Replace deprecated audioop with NumPy and add silent join sounds

- Replace audioop module (removed in Python 3.13) with NumPy
  - Implement _apply_volume() method using NumPy for volume adjustment
  - Update RMS calculation in ducking_sound_received() to use NumPy
  - Add numpy to requirements.txt
- Add silent flag to CachedItemWrapper for join sounds
  - Join sounds now play without channel announcements
  - Regular playlist items still respect announce_current_music config
- Update CLAUDE.md to document changes

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

bragi.py

@@ -15,7 +15,7 @@ import sys
 import math
 import signal
 import configparser
-import audioop
+import numpy as np
 import subprocess as sp
 import argparse
 import os.path
@@ -427,7 +427,7 @@ class MumbleBot:
             if sound_config.startswith('http://') or sound_config.startswith('https://'):
                 # It's a URL
                 self.log.info(f'bot: Playing join sound URL for {username}: {sound_config}')
-                music_wrapper = get_cached_wrapper_from_scrap(type='url', url=sound_config, user='system')
+                music_wrapper = get_cached_wrapper_from_scrap(type='url', url=sound_config, user='system', silent=True)
                 self.log.info(f'bot: Successfully created music wrapper for {username}')
             else:
                 # It's a file path - search database for first match
@@ -440,7 +440,7 @@ class MumbleBot:
 
                 # Use first match
                 self.log.info(f'bot: Playing join sound file for {username}: {matches[0]["path"]}')
-                music_wrapper = get_cached_wrapper_from_dict(matches[0], 'system')
+                music_wrapper = get_cached_wrapper_from_dict(matches[0], 'system', silent=True)
                 self.log.info(f'bot: Successfully created music wrapper for {username}')
 
             # Add to playlist
@@ -509,7 +509,8 @@ class MumbleBot:
 
         self.log.info("bot: play music " + music_wrapper.format_debug_string())
 
-        if var.config.getboolean('bot', 'announce_current_music'):
+        # Only announce if configured and not a silent item (e.g., join sounds)
+        if var.config.getboolean('bot', 'announce_current_music') and not music_wrapper.silent:
             self.send_channel_msg(music_wrapper.format_current_playing())
 
         if var.config.getboolean('debug', 'ffmpeg'):
@@ -628,13 +629,13 @@ class MumbleBot:
 
                     if not self.on_interrupting and len(raw_music) == self.pcm_buffer_size:
                         self.mumble.sound_output.add_sound(
-                            audioop.mul(raw_music, 2, self.volume_helper.real_volume))
+                            self._apply_volume(raw_music, self.volume_helper.real_volume))
                     elif self.read_pcm_size == 0:
                         self.mumble.sound_output.add_sound(
-                            audioop.mul(self._fadeout(raw_music, self.stereo, fadein=True), 2, self.volume_helper.real_volume))
+                            self._apply_volume(self._fadeout(raw_music, self.stereo, fadein=True), self.volume_helper.real_volume))
                     elif self.on_interrupting or len(raw_music) < self.pcm_buffer_size:
                         self.mumble.sound_output.add_sound(
-                            audioop.mul(self._fadeout(raw_music, self.stereo, fadein=False), 2, self.volume_helper.real_volume))
+                            self._apply_volume(self._fadeout(raw_music, self.stereo, fadein=False), self.volume_helper.real_volume))
                         self._cleanup_ffmpeg_process()
                         time.sleep(0.1)
                         self.on_interrupting = False
@@ -728,7 +729,9 @@ class MumbleBot:
             self.last_volume_cycle_time = time.time()
 
     def ducking_sound_received(self, user, sound):
-        rms = audioop.rms(sound.pcm, 2)
+        # Calculate RMS (root mean square) for volume ducking
+        audio_array = np.frombuffer(sound.pcm, dtype=np.int16)
+        rms = int(np.sqrt(np.mean(audio_array.astype(np.float64) ** 2)))
         self._max_rms = max(rms, self._max_rms)
         if self._display_rms:
             if rms < self.ducking_threshold:
@@ -743,6 +746,27 @@ class MumbleBot:
                 self.on_ducking = True
             self.ducking_release = time.time() + 1  # ducking release after 1s
 
+    def _apply_volume(self, pcm_data, volume):
+        """Apply volume adjustment to 16-bit PCM audio data.
+
+        Replaces deprecated audioop.mul() with modern NumPy implementation.
+
+        Args:
+            pcm_data: bytes containing 16-bit signed PCM samples
+            volume: float multiplier (0.0 to 1.0+)
+
+        Returns:
+            bytes with volume-adjusted audio
+        """
+        # Convert bytes to numpy array of 16-bit signed integers
+        audio_array = np.frombuffer(pcm_data, dtype=np.int16)
+
+        # Apply volume (multiply and clip to prevent overflow)
+        adjusted = np.clip(audio_array * volume, -32768, 32767).astype(np.int16)
+
+        # Convert back to bytes
+        return adjusted.tobytes()
+
     def _fadeout(self, _pcm_data, stereo=False, fadein=False):
         pcm_data = bytearray(_pcm_data)
         if stereo:

media/cache.py

@@ -139,11 +139,12 @@ class MusicCache(dict):
 
 
 class CachedItemWrapper:
-    def __init__(self, lib, id, type, user):
+    def __init__(self, lib, id, type, user, silent=False):
         self.lib = lib
         self.id = id
         self.user = user
         self.type = type
+        self.silent = silent  # If True, don't announce when playing (for join sounds)
         self.log = logging.getLogger("bot")
         self.version = 0
 
@@ -216,10 +217,10 @@ class CachedItemWrapper:
 
 
 # Remember!!! Get wrapper functions will automatically add items into the cache!
-def get_cached_wrapper(item, user):
+def get_cached_wrapper(item, user, silent=False):
     if item:
         var.cache[item.id] = item
-        return CachedItemWrapper(var.cache, item.id, item.type, user)
+        return CachedItemWrapper(var.cache, item.id, item.type, user, silent=silent)
     return None
 
 def get_cached_wrappers(items, user):
@@ -234,12 +235,13 @@ def get_cached_wrapper_from_scrap(**kwargs):
     item = var.cache.get_item(**kwargs)
     if 'user' not in kwargs:
         raise KeyError("Which user added this song?")
-    return CachedItemWrapper(var.cache, item.id, kwargs['type'], kwargs['user'])
+    silent = kwargs.get('silent', False)
+    return CachedItemWrapper(var.cache, item.id, kwargs['type'], kwargs['user'], silent=silent)
 
-def get_cached_wrapper_from_dict(dict_from_db, user):
+def get_cached_wrapper_from_dict(dict_from_db, user, silent=False):
     if dict_from_db:
         item = dict_to_item(dict_from_db)
-        return get_cached_wrapper(item, user)
+        return get_cached_wrapper(item, user, silent=silent)
     return None
 
 def get_cached_wrappers_from_dicts(dicts_from_db, user):
@@ -250,10 +252,10 @@ def get_cached_wrappers_from_dicts(dicts_from_db, user):
 
     return items
 
-def get_cached_wrapper_by_id(id, user):
+def get_cached_wrapper_by_id(id, user, silent=False):
     item = var.cache.get_item_by_id(id)
     if item:
-        return CachedItemWrapper(var.cache, item.id, item.type, user)
+        return CachedItemWrapper(var.cache, item.id, item.type, user, silent=silent)
 
 def get_cached_wrappers_by_tags(tags, user):
     items = var.cache.get_items_by_tags(tags)

requirements.txt

@@ -5,4 +5,5 @@ mutagen
 requests
 packaging
 pyradios
-opuslib==3.0.1
+opuslib==3.0.1
+numpy