feat: add SYNC command to merge entries with delta

merge_srt.py

@@ -63,18 +63,60 @@ def read_commands(filename):
             except (ValueError, AttributeError):
                 continue
             commands.append(('MAP', text_source, text_start, time_source, time_start, count))
+        elif command == 'SYNC':
+            if len(parts) != 3:
+                continue
+            try:
+                source1_index = int(parts[1])
+                source2_index = int(parts[2])
+            except (ValueError, AttributeError):
+                continue
+            commands.append(('SYNC', source1_index, source2_index))
         else:
             continue
     return commands
 
+def parse_timestamp(ts_str):
+    start_str, end_str = ts_str.split(' --> ')
+    def parse_part(part):
+        parts = part.split(':')
+        hours = int(parts[0])
+        minutes = int(parts[1])
+        sec_ms = parts[2].split(',')
+        seconds = int(sec_ms[0])
+        ms = int(sec_ms[1])
+        return hours * 3600000 + minutes * 60000 + seconds * 1000 + ms
+    return parse_part(start_str), parse_part(end_str)
+
+def compute_delta(prev_ts_str, curr_ts_str):
+    prev_start, _ = parse_timestamp(prev_ts_str)
+    curr_start, _ = parse_timestamp(curr_ts_str)
+    return curr_start - prev_start
+
+def format_time(ms):
+    total_seconds = ms // 1000
+    ms_part = ms % 1000
+    hours = total_seconds // 3600
+    minutes = (total_seconds // 60) % 60
+    seconds = total_seconds % 60
+    return hours, minutes, seconds, ms_part
+
+def format_timestamp(start_ms, end_ms):
+    def format_part(h, m, s, ms):
+        return f"{h:02}:{m:02}:{s:02},{ms:03}"
+    start_h, start_m, start_s, start_ms_part = format_time(start_ms)
+    end_h, end_m, end_s, end_ms_part = format_time(end_ms)
+    return f"{format_part(start_h, start_m, start_s, start_ms_part)} --> {format_part(end_h, end_m, end_s, end_ms_part)}"
+
+def add_delta_to_timestamp(ts_str, delta):
+    start, end = parse_timestamp(ts_str)
+    return format_timestamp(start + delta, end + delta)
+
 if __name__ == '__main__':
     if len(sys.argv) != 5:
         print("Usage: python merge_srt.py <srt1> <srt2> <commands> <output>")
         sys.exit(1)
-    srt1_filename = sys.argv[1]
-    srt2_filename = sys.argv[2]
-    commands_filename = sys.argv[3]
-    output_filename = sys.argv[4]
+    srt1_filename, srt2_filename, commands_filename, output_filename = sys.argv[1:5]
 
     srt1 = parse_srt(srt1_filename)
     srt2 = parse_srt(srt2_filename)
@@ -83,42 +125,47 @@ if __name__ == '__main__':
     output = []
     for cmd in commands:
         if cmd[0] == 'COPY':
-            _, source, start, end = cmd
+            source, start, end = cmd[1:]
             source_list = srt1 if source == 1 else srt2
-            start_idx = start - 1
-            end_idx = end - 1
+            start_idx, end_idx = start - 1, end - 1
             if start_idx < 0 or end_idx >= len(source_list) or start_idx > end_idx:
                 print(f"Skipping invalid COPY command: source {source}, range {start}-{end}")
                 continue
-            entries = source_list[start_idx:end_idx + 1]
-            output.extend(entries)
+            output.extend(source_list[start_idx:end_idx + 1])
         elif cmd[0] == 'MAP':
-            _, text_source, text_start, time_source, time_start, count = cmd
+            text_source, text_start, time_source, time_start, count = cmd[1:]
             text_list = srt1 if text_source == 1 else srt2
             time_list = srt1 if time_source == 1 else srt2
             text_start_idx = text_start - 1
             time_start_idx = time_start - 1
-            if text_start_idx < 0 or text_start_idx + count > len(text_list):
-                print(f"Skipping invalid MAP command: text source {text_source}, start {text_start}, count {count}")
-                continue
-            if time_start_idx < 0 or time_start_idx + count > len(time_list):
-                print(f"Skipping invalid MAP command: time source {time_source}, start {time_start}, count {count}")
+            if (text_start_idx < 0 or text_start_idx + count > len(text_list) or
+                time_start_idx < 0 or time_start_idx + count > len(time_list)):
+                print(f"Skipping invalid MAP command: text source {text_source}, start {text_start}, count {count} or time source {time_source}, start {time_start}, count {count}")
                 continue
             for i in range(count):
                 text_entry = text_list[text_start_idx + i]
                 time_entry = time_list[time_start_idx + i]
-                new_entry = {
+                output.append({
                     'index': len(output) + 1,
                     'timestamp': time_entry['timestamp'],
                     'text': text_entry['text'],
-                }
-                output.append(new_entry)
+                })
+        elif cmd[0] == 'SYNC':
+            source1_index, source2_index = cmd[1:]
+            if source1_index < 2 or source1_index > len(srt1) or source2_index < 1 or source2_index > len(srt2):
+                print(f"Skipping invalid SYNC command: source1 index {source1_index} must be >=2 and <= {len(srt1)}, source2 index {source2_index} must be >=1 and <= {len(srt2)}")
+                continue
+            entry_prev = srt1[source1_index - 2]
+            entry1 = srt1[source1_index - 1]
+            entry2 = srt2[source2_index - 1]
+            delta = compute_delta(entry_prev['timestamp'], entry1['timestamp'])
+            new_ts = add_delta_to_timestamp(entry2['timestamp'], delta)
+            output.append({
+                'index': len(output) + 1,
+                'timestamp': new_ts,
+                'text': entry2['text'],
+            })
 
     with open(output_filename, 'w', encoding='utf-8') as f:
-        for entry in output:
-            f.write(f"{entry['index']}\n")
-            f.write(entry['timestamp'] + '\n')
-            text = entry['text'].strip()
-            for line in text.split('\n'):
-                f.write(line + '\n')
-            f.write('\n')
+        for i, entry in enumerate(output, start=1):
+            f.write(f"{i}\n{entry['timestamp']}\n{entry['text'].strip()}\n\n")