Spaces:
Sleeping
Sleeping
File size: 5,822 Bytes
b165696 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 | """
Parse VTT and SRT subtitle files into timed segments.
Supports:
- WebVTT (.vtt) — primary format from YouTube via yt-dlp
- SubRip (.srt) — fallback format
"""
from __future__ import annotations
import logging
import re
from pathlib import Path
logger = logging.getLogger(__name__)
_SRT_TIMESTAMP_RE = re.compile(
r"(\d{2}):(\d{2}):(\d{2})[,.](\d{3})\s*-->\s*(\d{2}):(\d{2}):(\d{2})[,.](\d{3})"
)
_VTT_TIMESTAMP_RE = re.compile(
r"(\d{1,2}):(\d{2}):(\d{2})\.(\d{3})\s*-->\s*(\d{1,2}):(\d{2}):(\d{2})\.(\d{3})"
)
_VTT_SHORTTS_RE = re.compile(
r"(\d{2}):(\d{2})\.(\d{3})\s*-->\s*(\d{2}):(\d{2})\.(\d{3})"
)
_HTML_TAG_RE = re.compile(r"<[^>]+>")
_CUE_META_RE = re.compile(r"^(align|line|position|size):", re.IGNORECASE)
def _ts_to_seconds(h: str, m: str, s: str, ms: str) -> float:
return int(h) * 3600 + int(m) * 60 + int(s) + int(ms) / 1000
class SubtitleParser:
def parse(self, path: str) -> list[dict]:
"""
Auto-detect format and parse subtitle file.
Returns
-------
list of { "start": float, "end": float, "text": str }
"""
path = Path(path)
if not path.exists():
raise FileNotFoundError(f"Subtitle file not found: {path}")
content = path.read_text(encoding="utf-8", errors="replace")
if path.suffix.lower() == ".vtt" or content.strip().startswith("WEBVTT"):
return self._parse_vtt(content)
return self._parse_srt(content)
# ------------------------------------------------------------------
# VTT
# ------------------------------------------------------------------
def _parse_vtt(self, content: str) -> list[dict]:
segments = []
lines = content.splitlines()
i = 0
while i < len(lines):
line = lines[i].strip()
# Try full HH:MM:SS.mmm --> HH:MM:SS.mmm
m = _VTT_TIMESTAMP_RE.search(line)
if m:
start = _ts_to_seconds(m.group(1), m.group(2), m.group(3), m.group(4))
end = _ts_to_seconds(m.group(5), m.group(6), m.group(7), m.group(8))
i += 1
text_lines = []
while i < len(lines) and lines[i].strip():
raw = lines[i].strip()
if not _CUE_META_RE.match(raw):
cleaned = _HTML_TAG_RE.sub("", raw).strip()
if cleaned:
text_lines.append(cleaned)
i += 1
text = " ".join(text_lines).strip()
if text:
segments.append({"start": start, "end": end, "text": text})
continue
# Try short MM:SS.mmm --> MM:SS.mmm
m2 = _VTT_SHORTTS_RE.search(line)
if m2:
start = _ts_to_seconds("0", m2.group(1), m2.group(2), m2.group(3))
end = _ts_to_seconds("0", m2.group(4), m2.group(5), m2.group(6))
i += 1
text_lines = []
while i < len(lines) and lines[i].strip():
raw = lines[i].strip()
if not _CUE_META_RE.match(raw):
cleaned = _HTML_TAG_RE.sub("", raw).strip()
if cleaned:
text_lines.append(cleaned)
i += 1
text = " ".join(text_lines).strip()
if text:
segments.append({"start": start, "end": end, "text": text})
continue
i += 1
logger.info("VTT parsed: %d segments", len(segments))
return self._merge_overlapping(segments)
# ------------------------------------------------------------------
# SRT
# ------------------------------------------------------------------
def _parse_srt(self, content: str) -> list[dict]:
segments = []
blocks = re.split(r"\n\n+", content.strip())
for block in blocks:
lines = block.strip().splitlines()
if len(lines) < 2:
continue
# Find timestamp line
ts_line = None
text_start = 0
for idx, l in enumerate(lines):
m = _SRT_TIMESTAMP_RE.search(l)
if m:
ts_line = m
text_start = idx + 1
break
if ts_line is None:
continue
start = _ts_to_seconds(
ts_line.group(1), ts_line.group(2), ts_line.group(3), ts_line.group(4)
)
end = _ts_to_seconds(
ts_line.group(5), ts_line.group(6), ts_line.group(7), ts_line.group(8)
)
text = " ".join(
_HTML_TAG_RE.sub("", l).strip()
for l in lines[text_start:]
if l.strip()
).strip()
if text:
segments.append({"start": start, "end": end, "text": text})
logger.info("SRT parsed: %d segments", len(segments))
return self._merge_overlapping(segments)
# ------------------------------------------------------------------
# Helpers
# ------------------------------------------------------------------
def _merge_overlapping(self, segments: list[dict]) -> list[dict]:
"""Merge consecutive segments with identical text (YouTube often duplicates cues)."""
if not segments:
return []
merged = [segments[0].copy()]
for seg in segments[1:]:
prev = merged[-1]
if seg["text"] == prev["text"] and seg["start"] <= prev["end"] + 0.1:
prev["end"] = max(prev["end"], seg["end"])
else:
merged.append(seg.copy())
return merged
|