#!/usr/bin/env python3 import argparse import csv import json from pathlib import Path from typing import Dict, List, Tuple DEFAULT_COLUMNS = [ 'video_id', 'sign_language', 'title', 'duration_sec', 'start_sec', 'end_sec', 'subtitle_languages', 'subtitle_dir_path', 'subtitle_en_source', 'raw_video_path', 'raw_metadata_path', 'metadata_status', 'subtitle_status', 'download_status', 'failure_count', 'error', 'processed_at', 'subtitle_json_path', 'subtitle_en', 'subtitle_texts_json', 'process_status', 'upload_status', 'local_cleanup_status', 'archive_name', ] VIDEO_EXTS = {'.mp4', '.mkv', '.webm', '.mov'} def read_csv_rows(path: Path) -> Tuple[List[Dict[str, str]], List[str]]: text = path.read_text(encoding='utf-8-sig') lines = [line for line in text.splitlines() if line.strip()] if not lines: return [], [] first = next(csv.reader([lines[0]]), []) first0 = first[0].strip() if first else '' if first0 == 'video_id' or 'download_status' in first or 'process_status' in first: with path.open('r', encoding='utf-8-sig', newline='') as f: reader = csv.DictReader(f) rows = [dict(r) for r in reader] fieldnames = list(reader.fieldnames or []) return rows, fieldnames rows: List[Dict[str, str]] = [] with path.open('r', encoding='utf-8-sig', newline='') as f: reader = csv.reader(f) for parts in reader: if not parts: continue video_id = (parts[0] or '').strip() sign_language = (parts[1] or '').strip() if len(parts) > 1 else '' if not video_id: continue rows.append({'video_id': video_id, 'sign_language': sign_language}) return rows, ['video_id', 'sign_language'] def write_csv_rows(path: Path, rows: List[Dict[str, str]], fieldnames: List[str]): tmp = path.with_suffix(path.suffix + '.tmp') with tmp.open('w', encoding='utf-8', newline='') as f: writer = csv.DictWriter(f, fieldnames=fieldnames) writer.writeheader() for row in rows: writer.writerow({k: row.get(k, '') for k in fieldnames}) tmp.replace(path) def load_progress(progress_path: Path): if not progress_path.exists(): return {}, {} obj = json.loads(progress_path.read_text()) return obj.get('uploaded_folders', {}), obj.get('archives', {}) def load_journal(journal_path: Path): updates: Dict[str, Dict[str, str]] = {} if not journal_path.exists(): return updates for line in journal_path.read_text(encoding='utf-8').splitlines(): line = line.strip() if not line: continue try: obj = json.loads(line) except Exception: continue row_updates = {k: str(v) for k, v in (obj.get('updates') or {}).items()} for vid in obj.get('video_ids') or []: updates[str(vid)] = row_updates return updates def main(): ap = argparse.ArgumentParser() ap.add_argument('--source-metadata-csv', type=Path, required=True) ap.add_argument('--output-metadata-csv', type=Path, required=True) ap.add_argument('--raw-video-dir', type=Path, required=True) ap.add_argument('--raw-caption-dir', type=Path, required=True) ap.add_argument('--raw-metadata-dir', type=Path, required=True) ap.add_argument('--dataset-dir', type=Path, required=True) ap.add_argument('--progress-path', type=Path, required=True) ap.add_argument('--status-journal-path', type=Path, required=True) args = ap.parse_args() source_rows, source_fields = read_csv_rows(args.source_metadata_csv) output_rows, output_fields = (read_csv_rows(args.output_metadata_csv) if args.output_metadata_csv.exists() else ([], [])) fields: List[str] = [] for col in DEFAULT_COLUMNS + source_fields + output_fields: if col and col not in fields: fields.append(col) out_by_id = {r.get('video_id', '').strip(): r for r in output_rows if r.get('video_id', '').strip()} rows: List[Dict[str, str]] = [] for src in source_rows: vid = (src.get('video_id') or '').strip() if not vid: continue merged = {k: src.get(k, '') for k in fields} if vid in out_by_id: for k in fields: if k in out_by_id[vid]: merged[k] = out_by_id[vid].get(k, '') rows.append(merged) raw_videos = {p.stem: p for p in args.raw_video_dir.iterdir() if p.is_file() and p.suffix.lower() in VIDEO_EXTS} if args.raw_video_dir.exists() else {} raw_metadata = {p.stem: p for p in args.raw_metadata_dir.glob('*.json')} if args.raw_metadata_dir.exists() else {} complete = {p.parent.parent.name for p in args.dataset_dir.glob('*/npz/.complete')} if args.dataset_dir.exists() else set() process_claims_dir = args.dataset_dir.parent / 'slurm' / 'state' / 'claims' download_claims_dir = args.dataset_dir.parent / 'slurm' / 'state' / 'download_claims' process_claims = {p.stem for p in process_claims_dir.glob('*.claim')} if process_claims_dir.exists() else set() download_claims = {p.stem for p in download_claims_dir.glob('*.claim')} if download_claims_dir.exists() else set() uploaded_folders, _archives = load_progress(args.progress_path) journal_updates = load_journal(args.status_journal_path) for row in rows: vid = (row.get('video_id') or '').strip() if not vid: continue if vid in raw_metadata: row['raw_metadata_path'] = str(raw_metadata[vid]) row['metadata_status'] = 'ok' if vid in raw_videos: row['raw_video_path'] = str(raw_videos[vid]) row['download_status'] = 'ok' elif vid in download_claims and row.get('download_status', '') not in {'ok', 'skipped'}: row['download_status'] = 'running' if vid in complete: row['process_status'] = 'ok' elif vid in process_claims and row.get('process_status', '') != 'ok': row['process_status'] = 'running' if vid in uploaded_folders: row['upload_status'] = 'uploaded' row['archive_name'] = uploaded_folders[vid] row['local_cleanup_status'] = 'deleted' row['process_status'] = 'ok' row['download_status'] = 'ok' if not row.get('metadata_status'): row['metadata_status'] = 'ok' elif vid in complete: row['upload_status'] = '' row['archive_name'] = '' row['local_cleanup_status'] = '' elif vid in journal_updates: for k, v in journal_updates[vid].items(): if k in {'upload_status', 'archive_name', 'local_cleanup_status'}: row[k] = v write_csv_rows(args.output_metadata_csv, rows, fields) print(f'synced_rows={len(rows)}') if __name__ == '__main__': main()