#!/usr/bin/env python3
"""Regenerate the φ_PRG anchor-map known-answer-test fixture.

Writes ``bench/fixtures/phi-prg/known-answer-tests.jsonl`` from the
fixed (seed, hard_hash, dim_h) list below. Each entry records the
SHA-256 of the raw HMAC-SHA-512 expansion output (before float
conversion) — the durable contract; the float layout can change
without invalidating the fixture (the byte stream doesn't).

Run this whenever the φ_PRG algorithm changes — and then bump
``PHI_PRG_VERSION`` in ``arborist/substrate/anchor_prg.py`` so the
fixture's ``version`` field changes too. The KAT regression test
(``tests/test_anchor_prg.py::test_phi_prg_known_answer_tests``)
pins these values and asserts the version matches the module.

Usage::

    python -m scripts.generate_phi_prg_kat       # from repo root
    # or:  python scripts/generate_phi_prg_kat.py

Config list (10 entries): the placeholder seed against zero / all-
ones hashes at several dim_h; a random (seed, hash) pair with two
one-bit-flip variants (avalanche cross-check); block-boundary cases
(dim_h=16 = exactly one HMAC block; dim_h=17 = two blocks with
truncation); a 4096-element counter-rollover stress sample. Note:
at counter=0 the bytes are identical regardless of endianness, so
the single-block (dim_h ≤ 16) entries are endianness-invariant; the
multi-block entries pin the little-endian counter encoding.
"""

from __future__ import annotations

import hashlib
import json
import pathlib
import sys

sys.path.insert(0, str(pathlib.Path(__file__).resolve().parents[1]))

from arborist.substrate.anchor_prg import _expand, PHI_PRG_VERSION  # noqa: E402

_FIXTURE = (
    pathlib.Path(__file__).resolve().parents[1]
    / "bench" / "fixtures" / "phi-prg" / "known-answer-tests.jsonl"
)

# (label, seed_hex, hard_hash_hex, dim_h)
_ENTRIES: list[tuple[str, str, str, int]] = [
    ("placeholder-seed/zero-hash/dim_h=1",
     "ef532720a49159beb6816d98e13a162bac63c531b631bd1adb0fcca96b467ff3",
     "0000000000000000000000000000000000000000000000000000000000000000", 1),
    ("placeholder-seed/zero-hash/dim_h=8",
     "ef532720a49159beb6816d98e13a162bac63c531b631bd1adb0fcca96b467ff3",
     "0000000000000000000000000000000000000000000000000000000000000000", 8),
    ("placeholder-seed/zero-hash/dim_h=32",
     "ef532720a49159beb6816d98e13a162bac63c531b631bd1adb0fcca96b467ff3",
     "0000000000000000000000000000000000000000000000000000000000000000", 32),
    ("placeholder-seed/all-ones-hash/dim_h=16",
     "ef532720a49159beb6816d98e13a162bac63c531b631bd1adb0fcca96b467ff3",
     "ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff", 16),
    ("seed=A/hash=B/dim_h=64",
     "14ab2dab0d3cddeaa58ec70632d3ff4f5de2c514004a92144e260dfe384d912e",
     "a23cb10b94660f062b467f313a9dc9d84f2dc2748c3627c661082dda7f55e3bb", 64),
    ("seed=A/hash=B'/dim_h=64 (one-bit-flip from prior)",
     "14ab2dab0d3cddeaa58ec70632d3ff4f5de2c514004a92144e260dfe384d912e",
     "223cb10b94660f062b467f313a9dc9d84f2dc2748c3627c661082dda7f55e3bb", 64),
    ("seed=A'/hash=B/dim_h=64 (one-bit-flip seed)",
     "94ab2dab0d3cddeaa58ec70632d3ff4f5de2c514004a92144e260dfe384d912e",
     "a23cb10b94660f062b467f313a9dc9d84f2dc2748c3627c661082dda7f55e3bb", 64),
    ("block-boundary/dim_h=16",
     "cfd60c2bda64ebcefbb23a5b28d98269c9c4f8b8ac77f6f9ca7a0f4865b10f58",
     "724cd966a7bfe78ba802877510ffb90c67f385a1d3135e4e1b8a1b38f744c6da", 16),
    ("block-boundary/dim_h=17",
     "cfd60c2bda64ebcefbb23a5b28d98269c9c4f8b8ac77f6f9ca7a0f4865b10f58",
     "724cd966a7bfe78ba802877510ffb90c67f385a1d3135e4e1b8a1b38f744c6da", 17),
    ("stress/dim_h=4096",
     "0ddd62c311f88ebe2d4f6cd5d9d1374474dfd645e012043648dd966a71785c95",
     "e605ede3d9d0d13c6d7d32c5c424b998677eef0689a0d9f0fa4ebd1bb4307cb9", 4096),
]


def build_lines() -> list[str]:
    lines = [
        f"# arborist v7 phi_prg known-answer tests — version {PHI_PRG_VERSION}",
        "# Regenerated by scripts/generate_phi_prg_kat.py. Records SHA-256 of",
        "# the raw HMAC-SHA-512 (little-endian counter-mode) expansion output,",
        "# before float conversion. Algorithm change MUST bump PHI_PRG_VERSION",
        "# and re-run this script; old runs replay against the old fixture.",
    ]
    for label, seed_hex, hh_hex, dim_h in _ENTRIES:
        seed = bytes.fromhex(seed_hex)
        hh = bytes.fromhex(hh_hex)
        raw = _expand(seed, hh, dim_h * 4)
        lines.append(json.dumps({
            "label": label,
            "version": PHI_PRG_VERSION,
            "seed_hex": seed_hex,
            "hard_hash_hex": hh_hex,
            "dim_h": dim_h,
            "output_sha256": hashlib.sha256(raw).hexdigest(),
            "output_bytes": dim_h * 4,
        }, ensure_ascii=False))
    return lines


def main() -> int:
    lines = build_lines()
    _FIXTURE.parent.mkdir(parents=True, exist_ok=True)
    _FIXTURE.write_text("\n".join(lines) + "\n", encoding="utf-8")
    print(f"wrote {len(_ENTRIES)} KAT entries to {_FIXTURE} (version {PHI_PRG_VERSION})")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())