kangshifu-board/docs/workshop-report-design-20261109/finalize_local_revision.py

"""Validate and record this local report-design revision, preserving sealed releases."""
from pathlib import Path
from urllib.parse import unquote
import difflib
import hashlib
import json
import re

OUT = Path(__file__).resolve().parent
ROOT = OUT.parents[1]
SKILL = ROOT / 'skills/ksf-workshop-runtime'
CHANGED = {
    'SKILL.md',
    'references/workshop-feedback.md',
    'references/workshop-delivery.md',
    'references/workshop-innovation-reasoning.md',
    'references/workshop-verification.md',
}
ADDED = {'references/workshop-report-design.md'}


def sha(path):
    return hashlib.sha256(path.read_bytes()).hexdigest()


def main():
    baseline = json.loads((OUT / 'before-fingerprints.json').read_text(encoding='utf-8'))
    original_manifest = dict(line.split('  ', 1)[::-1] for line in
                             (OUT / 'before/FILES.sha256').read_text(encoding='utf-8').splitlines())
    assert baseline == original_manifest and len(baseline) == 41
    files = {p.relative_to(SKILL).as_posix() for p in SKILL.rglob('*') if p.is_file()}
    expected = set(baseline) | ADDED | {'FILES.sha256'}
    assert files == expected and len(files) == 43, files ^ expected
    changed = {name for name, digest in baseline.items() if sha(SKILL / name) != digest}
    assert changed == CHANGED, changed ^ CHANGED

    runtime = [name for name in baseline if name.startswith('scripts/') or name == 'runtime-lock.json']
    assert len(runtime) == 7
    assert all(sha(SKILL / name) == baseline[name] for name in runtime)
    for name in ('references/workshop-course-flow.md', 'references/workshop-intake.md',
                 'references/workshop-commercial-feasibility.md'):
        assert sha(SKILL / name) == baseline[name], name
    intake = (SKILL / 'references/workshop-intake.md').read_text(encoding='utf-8')
    assert len(re.findall(r'^\| D[12]-(?:\d+|N) ', intake, re.M)) == 13
    refs = [p for p in (SKILL / 'references').rglob('*') if p.is_file()]
    assert len(refs) == 34
    assert sum(p.suffix == '.md' for p in refs) == 32
    assert sum(p.suffix == '.json' for p in refs) == 2

    links = 0
    for path in SKILL.rglob('*.md'):
        content = path.read_text(encoding='utf-8')
        assert not re.search(r'豆包|doubao', content, re.I), path
        assert not re.search(r'付费|收费|计费|停费|paidCalls|零费用', content), path
        for target in re.findall(r'\[[^\]]*\]\(([^)\s]+)\)', content):
            if target.startswith(('https:', 'http:', '#')):
                continue
            target_path = unquote(target.split('#', 1)[0])
            if target_path:
                assert (path.parent / target_path).is_file(), (path, target)
                links += 1

    sealed = ROOT / 'docs/workshop-skill-release-20261109-r6/ksf-workshop-skill-update-20261109-r6.zip'
    sealed_hash = sha(sealed)
    assert sealed_hash == 'c329432b9773cfb0a8e5d4a3e5c162b3778eab28d576b79875da6854f33c4dc5'

    # The manifest excludes itself, retaining the previous public paths and adding one guide.
    manifest_names = sorted(set(baseline) | ADDED)
    manifest = ''.join(sha(SKILL / name) + '  ' + name + '\n' for name in manifest_names)
    (SKILL / 'FILES.sha256').write_text(manifest, encoding='utf-8', newline='\n')
    written = dict(line.split('  ', 1)[::-1] for line in
                   (SKILL / 'FILES.sha256').read_text(encoding='utf-8').splitlines())
    assert len(written) == 42 and set(written) | {'FILES.sha256'} == files
    assert all(sha(SKILL / name) == digest for name, digest in written.items())

    changes = []
    for name in sorted(CHANGED | ADDED | {'FILES.sha256'}):
        before = OUT / 'before' / name
        if name not in ADDED:
            assert before.is_file(), name
            if name != 'FILES.sha256':
                assert sha(before) == baseline[name], name
        old = before.read_text(encoding='utf-8') if before.is_file() else ''
        new = (SKILL / name).read_text(encoding='utf-8')
        target = OUT / 'diff' / (name + '.diff')
        target.parent.mkdir(parents=True, exist_ok=True)
        target.write_text(''.join(difflib.unified_diff(
            old.splitlines(True), new.splitlines(True),
            fromfile='before-report-design/' + name if before.is_file() else '/dev/null',
            tofile='local-report-design/' + name)), encoding='utf-8', newline='\n')
        changes.append({'path': name, 'action': 'add' if name in ADDED else 'update',
                        'beforeSha256': sha(before) if before.is_file() else None,
                        'afterSha256': sha(SKILL / name)})

    report = {
        'ok': True,
        'kind': 'local-instruction-and-report-design-draft',
        'comparisonBaseline': 'local draft immediately before report-design changes; not remote r6',
        'changes': changes,
        'skillSha256': sha(SKILL / 'SKILL.md'),
        'manifestSha256': sha(SKILL / 'FILES.sha256'),
        'publicFilesVerified': len(files), 'manifestEntriesVerified': len(written),
        'relativeFileLinksVerified': links, 'referenceFilesVerified': len(refs),
        'intakeStagesPreserved': 13, 'courseFlowAndIntakeUnchanged': True,
        'runtimeFilesVerifiedUnchanged': len(runtime),
        'businessSaveContractChanged': False,
        'platformBindingRemoved': True,
        'sealedR6PackageModified': False, 'sealedR6Sha256': sealed_hash,
        'remoteInstalled': False, 'researchExecuted': False,
        'employeeUsefulnessValidated': False,
    }
    (OUT / 'change-receipt.json').write_text(
        json.dumps(report, ensure_ascii=False, indent=2) + '\n', encoding='utf-8')
    print(json.dumps({k: v for k, v in report.items() if k != 'changes'}, ensure_ascii=False, indent=2))


if __name__ == '__main__':
    main()