Add an API for plugins

.gitignore

@@ -1,12 +1,8 @@
 # Files from setup
 .gitmodules
-akulai/akulai-plugins
 
-# Ignore the VOSK model folder
-akulai/model
-
-# Ignore the plugins directory
-akulai/plugins
+# Ignore the VOSK model folder and everything in it
+model/vosk_model
 
 # IDE files and folders
 ./idea
@@ -177,3 +173,6 @@ cython_debug/
 #  and can be added to the global gitignore or merged into this file.  For a more nuclear
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
+
+# JS Plugin-related
+node_modules/

.gitmodules

@@ -1,3 +0,0 @@
-[submodule "akulai-plugins"]
-	path = akulai-plugins
-	url = https://github.com/Akul-AI/akulai-plugins

.vscode/settings.json

@@ -0,0 +1,3 @@
+{
+    "python.analysis.typeCheckingMode": "basic"
+}

README.md

@@ -6,14 +6,15 @@ If you want a more stable version (the current version is not done yet and may n
 
 ### Manual installation
 
-First of all, download the ASR model from [here](https://alphacephei.com/vosk/models), unzip it, rename it to `vosk_model` and place it inside `models` folder, which is located in the `akulai` directory. Make sure that you have Node.js and any installation of Perl installed, and make sure cpamn (Perl's package manager) is in the system PATH for Windows.
+First of all, download the ASR model from [here](https://alphacephei.com/vosk/models), unzip it, rename it to `vosk_model` and place it inside `models` folder, which is located in the `akulai` directory. Make sure that you have Node.js and any installation of Perl installed (Strawberry or System Perl recommended), and make sure `cpamn` (Perl's package manager) is in the system PATH for Windows.
 
 ### Automatic installation (with script)
 
-Again, you may either follow the steps above, or you may use the provided python script. We have also provided a Batch (Windows CMD) and Bash (Shell) script. If you face any errors, however, please use the python script before reporting it.
+Again, you may either follow the steps above, or you may use the provided python script. We have also provided a Batch (Windows CMD) and Bash (Shell) script. If you face any errors, please let us know.
 
 Keep in mind that:
- - All the scripts assume you have a working installation of Python installed.
+ - All the scripts assume you have a working installation of Python installed. They do not provide a Python installation like they do for Strawberry Perl and Node.js.
+ - If one of the native scripts don't work, try using the Python script.
 
 ## How do I create a skill...??
 
@@ -25,7 +26,7 @@ Go to the [master-v1 branch](https://github.com/Akul-AI/akulai/tree/master-v1) f
 
 ## Contribution
 
-You can look at the `todo.md` file (not always updated) or you can look at the issues. Make sure to check the CONTRIBUTING.md guide before making a new feature/fix, though. Thanks to all the contributors!
+You can look at the `todo.md` file (not always updated) or you can look at the issues. Make sure to check the `CONTRIBUTING.md` guide before making a new feature/fix, though. Thanks to all the contributors!
 
 <a href="https://github.com/Akul-AI/akulai/graphs/contributors">
   <img src="https://contrib.rocks/image?repo=Akul-AI/akulai" />
@@ -34,4 +35,5 @@ You can look at the `todo.md` file (not always updated) or you can look at the i
 
 ## Other Information
 Version system - semantic versioning
+
 Compatible with - Windows, Linux, Raspberry Pi

akulai/akulai.py

@@ -3,11 +3,14 @@
 import importlib.util
 import json
 import os
-import pyttsx3
+import rlvoice
 import pyaudio
 import subprocess
 import threading
 import vosk
+import sys
+from fastapi import FastAPI
+import time
 
 
 class JSPlugin:
@@ -33,10 +36,15 @@ def __init__(self):
         # Initialize the pyaudio device
         self.p = pyaudio.PyAudio()
         self.stream = self.p.open(format=pyaudio.paInt16, channels=1, rate=16000, input=True, frames_per_buffer=8000)
-        # Initialize the pyttsx3 speech engine
-        self.engine = pyttsx3.init()
+        # Initialize the rlvoice speech engine
+        self.engine = rlvoice.init()
         self.voices = self.engine.getProperty('voices')
+        self.engine.setProperty('rate', 120)
         self.engine.setProperty('voice', self.voices[0].id)
+        # Create the listening thread
+        self.stop_listening = threading.Event()
+        self.listening_thread = threading.Thread(target=self.listen)
+        self.listening_thread.start()
         # load the plugins
         self.discover_plugins()
 
@@ -55,12 +63,12 @@ def discover_plugins(self):
                         # Load the Python plugin using importlib
                         plugin_path = os.path.join(dirpath, filename)
                         spec = importlib.util.spec_from_file_location("plugin", plugin_path)
-                        plugin = importlib.util.module_from_spec(spec)
-                        spec.loader.exec_module(plugin)
+                        plugin = importlib.util.module_from_spec(spec) # type: ignore
+                        spec.loader.exec_module(plugin) # type: ignore
 
                         # Add the plugin to the list
                         self.plugins.append(plugin)
-                        print(f"Loaded plugin from {plugin_path}")
+                        print(f"Loaded python plugin: {plugin_path}")
 
                     elif filename.endswith(".pl"):
                         # Load the Perl plugin using subprocess
@@ -71,7 +79,7 @@ def discover_plugins(self):
                             # Add the plugin to the list
                             print(f"Loaded perl plugin: {plugin_path}")
                         except subprocess.CalledProcessError:
-                            print(f"Error loading perl plugin: {plugin_path}")
+                            print(f"Error loading perl plugin: {plugin_path}") # type: ignore
 
                     elif filename.endswith(".js"):
                         # Load the JavaScript plugin using subprocess
@@ -82,11 +90,11 @@ def discover_plugins(self):
                             # Add the plugin to the list
                             print(f"Loaded node plugin: {plugin_path}")
                         except subprocess.CalledProcessError:
-                            print(f"Error loading node plugin: {plugin_path}")
+                            print(f"Error loading node plugin: {plugin_path}") # type: ignore
 
     # Listen for audio input through mic with pyaudio and vosk
     def listen(self):
-        while not self.stop_listening.is_set():
+        while not self.stop_listening.is_set(): # type: ignore
             data = self.stream.read(16000, exception_on_overflow=False)
             if len(data) == 0:
                 break
@@ -116,22 +124,30 @@ def execute_command(self, command):
         if(not handled):
             self.speak("Sorry, I didn't understand that.")
 
-    def start(self):
-        self.speak("Hello, I am AkulAI. How can I help you today?")
-        # Create the listening thread
-        self.stop_listening = threading.Event()
-        self.listening_thread = threading.Thread(target=self.listen)
-        self.listening_thread.start()
-
     # shut down the program and all threads
     def stop(self):
-        self.stop_listening.set()
+        self.stop_listening.set() # type: ignore
         self.stream.stop_stream()
         self.stream.close()
         self.p.terminate()
-        exit()
+        sys.exit()
+
+# Set up API
+app = FastAPI()
+akulAI = AkulAI()
+
+# Run API server
+os.system("uvicorn akulai:app --reload")
+time.sleep(5)
+
+@app.post("/speak/{text}")
+async def speak(text: str):
+    akulAI.speak(text)
+    return {"message": "Text synthesized"}
 
+@app.post("/listen")
+async def listen():
+    akulAI.listen()
+    return {"message": "Listening..."}
 
-if __name__ == '__main__':
-    akulai = AkulAI()
-    akulai.start()
+akulAI.speak("Hello, I am AkulAI. How can I help you today?")

docs/create_plugins.md

@@ -12,33 +12,40 @@ Next, create a file in your sub-directory called `plugin.info`. It should look s
 ```
 author: John Doe
 dependencies: requests, pandas
-description: Lorem ipsum di olor nulla quis lorem ut libero malesuada feugiat. This plugin.info file is an example. See the akulai_plugins repository for more examples.
+description: Lorem ipsum di olor nulla quis lorem ut libero malesuada feugiat. This plugin.info file is an example.
 ```
 
-The dependencies may vary based on your project. Note that when listing the dependencies, list them by the name you installed them. For example, if you installed a dependency with `pip install py-example`(note that this is an example, and applies to all languages), but imported it with `import example`, you would still list the dependency `as py-example`. If you have no dependencies required to be installed, just leave it blank. 
+The dependencies may vary based on your project. Note that when listing the dependencies, list them by the name you installed them. For example, if you installed a dependency with `pip install py-example`(note that this is an example, and applies to all languages), but imported it with `import example`, you would still list the dependency `as py-example`. If you have no dependencies required to be installed, just leave it blank. Note that default packages (such as `os` in python, which comes preinstalled with python) should nnot be included.
+
+Moreover, all plugins now require calling an API which gives them access to the speak and listen function, allowing them to speak and listen when needed in the plugin. For using the speak function, it requires a GET request, and for listen, it requires a POST request. I have only included GET in this so far. If, for whatever reason, you do not require this, you may skip calling the API, and print in the console rather than using the `speak()` and `listen()` functions.
 
 ## Python Plugins
 Python plugins should be written as a function called `handle()` with one parameter, `command`, which is the text of what the user said. It should return the text of AkulAI would like to say to the user.
 
 Here is an example of a Python plugin that says "Hello, World!" when the user says "hello":
 
-``` python
+```python
+import requests
+
+response = requests.get('http://127.0.0.1:8000/speak')
 def handle(command):
     if "hello" in command:
-        return "Hello there!"
+        response.speak("Hello there!")
 ```
 ## Javascript Plugins
 JavaScript plugins should read from the commandline and write to stdout using console.log().
 
 Here is an example of a JavaScript plugin that says "Hello, World!" when the user says "hello":
 
-``` javascript
+```javascript
+fetch('http://127.0.0.1:8000/speak')
+
 command = ""
 if(process.argv.length > 2){
     command = process.argv[2]
 }
 if (command.indexOf("hello") !== -1){
-    console.log("Hello there!")
+    speak("Hello there!")
 }
 ```
 If you want to check for multiple words, you can use the `.indexOf()` method multiple times and use logical operators `(&&, ||, etc)` to check if multiple conditions are met.
@@ -54,29 +61,53 @@ if (command.includes("hello")){
 ```
 or you can use `RegExp` to check if a string matches a specific pattern, like this:
 
-``` javascript
+```javascript
 let match = command.match(/hello/);
 if (match) {
     console.log("Hello there!")
 }
 ```
 It all depends on your plugin and preference.
 
-Keep in mind that when making plugins which require packages, the `node_modules` directory and the `package.json` file belong in the root directory.
+Keep in mind that when making plugins which require packages, the `node_modules` directory and the `package.json` file belong in the directory of the plugin.
+
+Here is an example of a `package.json` file for the inbuilt calendar skill:
+```json
+{
+  "name": "calendar",
+  "version": "1.0.0",
+  "description": "",
+  "main": "main.js",
+  "scripts": {
+    "test": "npm run dev",
+    "start": "node main.js"
+  },
+  "author": "Akul Goel",
+  "license": "ISC",
+  "dependencies": {
+    "fs": "^0.0.1-security"
+  }
+}
+```
+<b>MAKE SURE YOU ONLY USE NPM, OTHER JS PACKAGE MANAGERS AREN'T CURRENTLY SUPPORTED.</b>
 
 ## Perl Plugins
 
 Here is an example of a Perl plugin:
 
 ``` perl
 #!/usr/bin/perl
+use LWP::UserAgent;
+
+my $ua = LWP::UserAgent->new;
+my $response = $ua->get('http://127.0.0.1:8000/speak');
 
 my $command = shift();
 if($command=~/hello/){
-    print("Hello there!\n");
+    speak("Hello there!\n");
 }
 ```
-Perl plugins should read from the commandline and write to stdout using print(). This example uses the command variable to check if the command contains the word "hello" and if true, print a response for AkulAI to read.
+This example uses the command variable to check if the command contains the word "hello" and if true, speak a response for AkulAI to read.
 
 ## Using a Plugin
 Once a plugin has been created, it will automatically be loaded and available for use when the `AkulAI` class is instantiated. The `AkulAI` class will search for any files with the extensions of ".py" or ".js" in the "plugins" directory, and will add them to the list of available plugins.

docs/strawberry-perl-docs.md

@@ -0,0 +1,50 @@
+=== Strawberry Perl (64-bit) 5.38.0.1-64bit README ===
+
+What is Strawberry Perl?
+------------------------
+
+* 'Perl' is a programming language suitable for writing simple scripts as well
+  as complex applications. See https://perldoc.perl.org/perlintro.html
+
+* 'Strawberry Perl' is a perl environment for Microsoft Windows containing all
+  you need to run and develop perl applications. It is designed to be as close
+  as possible to perl environment on UNIX systems. See https://strawberryperl.com/
+
+* If you are completely new to perl consider visiting http2://learn.perl.org/
+
+Installation instructions: (.ZIP distribution only, not .MSI installer)
+-----------------------------------------------------------------------
+
+* If installing this version from a .zip file, you MUST extract it to a 
+  directory that does not have spaces in it - e.g. c:\myperl\
+  and then run some commands and manually set some environment variables:
+
+  c:\myperl\relocation.pl.bat         ... this is REQUIRED!
+  c:\myperl\update_env.pl.bat         ... this is OPTIONAL
+
+  You can specify " --nosystem" after update_env.pl.bat to install Strawberry 
+  Perl's environment variables for the current user only.
+
+* If having a fixed installation path does not suit you, try "Strawberry Perl
+  Portable Edition" from https://strawberryperl.com/releases.html
+
+How to use Strawberry Perl?
+---------------------------
+
+* In the command prompt window you can:
+
+  1. run any perl script by launching
+
+     c:\> perl c:\path\to\script.pl
+
+  2. install additional perl modules (libraries) from https://www.metacpan.org/ by
+
+     c:\> cpanm Module::Name
+
+  3. run other tools included in Strawberry Perl like: perldoc, gcc, gmake ...
+
+* You'll need a text editor to create perl scripts.  One is NOT included with 
+  Strawberry Perl. A few options are Padre (which can be installed by running 
+  "cpan Padre" from the command prompt) and Notepad++ (which is downloadable at
+  https://notepad-plus-plus.org/ ) which both include syntax highlighting
+  for perl scripts. You can even use Notepad, if you wish.