be made with a Box account. It sounds like you are trying to do an anonymous upload. I think box, like most services, including stackoverflow don't want anonymous uploads.
You could create a system user. Go do the Oauth2 dance and store just the refresh token somewhere safe. Then as the first step of your script waking up go use the refresh token and store the new refresh token. Then upload all your files.
I have a rails application which processes incoming emails via IMAP. Currently a method is used that searches the parts of a TMail object for a given content_type:
def self.search_parts_for_content_type(parts, content_type = 'text/html')
parts.each do |part|
if part.content_type == content_type
return part.body
else
if part.multipart?
if body = self.search_parts_for_content_type(part.parts, content_type)
return body
end
end
end
end
return false
end
These emails are generally in response to a html email it sent out in the first place. (The original outbound email is never the same.) The body text the method above returns contains the full history of the email and I would like to just parse out the reply text.
I'm wondering whether it's reasonable to place some '---please reply above this line---' text at the top of the mail as I have seen in a 37 signals application.
Is there another way to ignore the client specific additions to the email, other than write a multitude of regular expressions (which I haven't yet attempted) for each and every mail client? They all seem to tack on their own bit at the top of any replies.
I have to do email reply parsing on a project I'm working on right now. I ended up using pattern matching to identify the response part, so users wouldn't have to worry about where to insert their reply.
The good news is that the implementation really isn't too difficult. The hard part is just testing all the different email clients and services you want to support and figuring out how to identify each one. Generally, you can use either the message ID or the X-Mailer or Return-Path header to determine where an incoming email came from.
Here's a method that takes a TMail object and extracts the response part of the message and returns that along with the email client/service it was sent from. It assumes you have the original message's From: name and address in the constants FROM_NAME and FROM_ADDRESS.
def find_reply(email)
message_id = email.message_id('')
x_mailer = email.header_string('x-mailer')
# For optimization, this list could be sorted from most popular to least popular email client/service
rules = [
[ 'Gmail', lambda { message_id =~ /.+gmail\.com>\z/}, /^.*#{FROM_NAME}\s+<#{FROM_ADDRESS}>\s*wrote:.*$/ ],
[ 'Yahoo! Mail', lambda { message_id =~ /.+yahoo\.com>\z/}, /^_+\nFrom: #{FROM_NAME} <#{FROM_ADDRESS}>$/ ],
[ 'Microsoft Live Mail/Hotmail', lambda { email.header_string('return-path') =~ /<.+#(hotmail|live).com>/}, /^Date:.+\nSubject:.+\nFrom: #{FROM_ADDRESS}$/ ],
[ 'Outlook Express', lambda { x_mailer =~ /Microsoft Outlook Express/ }, /^----- Original Message -----$/ ],
[ 'Outlook', lambda { x_mailer =~ /Microsoft Office Outlook/ }, /^\s*_+\s*\nFrom: #{FROM_NAME}.*$/ ],
# TODO: other email clients/services
# Generic fallback
[ nil, lambda { true }, /^.*#{FROM_ADDRESS}.*$/ ]
]
# Default to using the whole body as the reply (maybe the user deleted the original message when they replied?)
notes = email.body
source = nil
# Try to detect which email service/client sent this message
rules.find do |r|
if r[1].call
# Try to extract the reply. If we find it, save it and cancel the search.
reply_match = email.body.match(r[2])
if reply_match
notes = email.body[0, reply_match.begin(0)]
source = r[0]
next true
end
end
end
[notes.strip, source]
end
I think you will be stuck on this one. I have been doing some stuff with emails myself in TMail recently, and what you will generally find is that an email that has an HTML part is generally structured like:
part 1 - multipart/mixed
sub part 1 - text/plain
sub part 2 - text/html
end
The email clients I have played with Outlook and Gmail both generate replies in this format, and they just generally quote the original email inline in the reply. At first I though that the 'old' parts of the original email would be separate parts, but they are actually not - the old part is just merged into the reply part.
You could search the part for a line that begins 'From: ' (as most clients generally place a header at the top of the original email text detailing who sent it etc), but its probably not guaranteed.
I don't really see anything wrong with a --- please reply above this line --- generally, its not that invasive, and could make things a lot simpler.