open-standards/aac/aac-rs/tools/voices/check_packs.py

#!/usr/bin/env python3
"""Check a voice-pack directory before a site build ships it.

  python3 tools/voices/check_packs.py DIR

DIR must hold what tools/voices/build_voices.py built for the committed
catalogue (data/voices/index.json): its index.json points at
index.<sha12>.json, whose bytes are the committed catalogue's, and every file
that catalogue names is there with its exact size and SHA-256. Otherwise this
exits non-zero and says what is wrong, so a build never ships packs that
disagree with the catalogue the site describes, or that are incomplete.

Standard library only; Python 3.9 or later (macOS's own python3).
"""

import hashlib
import json
import sys
from pathlib import Path

HERE = Path(__file__).resolve().parent
COMMITTED = HERE.parents[1] / "data" / "voices" / "index.json"
SHOW = 10


def sha256(path):
    h = hashlib.sha256()
    with open(path, "rb") as f:
        for chunk in iter(lambda: f.read(1 << 20), b""):
            h.update(chunk)
    return h.hexdigest()


def file_refs(cat):
    """Every FileRef the catalogue names: runtimes, components and packs,
    each with its licence text, and each pack's sample."""
    refs = []
    owners = list(cat.get("runtimes", {}).values()) + list(cat.get("components", {}).values())
    for lang in cat.get("languages", {}).values():
        owners += lang.get("packs", [])
    for owner in owners:
        refs += owner.get("files", [])
        for one in ("licence_text", "sample"):
            if one in owner:
                refs.append(owner[one])
    return refs


def check(out):
    committed = COMMITTED.read_bytes()
    want = "index.%s.json" % hashlib.sha256(committed).hexdigest()[:12]
    try:
        pointer = json.loads((out / "index.json").read_text(encoding="utf-8"))
    except (OSError, ValueError) as e:
        return ["%s: no readable index.json (%s)" % (out, e)]
    problems = []
    if pointer.get("current") != want:
        problems.append(
            "index.json points at %s, but the committed catalogue is %s: rebuild the packs "
            "with tools/voices/build_voices.py" % (pointer.get("current"), want)
        )
    catalogue = out / want
    if not catalogue.is_file() or catalogue.read_bytes() != committed:
        problems.append("%s is missing, or differs from data/voices/index.json" % want)
        return problems
    for ref in file_refs(json.loads(committed)):
        path = out / ref["url"]
        if not path.is_file():
            problems.append("missing: %s" % ref["url"])
        elif path.stat().st_size != ref["bytes"]:
            problems.append("%s is %d bytes, not %d" % (ref["url"], path.stat().st_size, ref["bytes"]))
        elif sha256(path) != ref["sha256"]:
            problems.append("%s does not match its SHA-256" % ref["url"])
    return problems


def main(argv):
    if len(argv) != 2:
        print(__doc__.strip(), file=sys.stderr)
        return 2
    out = Path(argv[1])
    problems = check(out)
    if problems:
        print("✗ voice packs in %s do not match the committed catalogue:" % out, file=sys.stderr)
        for p in problems[:SHOW]:
            print("  - " + p, file=sys.stderr)
        if len(problems) > SHOW:
            print("  ... and %d more" % (len(problems) - SHOW), file=sys.stderr)
        return 1
    print("✓ voice packs in %s match the committed catalogue" % out)
    return 0


if __name__ == "__main__":
    sys.exit(main(sys.argv))