sebastjan/reels-app
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Subtitles: cap chunk duration at 2.5s, split long lines into multiple time slices for faster reels pacing

Browse Source
This commit is contained in:
Sebastjan Artič 2026-04-29 05:59:36 +00:00
parent e001387a89
commit ba787744a6
1 changed files with 24 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
scripts/subtitle.py
View File

@ -39,14 +39,32 @@ def transcribe(video, lang=None, model_size="small"):
sec = s % 60 sec = s % 60
return f"{h:02d}:{m:02d}:{sec:06.3f}".replace(".", ",") return f"{h:02d}:{m:02d}:{sec:06.3f}".replace(".", ",")
# Generiramo word-level chunked podnapise: 3-5 besed naenkrat # Generiramo word-level chunked podnapise: 3-4 besede naenkrat,
# Vse v VELIKIH TISKANIH ČRKAH za reels stil # max 2.5s na chunk za hiter pacing v reels stil.
# Vse v VELIKIH TISKANIH ČRKAH.
MAX_CHUNK_DURATION = 2.5 # sekund
def write_chunk(idx, start, end, text):
# Če je chunk daljši od MAX, razdeli ga na N enakih kosov z istim tekstom
duration = end - start
if duration <= MAX_CHUNK_DURATION:
srt_path.write(f"{idx}\n{fmt_ts(start)} --> {fmt_ts(end)}\n{text}\n\n")
return idx + 1
# Razdeli na N enakih kosov
n_parts = int(duration / MAX_CHUNK_DURATION) + 1
chunk_dur = duration / n_parts
for i in range(n_parts):
chunk_start = start + i * chunk_dur
chunk_end = start + (i + 1) * chunk_dur
srt_path.write(f"{idx}\n{fmt_ts(chunk_start)} --> {fmt_ts(chunk_end)}\n{text}\n\n")
idx += 1
return idx
idx = 1 idx = 1
for seg in segments: for seg in segments:
words = seg.words or [] words = seg.words or []
if not words: if not words:
srt_path.write(f"{idx}\n{fmt_ts(seg.start)} --> {fmt_ts(seg.end)}\n{seg.text.strip().upper()}\n\n") idx = write_chunk(idx, seg.start, seg.end, seg.text.strip().upper())
idx += 1
continue continue
# Združi v skupine po ~4 besede # Združi v skupine po ~4 besede
@ -57,15 +75,13 @@ def transcribe(video, lang=None, model_size="small"):
start = group[0].start start = group[0].start
end = group[-1].end end = group[-1].end
text = "".join(g.word for g in group).strip().upper() text = "".join(g.word for g in group).strip().upper()
srt_path.write(f"{idx}\n{fmt_ts(start)} --> {fmt_ts(end)}\n{text}\n\n") idx = write_chunk(idx, start, end, text)
idx += 1
group = [] group = []
if group: if group:
start = group[0].start start = group[0].start
end = group[-1].end end = group[-1].end
text = "".join(g.word for g in group).strip().upper() text = "".join(g.word for g in group).strip().upper()
srt_path.write(f"{idx}\n{fmt_ts(start)} --> {fmt_ts(end)}\n{text}\n\n") idx = write_chunk(idx, start, end, text)
idx += 1
srt_path.close() srt_path.close()
print(f"📝 SRT: {srt_path.name} ({idx - 1} segmentov)") print(f"📝 SRT: {srt_path.name} ({idx - 1} segmentov)")