kaylerch / alexa-skills-kit-tellask-java Goto Github PK

I tried to gave alexa-skills-kit-tellask a try, but when testing my skill, I get:

The remote endpoint could not be called, or the response it returned was invalid.

When I open my log, it says:

java.lang.ClassNotFoundException: io.klerch.alexa.tellask.util.factory.AlexaIntentHandlerFactoryImpl

Indeed, there is no such class but your AlexaIntentHandlerFactory tries to create one.

I directly followed your example.

What am I missing?

I took the default sample and created a jar from it via Maven, uploaded it to lambda under java 8 and as soon i hit the LaunchIntent I get this error

{
  "errorMessage": "Class not found: index.handler",
  "errorType": "java.lang.ClassNotFoundException"
}

The conventional way was that I used to create a command line java project, install lambda sdk and something like this

public class ClassName implements RequestHandler<Map<String, Object>, Object> {

@Override
    public Object handleRequest(Map<String, Object> input, Context context) {
        String data = input.toString();
        return  doAction(input.toString());
    }

Please do guide me if I am doing something wrong here

Hi Kay,

thanks for this great library. Really love the annotation based approach. 👍

Just one small question, feature request or wish: Is there any way to access the SystemState object (from AlexaInput) within AlexaIntentHandler? I think that SystemState object would be the only way to get alexa DeviceId and ApiEndpoint, which are needed in order to implement the new alexa address features.

Amzn sample implementation: https://github.com/amzn/alexa-skills-kit-java/blob/master/samples/src/main/java/address/DeviceAddressSpeechlet.java#L138

Thanks & Regards,
Philipp

Hi Kay,

Sorry to post this here though this is not an issue.

I' am an Alexa developer(Java) and I code the way that is given in the amazon sample code library.

I've seen your code, but sorry to say that I didn't understand this quite well.(Problem from my end, not your end). As this is divided into multiple files and you are using an YAML instead of the regular utterances and intents in the UI, also the divided files are confusing me a lot on terms like where does the flow start from and on how it goes on.

Can you please make one of the sample codes in the way that you created this, so that I can match the sample code(the default provided by alexa) to this and understand on how to be a better coder in terms of following the semantics and make it more readable and user friendly.

I request you to do it using the sessions example(the colors)

Thanks & Regards,
Rakesh

So I tried to dig into this library as it seems to make coding for Alexa using Java (and Scala) easier. I set up a build.sbt like this:

name := "hello-world"
version := "0.0.1-SNAPSHOT"
scalaVersion := "2.12.2"

libraryDependencies += "com.amazon.alexa" % "alexa-skills-kit" % "1.3.1"
libraryDependencies += "io.klerch" % "alexa-skills-kit-tellask-java" % "0.2.2"

I now created an utterances.yml in resources/de-DE like this:

SayWelcome:
  Utterances:
    - "Hallo. Das ist ein Test."

SayTest:
  Utterances:
    - "Das ist ein Test."

SaySorry:
  Utterances:
    - "Irgendwas ist schief gelaufen."

I created the speechlet handler:

@AlexaApplication(applicationIds = Array("amzn1.ask.skill.xxxxx"))
class HelloWorldSpeechletHandler extends AlexaRequestStreamHandler {
}

I created the error handler:

@AlexaLaunchListener
class HelloWorldLaunchHandler extends AlexaLaunchHandler {

  def handleError(exception: AlexaRequestHandlerException): AlexaOutput = AlexaOutput.tell("SaySorry").build()
  def handleRequest(input: AlexaInput): AlexaOutput = AlexaOutput.tell("SayWelcome").build()
}

And my test error handler (this time in Java to make sure Scala is not the cause for the problem):

@AlexaIntentListener(customIntents = "ThisIsATest")
public class ListThingsHandler implements AlexaIntentHandler {

    @Override
    public boolean verify(final AlexaInput input) {
        return false;
    }

    @Override
    public AlexaOutput handleRequest(final AlexaInput input) throws AlexaRequestHandlerException, AlexaStateException {
        return AlexaOutput.tell("SayTest").build();
    }

    @Override
    public AlexaOutput handleError(final AlexaRequestHandlerException exception) {
        return AlexaOutput.tell("SaySorry").build();
    }
}

In the interaction model builder (the new interactive one) I created a "ThisIsATest" intent and compiled the interaction model.

When I want to test my skill, I get this error in my log:

com.amazon.speech.speechlet.SpeechletException: Could not find a handler for speechlet request: java.io.IOException
java.io.IOException: com.amazon.speech.speechlet.SpeechletException: Could not find a handler for speechlet request
at io.klerch.alexa.tellask.model.wrapper.AlexaRequestStreamHandler.handleRequest(AlexaRequestStreamHandler.java:92)
Caused by: com.amazon.speech.speechlet.SpeechletException: Could not find a handler for speechlet request
at io.klerch.alexa.tellask.model.wrapper.AlexaSpeechlet.handleRequest(AlexaSpeechlet.java:105)
at io.klerch.alexa.tellask.model.wrapper.AlexaSpeechlet.onIntent(AlexaSpeechlet.java:90)
at com.amazon.speech.speechlet.SpeechletToSpeechletV2Adapter.onIntent(SpeechletToSpeechletV2Adapter.java:51)
at com.amazon.speech.speechlet.SpeechletRequestDispatcher.dispatchSpeechletCall(SpeechletRequestDispatcher.java:218)
at com.amazon.speech.speechlet.SpeechletRequestHandler.internalHandleSpeechletCall(SpeechletRequestHandler.java:152)
at com.amazon.speech.speechlet.SpeechletRequestHandler.handleSpeechletCall(SpeechletRequestHandler.java:91)
at io.klerch.alexa.tellask.model.wrapper.AlexaRequestStreamHandler.handleRequest(AlexaRequestStreamHandler.java:88)

I packed my Jar with the SBT assembly plugin which embeds all dependencies similar to the maven-shade-plugin.

What am I missing?

Hi, If I try to build AlexaOutput with a $ symbol in the string, for example: AlexaOutput.ask("UtteranceKey").putSlot("slot1","$5.00"); the handler fails to error state with the exception: "io.klerch.alexa.tellask.util.AlexaRequestHandlerException: General error occured." A little debugging digs up the real issue: "java.lang.indexoutofboundsexception no group 3" which happens in the while loop starting on line 109 of AlexaSpeechletResponse.class in the method resolveSlotsInUtterance. This happens because the regex engine thinks "$5" is a token, it should be escaped before processing.

A workaround is to use "Matcher.quoteReplacement" in the input for the slot, for example:
AlexaOutput.ask("UtteranceKey").putSlot("slot1",Matcher.quoteReplacement("$5.00"));

While the workaround is sufficient for my purposes, it would be great if the library took care of this detail for me and I didn't have to worry about it. Thanks!

Hi @KayLerch first a big thank you for this amazing sdk , this is a very well written piece and my concern is how often can I expect an update of the version in sync with the amazon skill set sdk . The current version 1.3.1 but the tell ask sdk has 1.3.0 on it . Any info on this , kindly advise