Skip to content

FunASR-powered video transcription, subtitle generation, and LLM-assisted clipping tool with a local Gradio UI.

fatal bug fix

维石ed1b49b5999b
diff --git a/funclip/utils/subtitle_utils.py b/funclip/utils/subtitle_utils.py
index ebcba4b..43a615e 100644
--- a/funclip/utils/subtitle_utils.py
+++ b/funclip/utils/subtitle_utils.py
@@ -45,10 +45,10 @@ class Text2SRT():
def generate_srt(sentence_list):
srt_total = ''
- for i, d in enumerate(sentence_list):
- t2s = Text2SRT(d['text'], d['timestamp'])
- if 'spk' in d:
- srt_total += "{} spk{}\n{}".format(i, d['spk'], t2s.srt())
+ for i, sent in enumerate(sentence_list):
+ t2s = Text2SRT(sent['text'], sent['timestamp'])
+ if 'spk' in sent:
+ srt_total += "{} spk{}\n{}".format(i, sent['spk'], t2s.srt())
else:
srt_total += "{}\n{}".format(i, t2s.srt())
return srt_total
@@ -58,48 +58,54 @@ def generate_srt_clip(sentence_list, start, end, begin_index=0, time_acc_ost=0.0
srt_total = ''
cc = 1 + begin_index
subs = []
- for i, d in enumerate(sentence_list):
- if d['timestamp'][-1][1] <= start:
+ for _, sent in enumerate(sentence_list):
+ if sent['timestamp'][-1][1] <= start:
+ print("CASE0")
continue
- if d['timestamp'][0][0] >= end:
+ if sent['timestamp'][0][0] >= end:
+ print("CASE4")
break
# parts in between
- if (d['timestamp'][-1][1] <= end and d['timestamp'][0][0] > start) or (d['timestamp'][-1][1] == end and d['timestamp'][0][0] == start):
- t2s = Text2SRT(d['text'], d['timestamp'], offset=start)
+ if (sent['timestamp'][-1][1] <= end and sent['timestamp'][0][0] > start) or (sent['timestamp'][-1][1] == end and sent['timestamp'][0][0] == start):
+ print("CASE1")
+ t2s = Text2SRT(sent['text'], sent['timestamp'], offset=start)
srt_total += "{}\n{}".format(cc, t2s.srt(time_acc_ost))
subs.append((t2s.time(time_acc_ost), t2s.text()))
cc += 1
continue
- if d['timestamp'][0][0] <= start:
- if not d['timestamp'][-1][1] > end:
- for j, ts in enumerate(d['timestamp']):
+ if sent['timestamp'][0][0] <= start:
+ print("CASE2")
+ if not sent['timestamp'][-1][1] > end:
+ for j, ts in enumerate(sent['timestamp']):
if ts[1] > start:
break
- _text = " ".join(d['text'].split()[j:])
- _ts = d['timestamp'][j:]
+ _text = " ".join(sent['text'][j:])
+ _ts = sent['timestamp'][j:]
else:
- for j, ts in enumerate(d['timestamp']):
+ for j, ts in enumerate(sent['timestamp']):
if ts[1] > start:
_start = j
break
- for j, ts in enumerate(d['timestamp']):
+ for j, ts in enumerate(sent['timestamp']):
if ts[1] > end:
_end = j
break
- _text = " ".join(d['text'].split()[_start:_end])
- _ts = d['timestamp'][_start:_end]
+ _text = " ".join(sent['text'][_start:_end])
+ _ts = sent['timestamp'][_start:_end]
if len(ts):
t2s = Text2SRT(_text, _ts, offset=start)
srt_total += "{}\n{}".format(cc, t2s.srt(time_acc_ost))
subs.append((t2s.time(time_acc_ost), t2s.text()))
cc += 1
+ # import pdb; pdb.set_trace()
continue
- if d['timestamp'][-1][1] > end:
- for j, ts in enumerate(d['timestamp']):
+ if sent['timestamp'][-1][1] > end:
+ print("CASE3")
+ for j, ts in enumerate(sent['timestamp']):
if ts[1] > end:
break
- _text = " ".join(d['text'].split()[:j])
- _ts = d['timestamp'][:j]
+ _text = " ".join(sent['text'][:j])
+ _ts = sent['timestamp'][:j]
if len(_ts):
t2s = Text2SRT(_text, _ts, offset=start)
srt_total += "{}\n{}".format(cc, t2s.srt(time_acc_ost))
diff --git a/funclip/videoclipper.py b/funclip/videoclipper.py
index cdbcae3..9d8e4d9 100644
--- a/funclip/videoclipper.py
+++ b/funclip/videoclipper.py
@@ -121,6 +121,7 @@ class VideoClipper():
video = mpy.VideoFileClip(video_filename)
# Extract the base name, add '_clip.mp4', and 'wav'
if output_dir is not None:
+ os.makedirs(output_dir, exist_ok=True)
_, base_name = os.path.split(video_filename)
base_name, _ = os.path.splitext(base_name)
clip_video_file = base_name + '_clip.mp4'
@@ -233,7 +234,7 @@ class VideoClipper():
clip_video_file_name, _ = os.path.splitext(file_with_extension)
print(output_dir, clip_video_file)
clip_video_file = os.path.join(output_dir, "{}_no{}.mp4".format(clip_video_file_name, self.GLOBAL_COUNT))
- temp_audio_file = os.path.join(output_dir, "{}_no{}.mp3".format(clip_video_file_name, self.GLOBAL_COUNT))
+ temp_audio_file = os.path.join(output_dir, "{}_tempaudio_no{}.mp4".format(clip_video_file_name, self.GLOBAL_COUNT))
else:
clip_video_file = clip_video_file[:-4] + '_no{}.mp4'.format(self.GLOBAL_COUNT)
temp_audio_file = clip_video_file[:-4] + '_tempaudio_no{}.mp4'.format(self.GLOBAL_COUNT)