From 67c7ad3d0fe3a3e481fc90aee29648d7c1978cd9 Mon Sep 17 00:00:00 2001
From: Shuming Liu <smshine@qq.com>
Date: Fri, 11 Sep 2026 12:21:02 +0800
Subject: [PATCH] fixed a bug

---
 sentence_api/transcription.py | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/sentence_api/transcription.py b/sentence_api/transcription.py
index 95583b2..ee6875a 100644
--- a/sentence_api/transcription.py
+++ b/sentence_api/transcription.py
@@ -99,7 +99,8 @@ class MossTranscriber:
         segments = _parse_json_segments(payload.get("segments"))
         if not segments:
             segments = _parse_compact_segments(raw_text)
-        plain_text = " ".join(segment.text for segment in segments).strip() or raw_text
+        plain_text = " ".join(segment.text for segment in segments).strip()
+        plain_text = _strip_speaker_tags(plain_text) or _strip_speaker_tags(raw_text)
         return Transcript(text=plain_text, segments=segments)
 
     def _transcribe_chunked(
@@ -142,7 +143,9 @@ class MossTranscriber:
                     )
                 )
         segments.sort(key=lambda item: (item.start_seconds, item.end_seconds))
-        plain_text = " ".join(segment.text for segment in segments).strip()
+        plain_text = _strip_speaker_tags(
+            " ".join(segment.text for segment in segments).strip()
+        )
         return Transcript(text=plain_text, segments=segments)
 
     def _post_audio(
@@ -249,6 +252,11 @@ _ABBREVIATION_RE = re.compile(
 _SPEAKER_TAG_RE = re.compile(r"\[(S\d+)\]")
 
 
+def _strip_speaker_tags(text: str) -> str:
+    """Remove inline ``[S01]`` speaker tags from a text string."""
+    return _SPEAKER_TAG_RE.sub("", text).strip()
+
+
 def _expand_speaker_segments(
     segment: TranscriptionSegment,
 ) -> List[TranscriptionSegment]: