Repository navigation
Expand file tree
/
Copy pathprepare_transcript_analysis.py
More file actions
346 lines (297 loc) · 11.5 KB
/
Copy pathprepare_transcript_analysis.py
File metadata and controls
346 lines (297 loc) · 11.5 KB
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
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
#!/usr/bin/env python3
"""
Prepare deterministic transcript artifacts for pi-driven editorial workflows.
Purpose
- Turn transcript outputs into chunked, timecoded artifacts that pi can reason over.
- Keep editorial judgment in pi while this script handles stable formatting and indexing.
- Standardize output names under dist/analysis/.
Usage
uv run python scripts/prepare_transcript_analysis.py <transcript-dir> [options]
Example
uv run python scripts/prepare_transcript_analysis.py \
dist/analysis/005/transcript \
--output-dir dist/analysis/005 \
--slug ep005 \
--plain-output-names
Outputs
- dist/analysis/<episode>/transcript_chunks.md
- dist/analysis/<episode>/transcript_index.json
"""
from __future__ import annotations
import argparse
import json
import math
import re
from pathlib import Path
from typing import Any
DEFAULT_OUTPUT_DIR = "dist/analysis"
DEFAULT_MAX_CHUNK_SECONDS = 300.0
DEFAULT_MAX_CHUNK_CHARS = 3500
DEFAULT_MAX_CHUNK_SEGMENTS = 80
def seconds_to_timecode(seconds: float) -> str:
total_ms = int(round(seconds * 1000.0))
ms = total_ms % 1000
total_s = total_ms // 1000
s = total_s % 60
m = (total_s // 60) % 60
h = total_s // 3600
if ms == 0:
return f"{h:02d}:{m:02d}:{s:02d}"
return f"{h:02d}:{m:02d}:{s:02d}.{ms:03d}"
def duration_to_label(seconds: float) -> str:
whole_seconds = int(math.floor(seconds + 0.5))
s = whole_seconds % 60
m = (whole_seconds // 60) % 60
h = whole_seconds // 3600
return f"{h:02d}:{m:02d}:{s:02d}"
def normalize_text(text: str) -> str:
return " ".join(text.strip().split())
def sanitize_slug(value: str) -> str:
slug = re.sub(r"[^a-zA-Z0-9]+", "_", value).strip("_").lower()
if not slug:
raise SystemExit("could not derive a slug; pass --slug explicitly")
return slug
def derive_slug(transcript_dir: Path) -> str:
name = transcript_dir.name
for suffix in ("_transcript", "-transcript", " transcript"):
if name.lower().endswith(suffix):
name = name[: -len(suffix)]
break
return sanitize_slug(name)
def load_segments(transcript_dir: Path) -> dict[str, Any]:
segments_path = transcript_dir / "segments.json"
if not segments_path.is_file():
raise SystemExit(f"segments.json not found in transcript dir: {transcript_dir}")
return json.loads(segments_path.read_text(encoding="utf-8"))
def normalize_segments(raw_segments: list[dict[str, Any]]) -> list[dict[str, Any]]:
segments: list[dict[str, Any]] = []
for idx, segment in enumerate(raw_segments, start=1):
start = float(segment.get("start", 0.0))
end = float(segment.get("end", start))
text = normalize_text(str(segment.get("text", "")))
if not text:
continue
segments.append(
{
"index": idx,
"start": start,
"end": end,
"start_timecode": seconds_to_timecode(start),
"end_timecode": seconds_to_timecode(end),
"duration_seconds": round(max(0.0, end - start), 3),
"text": text,
"char_count": len(text),
}
)
return segments
def chunk_segments(
segments: list[dict[str, Any]],
max_chunk_seconds: float,
max_chunk_chars: int,
max_chunk_segments: int,
) -> list[dict[str, Any]]:
chunks: list[dict[str, Any]] = []
current: list[dict[str, Any]] = []
def flush() -> None:
if not current:
return
start = float(current[0]["start"])
end = float(current[-1]["end"])
text = "\n".join(
f"[{segment['start_timecode']} -> {segment['end_timecode']}] {segment['text']}" for segment in current
)
chunks.append(
{
"chunk_index": len(chunks) + 1,
"start": start,
"end": end,
"start_timecode": seconds_to_timecode(start),
"end_timecode": seconds_to_timecode(end),
"duration_seconds": round(max(0.0, end - start), 3),
"segment_start_index": int(current[0]["index"]),
"segment_end_index": int(current[-1]["index"]),
"segment_count": len(current),
"char_count": sum(int(segment["char_count"]) for segment in current),
"text": text,
}
)
current.clear()
for segment in segments:
candidate = current + [segment]
start = float(candidate[0]["start"])
end = float(candidate[-1]["end"])
candidate_duration = end - start
candidate_chars = sum(int(item["char_count"]) for item in candidate)
candidate_count = len(candidate)
would_exceed = (
current
and (
candidate_duration > max_chunk_seconds
or candidate_chars > max_chunk_chars
or candidate_count > max_chunk_segments
)
)
if would_exceed:
flush()
current.append(segment)
flush()
return chunks
def build_index(
transcript_dir: Path,
slug: str,
data: dict[str, Any],
segments: list[dict[str, Any]],
chunks: list[dict[str, Any]],
index_path: Path,
chunks_path: Path,
max_chunk_seconds: float,
max_chunk_chars: int,
max_chunk_segments: int,
) -> dict[str, Any]:
duration_seconds = max((float(segment["end"]) for segment in segments), default=0.0)
return {
"schema_version": 1,
"slug": slug,
"source": {
"transcript_dir": str(transcript_dir),
"segments_json": str(transcript_dir / "segments.json"),
"transcript_txt": str(transcript_dir / "transcript.txt"),
"input_media": data.get("input"),
"backend": data.get("backend"),
"model": data.get("model"),
"language": data.get("language"),
},
"outputs": {
"transcript_chunks_markdown": str(chunks_path),
"transcript_index_json": str(index_path),
},
"chunking": {
"max_chunk_seconds": max_chunk_seconds,
"max_chunk_chars": max_chunk_chars,
"max_chunk_segments": max_chunk_segments,
},
"summary": {
"segment_count": len(segments),
"chunk_count": len(chunks),
"duration_seconds": round(duration_seconds, 3),
"duration_timecode": duration_to_label(duration_seconds),
},
"segments": segments,
"chunks": chunks,
}
def render_chunks_markdown(index: dict[str, Any]) -> str:
summary = index["summary"]
chunking = index["chunking"]
chunks = index["chunks"]
lines = [
f"# Prepared transcript chunks for {index['slug']}",
"",
"This file is deterministic prep for pi-driven editorial analysis.",
"",
f"- Source transcript dir: `{index['source']['transcript_dir']}`",
f"- Source segments: `{index['source']['segments_json']}`",
f"- Backend: `{index['source'].get('backend') or 'unknown'}`",
f"- Model: `{index['source'].get('model') or 'unknown'}`",
f"- Language: `{index['source'].get('language') or 'unknown'}`",
f"- Segment count: {summary['segment_count']}",
f"- Chunk count: {summary['chunk_count']}",
f"- Total duration: {summary['duration_timecode']}",
f"- Chunking: {chunking['max_chunk_seconds']}s / {chunking['max_chunk_chars']} chars / {chunking['max_chunk_segments']} segments max",
"",
"## Chunk index",
"",
]
for chunk in chunks:
lines.append(
"- "
f"Chunk {chunk['chunk_index']} — {chunk['start_timecode']} to {chunk['end_timecode']} "
f"(segments {chunk['segment_start_index']}-{chunk['segment_end_index']}, {chunk['char_count']} chars)"
)
for chunk in chunks:
lines.extend(
[
"",
f"## Chunk {chunk['chunk_index']}",
"",
f"- Time range: {chunk['start_timecode']} → {chunk['end_timecode']}",
f"- Duration: {duration_to_label(float(chunk['duration_seconds']))}",
f"- Segments: {chunk['segment_start_index']}-{chunk['segment_end_index']}",
f"- Segment count: {chunk['segment_count']}",
f"- Characters: {chunk['char_count']}",
"",
chunk["text"],
]
)
return "\n".join(lines).rstrip() + "\n"
def parse_args() -> argparse.Namespace:
parser = argparse.ArgumentParser(description="Prepare deterministic transcript artifacts for pi.")
parser.add_argument("transcript_dir", help="directory produced by scripts/transcribe_video.py")
parser.add_argument(
"--output-dir",
default=DEFAULT_OUTPUT_DIR,
help=f"directory to write prepared analysis outputs (default: {DEFAULT_OUTPUT_DIR})",
)
parser.add_argument("--slug", help="override the episode slug used in output filenames")
parser.add_argument(
"--plain-output-names",
action="store_true",
help="write transcript_chunks.md and transcript_index.json instead of slug-prefixed filenames",
)
parser.add_argument(
"--max-chunk-seconds",
type=float,
default=DEFAULT_MAX_CHUNK_SECONDS,
help=f"maximum duration per chunk in seconds (default: {DEFAULT_MAX_CHUNK_SECONDS})",
)
parser.add_argument(
"--max-chunk-chars",
type=int,
default=DEFAULT_MAX_CHUNK_CHARS,
help=f"maximum characters per chunk (default: {DEFAULT_MAX_CHUNK_CHARS})",
)
parser.add_argument(
"--max-chunk-segments",
type=int,
default=DEFAULT_MAX_CHUNK_SEGMENTS,
help=f"maximum segments per chunk (default: {DEFAULT_MAX_CHUNK_SEGMENTS})",
)
return parser.parse_args()
def resolve_output_paths(output_dir: Path, slug: str, plain_output_names: bool) -> tuple[Path, Path]:
if plain_output_names:
return output_dir / "transcript_index.json", output_dir / "transcript_chunks.md"
return output_dir / f"{slug}_transcript_index.json", output_dir / f"{slug}_transcript_chunks.md"
def main() -> None:
args = parse_args()
transcript_dir = Path(args.transcript_dir)
if not transcript_dir.is_dir():
raise SystemExit(f"transcript directory not found: {transcript_dir}")
output_dir = Path(args.output_dir)
output_dir.mkdir(parents=True, exist_ok=True)
slug = sanitize_slug(args.slug) if args.slug else derive_slug(transcript_dir)
index_path, chunks_path = resolve_output_paths(output_dir, slug, args.plain_output_names)
raw_data = load_segments(transcript_dir)
segments = normalize_segments(list(raw_data.get("segments", [])))
chunks = chunk_segments(
segments,
max_chunk_seconds=args.max_chunk_seconds,
max_chunk_chars=args.max_chunk_chars,
max_chunk_segments=args.max_chunk_segments,
)
index = build_index(
transcript_dir=transcript_dir,
slug=slug,
data=raw_data,
segments=segments,
chunks=chunks,
index_path=index_path,
chunks_path=chunks_path,
max_chunk_seconds=args.max_chunk_seconds,
max_chunk_chars=args.max_chunk_chars,
max_chunk_segments=args.max_chunk_segments,
)
index_path.write_text(json.dumps(index, indent=2) + "\n", encoding="utf-8")
chunks_path.write_text(render_chunks_markdown(index), encoding="utf-8")
print(f"Wrote {index_path}")
print(f"Wrote {chunks_path}")
if __name__ == "__main__":
main()