Update app.py

app.py

@@ -7,15 +7,43 @@ import whisperx
 import torch
 
 def convert_segments_to_text(data):
-    segments = data['segments']
     result = []
-    for segment in segments:
-        speaker = segment['speaker']
-        start = segment['start']
-        end = segment['end']
-        text = segment['text']
-        formatted_text = f'{speaker} ({start} : {end}) : {text}'
-        result.append(formatted_text)
+    for segment in data['segments']:
+        words = segment['words']
+        current_speaker = None
+        current_start = None
+        current_end = None
+        current_text = []
+
+        for word_info in words:
+            word = word_info['word']
+            start = word_info['start']
+            end = word_info['end']
+            speaker = word_info['speaker']
+
+            if current_speaker is None:
+                current_speaker = speaker
+                current_start = start
+
+            if speaker == current_speaker:
+                current_text.append(word)
+                current_end = end
+            else:
+                # Finish current segment
+                formatted_text = f'{current_speaker} ({current_start} : {current_end}) : {" ".join(current_text)}'
+                result.append(formatted_text)
+
+                # Start new segment
+                current_speaker = speaker
+                current_start = start
+                current_end = end
+                current_text = [word]
+
+        # Append the last segment
+        if current_text:
+            formatted_text = f'{current_speaker} ({current_start} : {current_end}) : {" ".join(current_text)}'
+            result.append(formatted_text)
+
     return '\n'.join(result)
     
 st.title('Audio Transcription App')