#!/usr/bin/env python3
"""Generate deterministic configuration startup closure evidence.

The resolver intentionally has a narrow search policy.  Ordinary relative
references resolve from their entrypoint directory, repository-qualified
references resolve from the repository root, and only an explicit ``LIB:``
prefix searches the locked LinuxCNC library roots.  Bare basenames are never
silently replaced by a same-named candidate elsewhere in the checkout.
"""

from __future__ import annotations

import argparse
import hashlib
import json
import re
import shlex
from collections import Counter
from pathlib import Path, PurePosixPath
from typing import Any


CONTEXT = Path("context-lock.json")
RESOURCES = Path("coverage/config-resources.jsonl")
REGRESSION = Path("coverage/configuration-regression.report.json")
GRAPH = Path("coverage/kinematics-config-graph.report.json")
GATE = Path("coverage/release-candidate-gate.report.json")
OUTPUT = Path("coverage/configuration-startup-closure.report.json")
SCHEMA = "configuration-startup-closure.v1"
LIBRARY_ROOTS = ("linuxcnc/lib/hallib", "linuxcnc/lib")
HOST_PREFIXES = ("~", "$", "${")


def canonical(value: Any) -> str:
    return json.dumps(value, ensure_ascii=False, sort_keys=True, separators=(",", ":"))


def digest(value: Any) -> str:
    return hashlib.sha256(canonical(value).encode("utf-8")).hexdigest()


def file_digest(path: Path) -> str | None:
    if not path.is_file():
        return None
    handle = hashlib.sha256()
    with path.open("rb") as stream:
        for block in iter(lambda: stream.read(1024 * 1024), b""):
            handle.update(block)
    return handle.hexdigest()


def artifact(root: Path, path: Path, role: str) -> dict[str, Any]:
    full = root / path
    sha = file_digest(full)
    if sha is None:
        return {"path": path.as_posix(), "role": role, "status": "missing"}
    return {"path": path.as_posix(), "role": role, "status": "present", "bytes": full.stat().st_size, "sha256": sha}


def read_json(root: Path, path: Path) -> tuple[dict[str, Any], str | None]:
    try:
        value = json.loads((root / path).read_text(encoding="utf-8"))
        if not isinstance(value, dict):
            raise ValueError(f"{path} must contain an object")
        return value, None
    except (OSError, UnicodeError, json.JSONDecodeError, ValueError) as exc:
        return {}, str(exc)


def read_jsonl(root: Path, path: Path) -> tuple[list[dict[str, Any]], str | None]:
    try:
        rows: list[dict[str, Any]] = []
        for line_number, line in enumerate((root / path).read_text(encoding="utf-8").splitlines(), 1):
            if not line.strip():
                continue
            value = json.loads(line)
            if not isinstance(value, dict):
                raise ValueError(f"{path}:{line_number} must contain an object")
            rows.append(value)
        return rows, None
    except (OSError, UnicodeError, json.JSONDecodeError, ValueError) as exc:
        return [], str(exc)


def normalize(value: str) -> tuple[str | None, str | None]:
    if not isinstance(value, str) or not value:
        return None, "invalid_reference"
    if "\x00" in value:
        return None, "invalid_reference"
    parts: list[str] = []
    for part in value.replace("\\", "/").split("/"):
        if not part or part == ".":
            continue
        if part == "..":
            if not parts:
                return None, "path_escape"
            parts.pop()
            continue
        parts.append(part)
    return "/".join(parts), None


def first_token(raw: Any) -> tuple[str | None, list[str], str | None]:
    if not isinstance(raw, str):
        return None, [], "invalid_reference"
    value = raw.strip().strip("\"'")
    if not value:
        return None, [], "empty_reference"
    try:
        tokens = shlex.split(value, posix=True)
    except ValueError:
        tokens = re.findall(r"[^\s]+", value)
    if not tokens:
        return None, [], "empty_reference"
    return tokens[0], tokens[1:], None


def candidates(raw_value: Any, entrypoint: str) -> tuple[list[str], str, str | None, list[str]]:
    value, args, error = first_token(raw_value)
    if error:
        return [], "blocked", error, args
    assert value is not None
    if value.startswith(HOST_PREFIXES):
        return [], "blocked", "host_path", args
    entry, error = normalize(entrypoint)
    if error or entry is None:
        return [], "blocked", error or "invalid_reference", args
    if value.startswith("LIB:"):
        name = value[4:]
        if not name or name.startswith("/") or ".." in PurePosixPath(name).parts:
            return [], "library", "invalid_reference", args
        paths = []
        for root in LIBRARY_ROOTS:
            path, path_error = normalize(f"{root}/{name}")
            if path_error is None and path is not None:
                paths.append(path)
        return paths, "library", None, args
    if value.startswith("/"):
        return [], "blocked", "outside_repo", args
    if value.startswith("linuxcnc/"):
        path, path_error = normalize(value)
        return ([path] if path else []), "repo_relative", path_error, args
    base = PurePosixPath(entry).parent.as_posix()
    path, path_error = normalize(f"{base}/{value}")
    if path_error:
        return [], "relative", path_error, args
    if path is not None and not (path == "linuxcnc" or path.startswith("linuxcnc/")):
        return [], "relative", "outside_repo", args
    return ([path] if path else []), "relative", None, args


def available_paths(root: Path) -> set[str]:
    return {path.relative_to(root).as_posix() for path in root.rglob("*") if path.is_file() or path.is_dir()}


def resolve_row(row: dict[str, Any], available: set[str]) -> dict[str, Any]:
    original_path = str(row.get("path", ""))
    original_status = str(row.get("parse_status", ""))
    raw_value = row.get("raw_value", original_path)
    if original_status in {"resolved_file", "resolved_directory"} and original_path in available:
        return {
            "resource_source_id": row.get("resource_source_id"),
            "entrypoint": row.get("entrypoint"),
            "original_path": original_path,
            "original_status": original_status,
            "raw_value": raw_value,
            "status": "resolved_existing",
            "code": None,
            "path": original_path,
            "args": [],
            "attempted_paths": [original_path],
            "resolution": "indexed",
        }
    paths, resolution, error, args = candidates(raw_value, str(row.get("entrypoint", "")))
    existing = [path for path in paths if path in available]
    if error:
        status = f"blocked_{error}"
        return {"resource_source_id": row.get("resource_source_id"), "entrypoint": row.get("entrypoint"), "original_path": original_path, "original_status": original_status, "raw_value": raw_value, "status": status, "code": error, "path": None, "args": args, "attempted_paths": paths, "resolution": resolution}
    if len(existing) > 1:
        return {"resource_source_id": row.get("resource_source_id"), "entrypoint": row.get("entrypoint"), "original_path": original_path, "original_status": original_status, "raw_value": raw_value, "status": "blocked_ambiguous", "code": "ambiguous", "path": None, "args": args, "attempted_paths": paths, "resolution": resolution}
    if len(existing) == 1:
        return {"resource_source_id": row.get("resource_source_id"), "entrypoint": row.get("entrypoint"), "original_path": original_path, "original_status": original_status, "raw_value": raw_value, "status": f"resolved_{resolution}", "code": None, "path": existing[0], "args": args, "attempted_paths": paths, "resolution": resolution}
    return {"resource_source_id": row.get("resource_source_id"), "entrypoint": row.get("entrypoint"), "original_path": original_path, "original_status": original_status, "raw_value": raw_value, "status": "blocked_missing", "code": "missing", "path": None, "args": args, "attempted_paths": paths, "resolution": resolution}


def generate(root: Path, output: Path = OUTPUT) -> dict[str, Any]:
    root = root.resolve()
    lock, lock_error = read_json(root, CONTEXT)
    regression, regression_error = read_json(root, REGRESSION)
    graph, graph_error = read_json(root, GRAPH)
    gate, gate_error = read_json(root, GATE)
    rows, rows_error = read_jsonl(root, RESOURCES)
    errors = {path.as_posix(): error for path, error in ((CONTEXT, lock_error), (REGRESSION, regression_error), (GRAPH, graph_error), (GATE, gate_error), (RESOURCES, rows_error)) if error}
    available = available_paths(root)
    records = [resolve_row(row, available) for row in rows]
    records.sort(key=lambda row: str(row.get("resource_source_id", "")))
    occurrence_counter: Counter[str] = Counter()
    for record in records:
        source_id = str(record.get("resource_source_id", ""))
        occurrence_counter[source_id] += 1
        record["resource_record_id"] = f"{source_id}#occurrence:{occurrence_counter[source_id]}"
    statuses = Counter(str(row["status"]) for row in records)
    codes = Counter(str(row["code"]) for row in records if row.get("code"))
    duplicate_source_ids = sorted(source_id for source_id, count in occurrence_counter.items() if count > 1)
    entrypoints = sorted({str(row.get("entrypoint")) for row in records})
    ids = [str(row.get("resource_source_id")) for row in records]
    config_entries = regression.get("configuration_entries", []) if isinstance(regression.get("configuration_entries"), list) else []
    context_id = lock.get("context_id")
    stop_ok = gate.get("release_action") == "STOP_RELEASE" and gate.get("candidate_eligible") is False
    checks = {
        "input_integrity": {"status": "verified" if not errors and context_id else "blocked", "context_id": context_id},
        "configuration_coverage": {"status": "verified" if len(config_entries) == 264 and len(entrypoints) == 233 else "blocked", "registered_entries": len(config_entries), "entrypoints": len(entrypoints)},
        "resource_identity": {"status": "verified" if len({str(row.get('resource_record_id')) for row in records}) == len(records) and ids == sorted(ids) else "blocked", "rows": len(ids), "duplicate_source_ids": duplicate_source_ids, "duplicate_policy": "source_id_preserved_with_deterministic_occurrence_suffix"},
        "explicit_library_policy": {"status": "verified" if all(row.get("resolution") in {"indexed", "relative", "repo_relative", "library", "blocked"} for row in records) else "blocked", "library_references": sum(1 for row in records if str(row.get("raw_value", "")).lstrip().startswith("LIB:")), "library_resolved": sum(1 for row in records if row.get("status") == "resolved_library")},
        "no_basename_substitution": {"status": "verified" if all(row.get("resolution") != "basename" for row in records) else "blocked"},
        "release_stop_preserved": {"status": "verified" if stop_ok else "blocked", "release_action": "STOP_RELEASE" if stop_ok else None},
    }
    blocked_count = sum(value for key, value in statuses.items() if key.startswith("blocked_"))
    report: dict[str, Any] = {
        "schema": SCHEMA,
        "context_id": context_id,
        "task_id": "M3-CFG-01",
        "status": "verified" if not errors and blocked_count == 0 else "partial" if not errors else "blocked",
        "implementation_status": "verified" if not errors else "blocked",
        "candidate_eligible": False,
        "release_action": "STOP_RELEASE",
        "inputs": [artifact(root, CONTEXT, "context_lock"), artifact(root, RESOURCES, "config_resources"), artifact(root, REGRESSION, "configuration_regression"), artifact(root, GRAPH, "kinematics_config_graph"), artifact(root, GATE, "release_candidate_gate")],
        "counts": {"resource_rows": len(records), "entrypoints": len(entrypoints), "registered_configuration_entries": len(config_entries), "resolved": len(records) - blocked_count, "blocked": blocked_count, "duplicate_source_ids": len(duplicate_source_ids), "statuses": dict(sorted(statuses.items())), "blocked_codes": dict(sorted(codes.items()))},
        "checks": checks,
        "resources": records,
        "next_transition": "execute M3-RT-02 HAL component manifest and executable configuration graph",
        "errors": errors,
        "report_sha256": None,
    }
    report["report_sha256"] = digest({key: value for key, value in report.items() if key != "report_sha256"})
    destination = output if output.is_absolute() else root / output
    destination.parent.mkdir(parents=True, exist_ok=True)
    destination.write_text(json.dumps(report, ensure_ascii=False, sort_keys=True, indent=2) + "\n", encoding="utf-8")
    return report


def main() -> int:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("--repo-root", type=Path, default=Path(__file__).resolve().parents[1])
    parser.add_argument("--output", type=Path, default=OUTPUT)
    args = parser.parse_args()
    report = generate(args.repo_root.resolve(), args.output)
    print(f"configuration_startup_closure={report['status']} resolved={report['counts']['resolved']} blocked={report['counts']['blocked']}")
    return 0


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