info-tech-canon/src/info_tech_canon/maintenance.py

109 lines
4.9 KiB
Python
Raw Normal View History

"""Reproducible projections, portable contract exports, and read measurements."""
from collections import Counter
from dataclasses import replace
from datetime import datetime, timezone
import hashlib
import io
import json
from pathlib import Path
import shutil
import statistics
import subprocess
import tarfile
import tempfile
import time
from . import generation
def scope_inventory(context) -> dict:
return {"artifact_count": len(context.infospace.artifacts),
"kinds": dict(sorted(Counter(a.kind for a in context.infospace.artifacts).items())),
"standards": sorted(a.id for a in context.infospace.artifacts if a.kind == "standard"),
"models": sorted(a.id for a in context.infospace.artifacts if a.kind == "model")}
def check_generated(context) -> dict:
"""Render only into a disposable copy; never repair the source as a check."""
with tempfile.TemporaryDirectory(prefix="canon-freshness-") as temporary:
target = Path(temporary) / "infospace"
shutil.copytree(context.infospace_root, target)
copied = replace(context, infospace_root=target)
files = {}
for render in (generation.generate_indexes, generation.generate_tree,
generation.generate_agent_briefs):
for item in render(copied)["files"]:
path = Path(item["path"])
files[str(path.relative_to(target))] = path.read_bytes()
stale = []
for relative, expected in files.items():
original = context.infospace_root / relative
if not original.is_file() or original.read_bytes() != expected:
stale.append(relative)
return {"ok": not stale, "checked": len(files), "stale": sorted(stale)}
def source_evidence(root: Path) -> dict:
digest = hashlib.sha256()
# Include source and generated corpus; exclude reports to avoid self-hashing.
for path in sorted(root.rglob("*")):
if path.is_file() and "validation" not in path.relative_to(root).parts:
digest.update(str(path.relative_to(root)).encode() + b"\0" + path.read_bytes() + b"\0")
def git(*args):
try:
result = subprocess.run(["git", "-C", str(root.parent), *args],
capture_output=True, text=True, check=False)
except FileNotFoundError:
return None
return result.stdout.strip() if result.returncode == 0 else None
status = git("status", "--porcelain")
return {"generated_at": datetime.now(timezone.utc).isoformat(),
"revision": git("rev-parse", "HEAD"),
"dirty": bool(status) if status is not None else None,
"corpus_sha256": digest.hexdigest(),
"digest_scope": "all infospace files except validation reports"}
def export_emission_bundle(root: Path, destination: Path) -> dict:
"""Content-addressed deterministic tar; a changed contract gets a new name."""
paths = ["schemas/emission-cadence.schema.yaml",
"standards/emission-cadence/InfoTechCanonEmissionCadenceStandard.md",
"standards/emission-cadence/examples/qonto-assistant.yaml"]
files = {path: (root / path).read_bytes() for path in paths}
manifest = {"contract": "emission-cadence/0.1", "status": "draft",
"files": {name: hashlib.sha256(data).hexdigest() for name, data in files.items()},
"semantic_checks": ["unique source_id in sources"],
"adoption_evidence": "Two independent source-owned implementations required; example does not count."}
files["manifest.json"] = (json.dumps(manifest, sort_keys=True, indent=2) + "\n").encode()
output = io.BytesIO()
with tarfile.open(fileobj=output, mode="w", format=tarfile.USTAR_FORMAT) as archive:
for name, data in sorted(files.items()):
info = tarfile.TarInfo(name)
info.size = len(data)
info.mode = 0o644
archive.addfile(info, io.BytesIO(data))
data = output.getvalue()
digest = hashlib.sha256(data).hexdigest()
destination.mkdir(parents=True, exist_ok=True)
path = destination / f"emission-cadence-0.1-{digest}.tar"
try:
with path.open("xb") as stream:
stream.write(data)
except FileExistsError:
if path.read_bytes() != data:
raise ValueError(f"Refusing to overwrite different bundle contents: {path}")
return {"ok": True, "path": str(path), "sha256": digest, "manifest": manifest}
def benchmark_reads(root: Path, runs: int = 10) -> dict:
from .service import inspect_canon
samples = []
for _ in range(runs):
started = time.perf_counter()
inspect_canon(root)
samples.append((time.perf_counter() - started) * 1000)
return {"ok": True, "operation": "inspect", "runs": runs,
"median_ms": statistics.median(samples), "max_ms": max(samples),
"cache": "none; benchmark does not establish a production latency requirement"}