diff --git a/va-recorder.py b/va-recorder.py index a4e9081..3f3e87d 100755 --- a/va-recorder.py +++ b/va-recorder.py @@ -1,318 +1,349 @@ #!/usr/bin/env python3 import pyaudio import wave from ctypes import * from optparse import OptionParser from pathlib import Path import math +import psutil import struct import wave import time import os import sys from datetime import datetime import functools print = functools.partial(print, flush=True) # remove useless alsa logging # From alsa-lib Git 3fd4ab9be0db7c7430ebd258f2717a976381715d # $ grep -rn snd_lib_error_handler_t # include/error.h:59:typedef void (*snd_lib_error_handler_t)(const char *file, int line, const char *function, int err, const char *fmt, ...) /* __attribute__ ((format (printf, 5, 6))) */; # Define our error handler type ERROR_HANDLER_FUNC = CFUNCTYPE(None, c_char_p, c_int, c_char_p, c_int, c_char_p) def py_error_handler(filename, line, function, err, fmt): pass #print('messages are yummy') c_error_handler = ERROR_HANDLER_FUNC(py_error_handler) asound = cdll.LoadLibrary('libasound.so') # Set error handler asound.snd_lib_error_set_handler(c_error_handler) # import sounddevice as sd #devices = sd.query_devices() #print(devices) f_name_directory = '%s/.sipclient/spool/playback' % Path.home() lock_file = '%s/.sipclient/spool/playback/playback.lock' % Path.home() +def checkIfProcessRunning(processName): + ''' + Check if there is any running process that contains the given name processName. + ''' + #Iterate over the all the running process + for proc in psutil.process_iter(): + try: + # Check if process name contains the given name string. + if processName.lower() in proc.name().lower(): + return True + except (psutil.NoSuchProcess, psutil.AccessDenied, psutil.ZombieProcess): + pass + return False; + + class Recorder: chunk = 1024 FORMAT = pyaudio.paInt16 CHANNELS = 1 @staticmethod def rms(frame): SHORT_NORMALIZE = (1.0/32768.0) count = len(frame) / 2 format = "%dh" % (count) shorts = struct.unpack(format, frame) sum_squares = 0.0 for sample in shorts: n = sample * SHORT_NORMALIZE sum_squares += n * n rms = math.pow(sum_squares / count, 0.5) return rms * 1000 def __init__(self, target, options): self.p = pyaudio.PyAudio() info = self.p.get_host_api_info_by_index(0) numdevices = info.get('deviceCount') self.target = target self.timeout_length = options.timeout self.RATE = options.rate self.device = options.device self.min_rec_time = options.min_rec_time self.max_rec_time = options.max_rec_time self.external_lock_file = options.external_lock_file - self.external_trigger_file = options.external_trigger_file + self.activate_by_file = options.activate_by_file + self.activate_by_level = options.activate_by_level > 0 + self.threshold = options.activate_by_level + self.quiet = options.quiet self.started_by_file = False self.started_by_level = False devices = {} orig_device = self.device devices_text = [] for i in range(0, numdevices): if (self.p.get_device_info_by_host_api_device_index(0, i).get('maxInputChannels')) > 0: devices[i] = self.p.get_device_info_by_host_api_device_index(0, i).get('name') if devices[i] == str(self.device): self.device = i elif str(i) == str(self.device): self.device = i devices_text.append("%s) %s" % (i, devices[i])) try: print('Using audio device %s) %s at sample rate %d' % (self.device, devices[self.device], self.RATE)) except KeyError: print('Non existent audio device %s' % orig_device) print('Available devices: %s' % ", ".join(devices_text)) sys.exit(0) - self.threshold = 0 if target == 'test' else options.threshold self.stream = self.p.open(format=self.FORMAT, channels=self.CHANNELS, rate=self.RATE, input=True, output=True, input_device_index=self.device, frames_per_buffer=self.chunk) def listen(self): wait_print = False i = 0 lock_print = False external_lock_print = False + festival_print = False while True: i = i + 1 input = self.stream.read(self.chunk, exception_on_overflow = False) rms_val = self.rms(input) now = datetime.now().strftime('%Y-%m-%d %H:%M:%S') if os.path.exists(lock_file): if not lock_print: print("%s - lock file %s present, listen paused" % (now, lock_file)) lock_print = True continue else: if lock_print: print("%s - lock file %s absent, listen resumed" % (now, lock_file)) lock_print = False if self.external_lock_file: if os.path.exists(self.external_lock_file): if not external_lock_print: print("%s - external lock file %s present, listen paused" % (now, self.external_lock_file)) external_lock_print = True continue else: if external_lock_print: print("%s - external lock file %s absent, listen resumed" % (now, self.external_lock_file)) external_lock_print = False + if checkIfProcessRunning('festival'): + if not festival_print: + print("%s - festival running, listen paused" % now) + festival_print = True + continue + else: + if festival_print: + print("%s - festival stopped, listen resumed" % now) + festival_print = False + wait_print = False - if self.external_trigger_file: - if os.path.exists(self.external_trigger_file): + if self.activate_by_file: + if os.path.exists(self.activate_by_file): self.started_by_file = True - print('%s - recording by file %s' % (now, self.external_trigger_file)) - elif rms_val >= self.threshold: + print('%s - recording by file %s' % (now, self.activate_by_file)) + + if not self.started_by_file and self.activate_by_level and rms_val >= self.threshold: print('%s - recording by level %3d > %d' % (now, rms_val, self.threshold)) self.started_by_level = True if self.started_by_level or self.started_by_file: self.record() else: if not self.quiet: - if self.external_trigger_file: + if self.activate_by_file: print("%s - listening, level %3d" % (now, rms_val), end='\r') else: print("%s - listening, level %3d < %d" % (now, rms_val, self.threshold), end='\r') def record(self): rec = [] current = time.time() end = time.time() + self.timeout_length start_time = current recording = False i = 0 #now = datetime.now().strftime('%Y-%m-%d %H:%M:%S') #print('%s - now recording...' % now) if self.started_by_level: while current <= end: i = i + 1 data = self.stream.read(self.chunk) rms_val = self.rms(data) now = datetime.now().strftime('%Y-%m-%d %H:%M:%S') if rms_val >= self.threshold: end = time.time() + self.timeout_length #if not recording: diff = time.time() - start_time if not self.quiet: print('%s - recording at level %3d for %.1f seconds' % (now, rms_val, diff), end='\r') recording = True current = time.time() rec.append(data) rec_time = time.time() - start_time if rec_time > self.max_rec_time: #print("%s - maximum recording time of %d seconds reached" % (now, rec_time)) break else: rec_time = time.time() - start_time - self.timeout_length if os.path.exists(lock_file): print("%s - lock file %s detected" % (now, lock_file)) break if rec_time > self.min_rec_time: self.write(b''.join(rec), rec_time) else: if rec_time > 1: - print("Skip too short recording %.1f seconds" % rec_time) + print("%s - skip short recording of %.1f seconds" % (now, rec_time)) else: print() if self.started_by_file: - while os.path.exists(self.external_trigger_file): + while os.path.exists(self.activate_by_file): i = i + 1 data = self.stream.read(self.chunk) rms_val = self.rms(data) now = datetime.now().strftime('%Y-%m-%d %H:%M:%S') if not self.quiet: print('%s - recording by file at level %3d' % (now, rms_val), end='\r') recording = True current = time.time() rec.append(data) rec_time = time.time() - start_time if rec_time > self.max_rec_time: break if os.path.exists(lock_file): print("%s - lock file %s detected" % (now, lock_file)) break print("%s - recorded %.1f seconds" % (now, rec_time)) self.write(b''.join(rec), rec_time) self.started_by_file = False self.started_by_level = False def play(self, file): p = pyaudio.PyAudio() wf = wave.open(file, 'rb') stream = p.open( format = p.get_format_from_width(wf.getsampwidth()), channels = wf.getnchannels(), rate = wf.getframerate(), output = True ) data = wf.readframes(1024) while data != b'': stream.write(data) data = wf.readframes(1024) stream.close() p.terminate() def write(self, recording, duration): now = datetime.now().strftime('%Y-%m-%d %H:%M:%S') if os.path.exists(lock_file): print("%s - lock file %s present, skip file" % (now, lock_file)) return if self.external_lock_file and os.path.exists(self.external_lock_file): print("%s - external lock file %s present, skip file" % (now, self.external_lock_file)) return n_files = len(os.listdir(f_name_directory)) tmp_filename = os.path.join(f_name_directory, '%s.tmp' % self.target) wf = wave.open(tmp_filename, 'wb') wf.setnchannels(self.CHANNELS) wf.setsampwidth(self.p.get_sample_size(self.FORMAT)) wf.setframerate(self.RATE) wf.writeframes(recording) wf.close() filename = os.path.join(f_name_directory, '%s.wav' % self.target) os.rename(tmp_filename, filename) print('%s - saved %d seconds audio to %s' % (now, duration, filename)) if self.target == 'test': self.play(filename) def end(self): # Reset to default error handler self.p.terminate() asound.snd_lib_error_set_handler(None) if __name__ == '__main__': description = 'This script is a voice activate recorder that saves individual recording in the folder %s that is polled by sip-session to initiate an outgoing call and playback the file. The filename is in the format user@domain.wav. Use test argument to test audio level.' % f_name_directory usage = '%prog [options] [user@domain]' parser = OptionParser(usage=usage, description=description) parser.print_usage = parser.print_help parser.add_option('-r', '--sample_rate', type='int', default='16000', dest='rate', help='Audio sample rate') parser.add_option('-d', '--device', type='string', default='pulse', dest='device', help='Use selected input audio device') parser.add_option('-t', '--timeout', type='int', default=2, dest='timeout', help='Silence timeout to stop recording') parser.add_option('-m', '--min_rec_time', type='int', default=2, dest='min_rec_time', help='Minimum recording time to save recording') - parser.add_option('-M', '--max_rec_time', type='int', default=5, dest='max_rec_time', help='Maximum recording time for each file') - parser.add_option('-l', '--threshold', type='int', default=30, dest='threshold', help='Minimum signal level to start recording') + parser.add_option('-M', '--max_rec_time', type='int', default=15, dest='max_rec_time', help='Maximum recording time for each file') +# parser.add_option('-t', '--threshold', type='int', default=30, dest='threshold', help='Minimum signal level to start recording') parser.add_option('-e', '--external_lock_file', type='string', dest='external_lock_file', help='Skip recording if file exists') - parser.add_option('-i', '--external_trigger_file', type='string', dest='external_trigger_file', help='Start recording if file exists, regardless of level') + parser.add_option('-f', '--activate_by_file', type='string', dest='activate_by_file', help='Start recording if file exists, regardless of level') + parser.add_option('-l', '--activate_by_level', type='int', default=30, dest='activate_by_level', help='Start recording if sound exceeds this threshold level') parser.add_option('-q', '--quiet', action='store_true', dest='quiet', default=False, help='Minimize logging.') options, args = parser.parse_args() try: target = args[0] except IndexError: parser.print_help() sys.exit(1) try: a = Recorder(args[0], options) a.listen() except KeyboardInterrupt: a.end() print() sys.exit(0) except OSError as e: print("Error: %s" % str(e)) sys.exit(0)