Add back words-in-cue support

src/edge_tts/data_classes.py

@@ -83,6 +83,7 @@ class UtilArgs(argparse.Namespace):
     rate: str
     volume: str
     pitch: str
+    words_in_cue: int
     write_media: str
     write_subtitles: str
     proxy: str

src/edge_tts/submaker.py

@@ -37,6 +37,38 @@ def feed(self, msg: TTSChunk) -> None:
             )
         )
 
+    def merge_cues(self, words: int) -> None:
+        """
+        Merge cues to reduce the number of cues.
+
+        Args:
+            words (int): The number of words to merge.
+
+        Returns:
+            None
+        """
+        if words <= 0:
+            raise ValueError("Invalid number of words to merge, expected > 0")
+
+        if len(self.cues) == 0:
+            return
+
+        new_cues: List[srt.Subtitle] = []  # type: ignore
+        current_cue: srt.Subtitle = self.cues[0]  # type: ignore
+        for cue in self.cues[1:]:
+            if len(current_cue.content.split()) < words:
+                current_cue = srt.Subtitle(
+                    index=current_cue.index,
+                    start=current_cue.start,
+                    end=cue.end,
+                    content=current_cue.content + " " + cue.content,
+                )
+            else:
+                new_cues.append(current_cue)
+                current_cue = cue
+        new_cues.append(current_cue)
+        self.cues = new_cues
+
     def get_srt(self) -> str:
         """
         Get the SRT formatted subtitles from the SubMaker object.

src/edge_tts/util.py

@@ -75,6 +75,9 @@ async def _run_tts(args: UtilArgs) -> None:
             elif chunk["type"] == "WordBoundary":
                 submaker.feed(chunk)
 
+        if args.words_in_cue > 0:
+            submaker.merge_cues(args.words_in_cue)
+
         if sub_file is not None:
             sub_file.write(submaker.get_srt())
     finally:
@@ -107,6 +110,12 @@ async def amain() -> None:
     parser.add_argument("--rate", help="set TTS rate. Default +0%%.", default="+0%")
     parser.add_argument("--volume", help="set TTS volume. Default +0%%.", default="+0%")
     parser.add_argument("--pitch", help="set TTS pitch. Default +0Hz.", default="+0Hz")
+    parser.add_argument(
+        "--words-in-cue",
+        help="number of words in a subtitle cue. Default: 10.",
+        default=10,
+        type=int,
+    )
     parser.add_argument(
         "--write-media", help="send media output to file instead of stdout"
     )