Files
Investidor-V2/tools/demography-requirements-sources.py
T

43 lines
2.4 KiB
Python

"""Read-only D01-D06 QA slices and cell verification; never modifies INE workbooks."""
import argparse
import hashlib
import json
from pathlib import Path
from demography_workbook import parse
cli = argparse.ArgumentParser()
cli.add_argument('directory', type=Path)
cli.add_argument('--output', type=Path, required=True)
args = cli.parse_args()
args.output.mkdir(parents=True, exist_ok=True)
slices, references = [], []
for name, territory, cell in [('Gaza.xlsx', 'Bilene', None),
('Inhambane.xlsx', 'Funhalouro', 'B54'),
('Maputo Provincia.xlsx', 'Moamba', 'B179')]:
original = args.directory / name
before = hashlib.sha256(original.read_bytes()).hexdigest()
parsed = parse(original)
rows = [r for r in parsed['records'] if r['territory'] == territory
and 2026 <= r['year'] <= 2028 and r['sex'] == 'all']
unchanged = before == hashlib.sha256(original.read_bytes()).hexdigest()
if not unchanged:
raise RuntimeError('Source changed during verification')
slices.append({'file': name, 'sha256': before, 'source_unchanged': unchanged, 'records': rows})
if cell:
matched = [r for r in rows if r['sheet'] == '2026' and r['cell'] == cell]
if len(matched) != 1:
raise RuntimeError('Reference not uniquely verified')
references.append({'file': name, 'sha256': before, 'source_unchanged': unchanged,
'state': 'verified_from_original_labels', 'observation': matched[0]})
initial = args.directory.parent / 'Inhambane.xlsx'
if initial.is_file():
before = hashlib.sha256(initial.read_bytes()).hexdigest()
matched = [r for r in parse(initial)['records'] if r['sheet'] == '2026' and r['cell'] == 'B54']
if len(matched) != 1 or before != hashlib.sha256(initial.read_bytes()).hexdigest():
raise RuntimeError('Initial Inhambane exemplar not verified')
references.append({'file': '../Inhambane.xlsx (exemplar inicial distinto)', 'sha256': before,
'source_unchanged': True, 'state': 'verified_from_original_labels', 'observation': matched[0]})
(args.output / 'source-slices.json').write_text(json.dumps(slices, ensure_ascii=False, indent=2), encoding='utf-8')
(args.output / 'source-references.json').write_text(json.dumps(references, ensure_ascii=False, indent=2), encoding='utf-8')
print(json.dumps({'files': len(slices), 'verified_cells': len(references), 'originals_unchanged': True}))