Simple test script for whisper transcript

tavern_talk/playback.py

@@ -0,0 +1,70 @@
+import json
+
+from pydub import AudioSegment
+from pydub.playback import play
+
+PATH_TEXT = "./2024-07-29_audio.json"
+PATH_AUDIO = "./2024-07-29_audio.wav"
+
+
+def play_audio_segment(file_path, start_time, duration):
+    """
+    Plays a segment of an audio file starting at `start_time` for `duration` seconds.
+
+    Parameters:
+    - file_path (str): Path to the audio file.
+    - start_time (float): The starting point in seconds.
+    - duration (float): The duration of playback in seconds.
+    """
+    # Load only the necessary segment of the audio file
+    audio_segment = AudioSegment.from_file(file_path)
+    start_ms = start_time * 1000  # Convert seconds to milliseconds
+    end_ms = start_ms + (duration * 1000)
+
+    # Extract the desired portion of the audio
+    segment = audio_segment[start_ms:end_ms]
+
+    # Play the extracted audio segment
+    play(segment)
+
+
+segments = []
+with open(PATH_TEXT, "r") as fp:
+    data = json.load(fp)
+    segments = data["segments"]
+
+print(f"Transcript {PATH_TEXT} loaded.")
+
+while True:
+    snippet = input("What are you looking for: ")
+
+    if snippet == "":
+        exit(0)
+
+    print(f"Looking for '{snippet}' ...")
+    targets = []
+    for seg in segments:
+        if snippet.lower() in seg["text"].lower():
+            targets.append(seg)
+
+        # if len(targets) >= 5:
+        #     break
+
+    print(f"Found {len(targets)} targets.")
+    if len(targets) == 0:
+        continue
+
+    for i, seg in enumerate(targets):
+        print(f"{i} | {seg['text']}")
+
+    while True:
+        index = input("Index to play: ")
+        if index == "":
+            break
+
+        try:
+            index = int(index)
+        except ValueError:
+            continue
+
+        play_audio_segment(PATH_AUDIO, targets[index]["start"] - 0.5, 5)