Enhanced version of gists-can-help-for-tropo-dev.js (https://gist.github.com/ObjectIsAdvantag/cc131aab9409f014db22e18e572f722d)

DevOps-with-CiscoSpark.js

// DO NOT DELETE : this gist is referenced by a live article

//
// Cisco Spark Logging Library for Tropo
//

// Factory for the Spark Logging Library, with 2 parameters
//    - the name of the application will prefix all your logs, 
//    - the Spark Incoming integration (to  which logs will be posted)
// To create an Incoming Integration
//   - click integrations in the right pane of a Spark Room (Example : I create a dedicated "Tropo Logs" room)
//   - select incoming integration
//   - give your integration a name, it will be displayed in the members lists (Example : I personally named it "from tropo scripting")
//   - copy your integration ID, you'll use it to initialize the SparkLibrary
function SparkLog(appName, incomingIntegrationID) {
	
	if (!appName) {
		log("SPARK_LOG : bad configuration, no application name, exiting...");
		throw createError("SparkLibrary configuration error: no application name specified");
	}
        this.tropoApp = appName;

	if (!incomingIntegrationID) {
		log("SPARK_LOG : bad configuration, no Spark incoming integration URI, exiting...");
		throw createError("SparkLibrary configuration error: no Spark incoming integration URI specified");
	}
        this.sparkIntegration = incomingIntegrationID;
        
	log("SPARK_LOG: all set for application:" + this.tropoApp + ", posting to integrationURI: " + this.sparkIntegration);
}

// This function sends the log entry to the registered Spark Room 
// Invoke this function from the Tropo token-url with the "sparkIntegration" parameter set to the incoming Webhook ID you'll have prepared
// Returns true if the log entry was acknowledge by Spark (ie, got a 2xx HTTP status code)
SparkLog.prototype.log = function(newLogEntry) {
	
    // Robustify
    if (!newLogEntry) {
    	newLogEntry = "";
    }
    
    var result;
    try {
        // Open Connection
        var url = "https://api.ciscospark.com/v1/webhooks/incoming/" + this.sparkIntegration;
        connection = new java.net.URL(url).openConnection();

        // Set timeout to 10s
        connection.setReadTimeout(10000);
        connection.setConnectTimeout(10000);

        // Method == POST
        connection.setRequestMethod("POST");
        connection.setRequestProperty("Content-Type", "application/json");
        
        // TODO : check if this cannot be removed
        connection.setRequestProperty("Content-Length", newLogEntry.length);
        connection.setUseCaches (false);
        connection.setDoInput(true);
        connection.setDoOutput(true); 

        //Send Post Data
        bodyWriter = new java.io.DataOutputStream(connection.getOutputStream());
        log("SPARK_LOG: posting: " + newLogEntry + " to: " + url);
        contents = '{ "text": "' + this.tropoApp + ': ' + newLogEntry + '" }'
        bodyWriter.writeBytes(contents);
        bodyWriter.flush ();
        bodyWriter.close (); 

        result = connection.getResponseCode();
        log("SPARK_LOG: read response code: " + result);

	if(result < 200 || result > 299) {
	        log("SPARK_LOG: could not log to Spark, message format not supported");
	        return false;
	 }
    }
    catch(e) {
        log("SPARK_LOG: could not log to Spark, socket Exception or Server Timeout");
        return false;
    }
    
    log("SPARK_LOG: log successfully sent to Spark, status code: " + result);
    return true; // success
}

// Let's create several instances for various log levels
// Note that you may spread logs to distinct rooms by changing the integrationId
var SparkInfo = new SparkLog("Spark Onboarding - INFO:", "YOUR_INTEGRATION_SUFFIX");
var SparkDebug = new SparkLog("Spark Onboarding - DEBUG:", "YOUR_INTEGRATION_SUFFIX");

//
// Log Configuration happens here
//

// info level used to get a synthetic sump up of what's happing
function info(logEntry) {
  log("INFO: " + logEntry);
  SparkInfo.log(logEntry);
  // Uncomment if you opt to go for 2 distinct Spark Rooms for DEBUG and INFO log levels
  //SparkDebug.log(logEntry); 
}

// debug level used to get detail informations
function debug(logEntry) {
  log("DEBUG: " + logEntry);
  SparkDebug.log(logEntry);
}

//
// Tropo Script custom logic starts here
//

// returns true or false
function isEmail(email) {
    // extract from http://stackoverflow.com/questions/46155/validate-email-address-in-javascript
    var re = /^(([^<>()\[\]\\.,;:\s@"]+(\.[^<>()\[\]\\.,;:\s@"]+)*)|(".+"))@((\[[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}])|(([a-zA-Z\-0-9]+\.)+[a-zA-Z]{2,}))$/;
    return re.test(email);
}

// returns an email address if found in the phrase specified
function extractEmail(phrase) {
    if (phrase) {
        var parts = phrase.split(" ");
        for (var i = 0; i < parts.length ; i++) {
            if (isEmail(parts[i])) {
                return parts[i];
            }
        }
    }
    return null;
}

// You may check currentCall features here : https://www.tropo.com/docs/scripting/currentcall
if (currentCall) {
  if (currentCall.network == "SMS") {
    var input = currentCall.initialText;
    debug("received: " + input  + ", from: " + currentCall.callerID );
    
    // check we received a valid email address
    var email = extractEmail(input);
    if (email) {
      info("new subscription for email: " + email);
      say ("Thanks for suscribing to our service.");  
    }
    else {
      debug("could not extract email from: " + input  + ", for: " + currentCall.callerID );
      say ("Sorry we could not read your email");  
    }
  }
  else {
    // Speak a welcome message if the incoming call is not a SMS
    info("incoming call from: " + currentCall.callerID );
    say("Please text your email to this number.");
  }
}
else {
  debug("no outgoing call support here");
}