"""Bounded local attempt records; never mark an audio clip accepted.""" import json from pathlib import Path import re import subprocess class Attempts: def __init__(self, root, code, key, run, text_sha256): if any(not re.fullmatch('[a-zA-Z0-9_-]{1,160}', value) for value in (code, key, run)): raise ValueError('Invalid narration diagnostic identifier') self.path = Path(root) / 'backup/narration/diagnostics' / code / key / (run + '.json') subprocess.run(['git', '-C', str(root), 'check-ignore', '-q', str(self.path)], check=True, capture_output=True) self.data = dict(schema=1, locale=code, key=key, text_sha256=text_sha256, attempts=[]) def save(self, row): if len(self.data['attempts']) >= 4: raise ValueError('Too many narration diagnostic attempts') row = dict(row) for key, limit in (('transcript', 8192), ('error', 2048)): if isinstance(row.get(key), str) and len(row[key]) > limit: row[key] = row[key][:limit] row[key + '_truncated'] = True self.data['attempts'].append(row) payload = json.dumps(self.data, ensure_ascii=False, allow_nan=False, indent=2) + '\n' if len(payload.encode()) > 65536: raise ValueError('Oversized narration diagnostic record') self.path.parent.mkdir(parents=True, exist_ok=True) temporary = self.path.with_suffix('.json.part') temporary.write_text(payload) temporary.replace(self.path)