myask

myask: My Alexa Skills Kit (ASK) helper library

(C) 2017 by "acmurmeltier69"

This library contains a set of functions and classes I found helpful when creating my own Alexa Skills

Modules in the package

myask_appdef.py

a class to manage the application data structure of the skill.

The application structure is is loaded from a separate file and is used to

1. create the "Interaction Model" for Alexa

• definition of intent structure
• definition of custom slot types

2. read, write and validate slots (using "myask_slots")
3. generate random Alexa output structures for testing the handler (using myask_localtest) Note:

• for each custom slottype, the application definition provides a set of standardized values (canonicals)
• for each canonical value, the application definition provides a set of spoken words (literals) that can be used to express this value
• I have deliberately decided to specify the application data as global variables in an importet python file rather than reading it from JSON file to
  • speed-up reading of data structure at run-time when handler function is invoked
  • avoid parsing error at run time
  • easily allow import of external data structures (e.g. long list of station names stored in a different file)

See Application Data Structure below

This module can also be called on the command line: myask_appdef [-out ROOTFILE] appdef Call with '--help' for syntax

myask_slots.py

Helper functions to read, normalize (i.e. convert to canonical) and write slot and session attributes

myask_alexaout.py

Functions to create output structure for Alexa

myask_log.py

Functions for logging (debug, print, error) and error counting

myask_dynamodb.py

A class and functions for storing user profile info in aws dynamodb

This tool can be called on the commandline to

• create new db tables
• scan all user profiles and access statistics in a db table
• print a specific user profile
• print access statistics for a specific user
• add a new user profile
• delete a user profile

Call with '--help' for syntax

myask_utterancegen.py

Tool to generate a large number of input samples for an Alexa skill from a compact generation grammar This tool can be called from the commandline

Call with '--help' for syntax

myask_localtest

Functions test an aw lambda function on windows Can currently not be called from the commandline (maybe later)

Application Data Structure

The core of myask functions is a data structure that describes the application data (intents, slots, values) Thee data structure consists of

• an application name (used in the title of output cards)

• an ASK application ID (used to verify the application in the handler)

• a list of intents with a list of slots for each intent

• the definition of the slot type for each slot

• the definition of custom slot types

• For each custom slottype, the application definition provides a set of standardized values (canonicals)

• for each canonical value, the application definition provides a set of spoken words (literals) that can be used to express this value

Example:

APPNAME = "SAMPLE"
APPID   = "amzn1.ask.skill.aaaaaaaa-bbbb-cccc-dddd-1234567890ab"
INTENTS = {
            "TellColor":    ["Color"],
            "MixColors":    ["MixColor1", "MixColor2"],
            "Confirm":      ["YesNo"]
            "AMAZON.HelpIntent"  : []
           }

SLOTS = {
            "Color" :     "LIST_OF_COLORS",
            "MixColor1" : "LIST_OF_COLORS",
            "MixColor2" : "LIST_OF_COLORS",
            "YesNo"     : "MY_YES_NO"
         }
SLOTTYPES = {
            "LIST_OF_COLORS":[
                ["RED", [u"red", u"fire red"]],
                ["GREEN", [u"green"]]
                ["YELLOW", [u"yellow"]]
                ["BLUE", [u"blue"]]                
            ],
            "YES_NO_TYPE" :[ 
                ["YES_CANONICAL",   [u"yes", u"yep", u"of course"]],
                ["NO_CANONICAL",    [u"no",  u"nope", u"no way"]] 
            ]
   }