Add an audio alert plugin

config.toml

@@ -52,3 +52,30 @@ enabled = false
 # # goes here!
 # # """
 # template="Pog count: {pog}"
+
+# # Plays a sound when specific criteria are met
+# [[plugin.AudioAlert]]
+# # The output device to use (run `python -m plugins.AudioAlert` to get a list of valid devices)
+# output = 'ALSA:default'
+# # The path to a sound you want to play. Must be in wave format. Depending on your output target, you may have to match channel count and sample rates to the output
+# soundpath = '../your-cool-sound.wav'
+# # An array of criteria. The sound will play if a message meets any of these criteria.
+# # Criteria can have any combination of three selectors:
+# # + event: The name of an event linked to the message
+# # + monitization: The amount of money attached to the message in USD. Can be an exact value or a range (like '1-10')
+# # + regex: A regex to match against the message text. This will partial match, so use start of string and end of string selectors if you want to match the whole message!
+# #
+# # To play a sound if a user donates exactly 4 dolalrs and twenty cents and said "blaze it" or "blazeit", you'd add
+# # criteria = [
+# #   { monitization = 4.20, regex = 'blaze\s?it' },
+# # ]
+# # or if you wanted a the same sound to play for either a monitization between 1 and 5 dollars or a sub, you could have
+# # critera = [
+# #   { monitization = '1-5' },
+# #   { event = 'SUB' },
+# # ]
+# critera = []
+# # How much of the sample to load at a time. Must be a power of two. Increase this if you have glitchy playback!
+# buffer_length = 4096
+# # Prevent cuttoff by waiting this number of buffers before closing the stream. Depending on your output target, this may not be neccesary
+# cutoff_prevention_buffers = 4

plugins/AudioAlert.py

@@ -0,0 +1,122 @@
+import re
+from contextlib import redirect_stdout
+import sys
+
+import pyaudio as pya
+import soundfile
+import numpy as np
+
+from .PluginBase import PluginBase
+
+
+class Plugin(PluginBase):
+    def __init__(self, critera=None, soundpath=None, output=None, buffer_length=1024, cutoff_prevention_buffers=0):
+        super().__init__()
+
+        self._criteria = critera
+        self._soundpath = soundfile
+        self._output = output
+        self._cutoff_prevent_length = cutoff_prevention_buffers
+        self._buffer_length = buffer_length
+
+        self._sound = soundfile.SoundFile(soundpath)
+
+        self._stream = None
+        self._cutoff_prevent_count = 0
+
+    def tick(self, dt):
+        pass
+
+    def handle_event(self, event):
+        pass
+
+    def on_message(self, message):
+        if self.should_alert(message):
+            self._play()
+        return message
+
+    def should_alert(self, message):
+        for criterion in self._criteria:
+            if criterion.get('event'):
+                if not (message.for_event is not None and message.for_event.name == criterion['event']):
+                    continue
+            if criterion.get('monitization'):
+                if '-' in criterion['monitization']:
+                    lower_bound, upper_bound = (float(bound) for bound in criterion['monitization'].split('-'))
+                    if not (message.monitization is not None and message.monitization >= lower_bound and message.monitization <= upper_bound):
+                        continue
+                else:
+                    if not (message.monitization is not None and message.monitization == float(criterion['monitization'])):
+                        continue
+            if criterion.get('regex'):
+                if re.search(criterion['regex'], message.text) is None:
+                    continue
+            return True
+
+    def prepare_stream(self):
+        print('Starting portaudio - prepare for debug spam!')
+        pyaudio = pya.PyAudio()
+        print('-' * 10)
+        if self._output is not None:
+            if ':' in self._output:
+                host_api_name, output_name = self._output.split(':', 1)
+            else:
+                host_api_name = self._output
+                output_name = None
+
+            for i in range(pyaudio.get_host_api_count()):
+                host_api_info = pyaudio.get_host_api_info_by_index(i)
+                if host_api_info['name'] == host_api_name:
+                    if output_name is None:
+                        output_index = host_api_info['defaultOutputDevice']
+                        break
+                    else:
+                        for j in range(host_api_info['deviceCount']):
+                            device_info = pyaudio.get_device_info_by_host_api_device_index(i, j)
+                            if device_info['name'] == output_name:
+                                output_index = device_info['index']
+                                break
+                        else:
+                            raise ValueError(f'Could not find requested output device: {output_name}')
+                    break
+            else:
+                raise ValueError(f'Could not find requested audio API: {host_api_name}')
+        else:
+            output_index = None
+
+        self._stream = pyaudio.open(
+            output_device_index=output_index,
+            format=pya.paFloat32,
+            channels=self._sound.channels,
+            rate=self._sound.samplerate,
+            frames_per_buffer=self._buffer_length,
+            output=True,
+            stream_callback=self._read_callback)
+
+    def _play(self):
+        self._sound.seek(0)
+        self._cutoff_prevent_count = 0
+
+        if self._stream is None:
+            self.prepare_stream()
+        if not self._stream.is_active():
+            self._stream.stop_stream()
+            self._stream.start_stream()
+
+    def _read_callback(self, in_data, frame_count, time_info, status):
+        if self._sound.tell() == self._sound.frames:
+            self._cutoff_prevent_count += 1
+            return np.zeros((frame_count, self._sound.channels)), pya.paContinue if self._cutoff_prevent_count < self._cutoff_prevent_length else pya.paComplete
+
+        return self._sound.read(frames=frame_count, dtype='float32', fill_value=0), pya.paContinue
+
+
+if __name__ == '__main__':
+    p = pya.PyAudio()
+    print('-' * 10)
+    for i in range(p.get_host_api_count()):
+        host_api_info = p.get_host_api_info_by_index(i)
+        for j in range(host_api_info['deviceCount']):
+            device_info = p.get_device_info_by_host_api_device_index(i, j)
+            if device_info['maxOutputChannels'] > 0:
+                print(f"{host_api_info['name']}:{device_info['name']}")