feat: format export timestamps as MM:SS instead of seconds

backend/utils/formatters.py

@@ -37,6 +37,30 @@ def format_result(diarized: list) -> list[dict]:
     return full_transcript
 
 
+def format_timestamp(seconds_str: str) -> str:
+    """
+    Format timestamp from seconds to MM:SS format.
+
+    Args:
+        seconds_str: Time in seconds as string (e.g., "65.50")
+
+    Returns:
+        Formatted timestamp in MM:SS format (e.g., "1:05")
+
+    Example:
+        >>> format_timestamp("65.50")
+        '1:05'
+        >>> format_timestamp("5.25")
+        '0:05'
+    """
+    try:
+        total_seconds = max(0, int(float(seconds_str)))
+        minutes, seconds = divmod(total_seconds, 60)
+        return f"{minutes}:{seconds:02d}"
+    except (ValueError, TypeError):
+        return "0:00"
+
+
 def format_speaker_name(speaker_name: str) -> str:
     """
     Format speaker name from SPEAKER_XX format to 'Speaker X' format.

backend/utils/markdown_generator.py

@@ -8,7 +8,7 @@
 from io import BytesIO
 
 from config import Settings, get_settings
-from utils.formatters import format_speaker_name
+from utils.formatters import format_speaker_name, format_timestamp
 
 
 def generate_summary_markdown(
@@ -58,9 +58,9 @@ def generate_summary_markdown(
         for entry in transcript_data:
             speaker = format_speaker_name(entry.get('speaker', 'Unknown Speaker'))
             text = entry.get('text', '')
-            start_time = entry.get('start', '0.00')
-            end_time = entry.get('end', '0.00')
-            markdown_content += f"**{speaker}** *({start_time}s - {end_time}s)*: {text}\n\n"
+            start_time = format_timestamp(entry.get('start', '0.00'))
+            end_time = format_timestamp(entry.get('end', '0.00'))
+            markdown_content += f"**{speaker}** *({start_time} - {end_time})*: {text}\n\n"
 
     # Return as BytesIO buffer
     return BytesIO(markdown_content.encode('utf-8'))
@@ -107,9 +107,9 @@ def generate_transcript_markdown(
         for entry in transcript_data:
             speaker = format_speaker_name(entry.get('speaker', 'Unknown Speaker'))
             text = entry.get('text', '')
-            start_time = entry.get('start', '0.00')
-            end_time = entry.get('end', '0.00')
-            markdown_content += f"**{speaker}** *({start_time}s - {end_time}s)*: {text}\n\n"
+            start_time = format_timestamp(entry.get('start', '0.00'))
+            end_time = format_timestamp(entry.get('end', '0.00'))
+            markdown_content += f"**{speaker}** *({start_time} - {end_time})*: {text}\n\n"
 
     # Return as BytesIO buffer
     return BytesIO(markdown_content.encode('utf-8'))

backend/utils/pdf_generator.py

@@ -20,7 +20,7 @@
 from svglib.svglib import svg2rlg
 
 from config import Settings, get_settings
-from utils.formatters import format_speaker_name
+from utils.formatters import format_speaker_name, format_timestamp
 
 
 def _process_markdown_text(text: str) -> str:
@@ -308,10 +308,10 @@ def _add_transcript_section(
     for entry in transcript_data:
         speaker = format_speaker_name(entry.get('speaker', 'Unknown Speaker'))
         text = entry.get('text', '')
-        start_time = entry.get('start', '0.00')
-        end_time = entry.get('end', '0.00')
+        start_time = format_timestamp(entry.get('start', '0.00'))
+        end_time = format_timestamp(entry.get('end', '0.00'))
 
-        timestamp_text = f"[{start_time}s - {end_time}s]"
+        timestamp_text = f"[{start_time} - {end_time}]"
         speaker_line = f"<b>{speaker}</b> {timestamp_text}"
         story.append(Paragraph(speaker_line, styles['speaker']))
         story.append(Paragraph(text, styles['transcript']))