ss

app.py

@@ -2585,9 +2585,9 @@ def format_beat_timeline(audio_file, lyrics=None):
         timeline += "S = Strong beat | m = Medium beat | w = Weak beat | · = No beat\n\n"
         
         # Calculate total duration and create time markers
-        if beats_info['beat_times']:
-            # Convert to float if it's a numpy array
-            max_beat_time = float(max(beats_info['beat_times'])) if isinstance(max(beats_info['beat_times']), np.ndarray) else max(beats_info['beat_times'])
+        if beats_info['beat_times'] and len(beats_info['beat_times']) > 0:
+            # Get the max value safely
+            max_beat_time = max([float(t) if isinstance(t, np.ndarray) else t for t in beats_info['beat_times']])
             total_duration = max_beat_time + 2  # Add 2 seconds of padding
         else:
             total_duration = 30  # Default duration if no beats found
@@ -2668,7 +2668,7 @@ def format_beat_timeline(audio_file, lyrics=None):
                 # Calculate end time (start of next measure or end of song)
                 if i < len(measure_starts) - 1:
                     end_time = measure_starts[i+1][1]
-                elif beats_info['beat_times']:
+                elif beats_info['beat_times'] and len(beats_info['beat_times']) > 0:
                     # Get the last beat time and convert to float if needed
                     last_beat = beats_info['beat_times'][-1]
                     end_time = float(last_beat) if isinstance(last_beat, np.ndarray) else last_beat
@@ -2685,17 +2685,19 @@ def format_beat_timeline(audio_file, lyrics=None):
                     break
         
         # Add phrase information
-        if beats_info['phrases']:
+        if 'phrases' in beats_info and beats_info['phrases']:
             timeline += "\n=== MUSICAL PHRASES ===\n\n"
             for i, phrase in enumerate(beats_info['phrases']):
                 if i < 10:  # Limit to first 10 phrases
                     if not phrase:
                         continue
-                        
-                    start_beat = phrase[0]
-                    end_beat = phrase[-1]
-                    if start_beat >= len(beats_info['beat_times']) or end_beat >= len(beats_info['beat_times']):
+                    
+                    # Safely check phrase indices
+                    if not (len(phrase) > 0 and len(beats_info['beat_times']) > 0):
                         continue
+                        
+                    start_beat = min(phrase[0], len(beats_info['beat_times'])-1)
+                    end_beat = min(phrase[-1], len(beats_info['beat_times'])-1)
                     
                     # Convert to float if needed
                     phrase_start = beats_info['beat_times'][start_beat]
@@ -2788,11 +2790,12 @@ def format_beat_timeline(audio_file, lyrics=None):
                 timeline += f"  Syllables: {syllable_count}\n"
                 
                 # Show ideal timing (if we have enough phrases)
-                if beats_info['phrases'] and i < len(beats_info['phrases']):
+                if 'phrases' in beats_info and beats_info['phrases'] and i < len(beats_info['phrases']):
                     phrase = beats_info['phrases'][i]
-                    if phrase and phrase[0] < len(beats_info['beat_times']) and phrase[-1] < len(beats_info['beat_times']):
-                        start_beat = phrase[0]
-                        end_beat = phrase[-1]
+                    # Safely check if phrase has elements and indices are valid
+                    if phrase and len(phrase) > 0 and len(beats_info['beat_times']) > 0:
+                        start_beat = min(phrase[0], len(beats_info['beat_times'])-1)
+                        end_beat = min(phrase[-1], len(beats_info['beat_times'])-1)
                         
                         start_time = beats_info['beat_times'][start_beat]
                         start_time = float(start_time) if isinstance(start_time, np.ndarray) else start_time
@@ -2814,7 +2817,11 @@ def format_beat_timeline(audio_file, lyrics=None):
                             if j < len(beats_info['beat_times']):
                                 beat_time = beats_info['beat_times'][j]
                                 beat_time = float(beat_time) if isinstance(beat_time, np.ndarray) else beat_time
-                                relative_pos = int((beat_time - start_time) / phrase_duration * syllable_count)
+                                # Handle edge case where phrase_duration is very small
+                                if phrase_duration > 0.001:  # Avoid division by very small numbers
+                                    relative_pos = int((beat_time - start_time) / phrase_duration * syllable_count)
+                                else:
+                                    relative_pos = 0
                                 
                                 while len(syllable_viz) <= relative_pos:
                                     syllable_viz.append("·")