#!/usr/bin/env python3
"""gen_inference_paths_truth.py — freeze/replay parity truth for
inference_paths (#145 census + trace wiring, dichotomy wave #2479).

Freeze mode (default): exercises EVERY public function of the LIVE Python
oracle (src/rune/inference_paths.py) across every branch and records
(kernel_input, expected_kernel_output) pairs into
src/go/inference-paths/truth.json (the GO port's battery replays this
VERBATIM).

Replay mode (--replay): re-runs the SAME scenarios against the live
Python oracle and asserts facade-level parity against the frozen file.
Kernel-level (Go) parity runs via `go test ./...` + the `replay` CLI in
src/go/inference-paths — the battery embeds the frozen file verbatim.

No-clock law: the wall clock is frozen by patching time.gmtime, so every
kernel payload carries its own `at` and replay is deterministic
regardless of run date.

Boundary law (ledger verdict: GO — the journal write is I/O-bound file
IO and the census is simple counting): the JOURNAL attempt
(rune.inference_trace import + record_trace + the journaled flag) and
the grep-verified census-completeness obligations stay in the Python
facade — the Go kernel owns the record shape (schema_version, kind,
module, function, model, task_id, duration_ms, at) and census_report.
Journal attempts in the battery write to throwaway temp dirs (never the
real journal); the failure branch passes a non-creatable dir so the
oracle's except path (journaled=False) is exercised deterministically.

Deep-copy law: every input is deep-copied before it reaches the oracle.
"""
import copy
import datetime
import json
import os
import sys
import tempfile
import time

REPO = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
TRUTH_PATH = os.path.join(REPO, "src", "go", "inference-paths", "truth.json")
REPLAY = "--replay" in sys.argv

os.environ["RUNE_IT_CORE"] = "python"  # poisoned-truth law: the journal
# path's own delegated serializer must not vary under the freeze

sys.path.insert(0, os.path.join(REPO, "src"))
sys.path.insert(0, os.path.join(REPO, "src", "rune"))

import inference_paths as oracle_mod  # noqa: E402

FROZEN_AT = "2026-10-12T12:34:56Z"
FROZEN_STRUCT = time.strptime(FROZEN_AT, "%Y-%m-%dT%H:%M:%SZ")

# no-clock law: freeze the wall clock BEFORE any scenario runs
oracle_mod.time.gmtime = lambda secs=None: FROZEN_STRUCT


class Case:
    def __init__(self, name, fn):
        self.name = name
        self.fn = fn


# ── trace_side_path ──────────────────────────────────────────────────────────

def _tsp_case(rec, name, trace_dir, **kwargs):
    args = {"module": "", "function": "", "model": "", "task_id": "",
            "duration_ms": 0.0}
    args.update(kwargs)
    args_c = copy.deepcopy(args)
    got = oracle_mod.trace_side_path(trace_dir=trace_dir, **copy.deepcopy(args))
    rec(name, {**args_c, "at": FROZEN_AT}, got,
        {"record": {k: v for k, v in got.items() if k != "journaled"}})


def c_tsp_defaults(rec):
    _tsp_case(rec, "tsp-defaults-side-path", tempfile.mkdtemp(prefix="ip_t_"),
              module="goal", function="_goal")


def c_tsp_full(rec):
    _tsp_case(rec, "tsp-full-fields", tempfile.mkdtemp(prefix="ip_t_"),
              module="rune_operator", function="_runtime_for_model",
              model="llama3", task_id="t1", duration_ms=12.5)


def c_tsp_zero_duration(rec):
    _tsp_case(rec, "tsp-zero-duration", tempfile.mkdtemp(prefix="ip_t_"),
              module="witness", function="_witness")


def c_tsp_float_precision(rec):
    _tsp_case(rec, "tsp-float-precision", tempfile.mkdtemp(prefix="ip_t_"),
              module="tiered_executor", function="_execute",
              model="qwen2.5", task_id="task-99", duration_ms=123.456)


def c_tsp_journal_failure(rec):
    # non-creatable dir → record_trace raises → journaled=False branch
    _tsp_case(rec, "tsp-journal-failure-false",
              "/nonexistent-ip-truth/nope", module="curator", function="_embed")


def c_tsp_empty_module(rec):
    _tsp_case(rec, "tsp-empty-module-function", tempfile.mkdtemp(prefix="ip_t_"))


# ── census_report ────────────────────────────────────────────────────────────

def c_cr_full(rec):
    got = oracle_mod.census_report()
    rec("cr-full-counts-and-exclusions", {}, got, got)


# ── driver ───────────────────────────────────────────────────────────────────

TRACE = [c_tsp_defaults, c_tsp_full, c_tsp_zero_duration,
         c_tsp_float_precision, c_tsp_journal_failure, c_tsp_empty_module]
REPORT = [c_cr_full]

GROUPS = [("trace_side_path", TRACE), ("census_report", REPORT)]


def canon(x) -> str:
    return json.dumps(json.loads(json.dumps(x, default=str)), sort_keys=True)


def main():
    frozen = None
    if REPLAY:
        with open(TRUTH_PATH) as f:
            frozen = json.load(f)["cases"]

    groups = {}
    for group, cases in GROUPS:
        recs = []

        def rec(name, inp, oracle_ret, expected, _r=recs):
            entry = {"name": name, "input": inp, "return": oracle_ret,
                     "expected": expected}
            if REPLAY:
                old = next(c for c in frozen[group] if c["name"] == name)
                for key_ in ("input", "return", "expected"):
                    got, want = entry[key_], old[key_]
                    if canon(got) != canon(want):
                        print(f"FACADE DIVERGE [{group}] {name} {key_}")
                        print("  got: ",
                              json.dumps(got, sort_keys=True, default=str)[:600])
                        print("  want:",
                              json.dumps(want, sort_keys=True, default=str)[:600])
                        recs.append({"name": name, "diverged": key_})
                        return
            _r.append(entry)

        for c in cases:
            c(rec)
        groups[group] = recs

    total = sum(len(v) for v in groups.values())
    if REPLAY:
        failures = sum(1 for v in groups.values()
                       for c in v if "diverged" in c)
        print(f"replay: facade {total - failures}/{total} "
              f"(kernel/Go parity: `go test ./...` + `replay` CLI, "
              f"battery embeds truth.json verbatim)")
        return 1 if failures else 0

    doc = {
        "frozen": datetime.datetime.now(datetime.timezone.utc).isoformat(),
        "oracle": "src/rune/inference_paths.py",
        "pin": "time.gmtime frozen at " + FROZEN_AT +
               " + RUNE_IT_CORE=python",
        "note": "Frozen via scripts/gen_inference_paths_truth.py against "
                "the live Python oracle. Ledger verdict: GO (the journal "
                "write is I/O-bound file IO; the census is simple "
                "counting). The journal attempt (rune.inference_trace + "
                "the journaled flag) stays in the Python facade — the Go "
                "kernel owns the record shape (schema_version, kind, "
                "module, function, model, task_id, duration_ms, at; `at` "
                "is a kernel argument — no-clock law) and census_report. "
                "Battery embeds this VERBATIM; float durations compare "
                "semantically (Python renders 0.0, Go marshals 0 — same "
                "value).",
        "cases": groups,
    }
    os.makedirs(os.path.dirname(TRUTH_PATH), exist_ok=True)
    with open(TRUTH_PATH, "w") as f:
        json.dump(doc, f, indent=1)
        f.write("\n")
    print(f"frozen {total} cases → {os.path.relpath(TRUTH_PATH, REPO)}")
    return 0


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