#!/usr/bin/env python3
"""Generate the reproducible LinuxCNC source and coverage seed indexes.

The source index is deliberately mechanical: every file in a declared input
root is recorded with its size and digest.  The coverage seed is not an
implementation claim; entries remain ``planned`` until an owner assigns a
feature, adapter, test and oracle.
"""

from __future__ import annotations

import argparse
import hashlib
import json
from pathlib import Path
from typing import Iterable


ROOTS = (
    ("doc", Path("linuxcnc/docs")),
    ("source", Path("linuxcnc/src")),
    ("config", Path("linuxcnc/configs/sim")),
    ("test", Path("linuxcnc/tests")),
    ("unit_test", Path("linuxcnc/unit_tests")),
    ("implementation", Path("wasm-core")),
    ("implementation", Path("web")),
)
BUILD_FILES = (
    Path("linuxcnc/VERSION"),
    Path("linuxcnc/meson.build"),
)


def sha256(path: Path) -> str:
    digest = hashlib.sha256()
    with path.open("rb") as handle:
        for block in iter(lambda: handle.read(1024 * 1024), b""):
            digest.update(block)
    return digest.hexdigest()


def source_kind(root_kind: str, path: Path) -> str:
    if root_kind == "implementation":
        if path.suffix in {".h", ".hpp"}:
            return "web_wasm_interface"
        if path.suffix in {".c", ".cc", ".cpp", ".mjs", ".js"}:
            return "web_wasm_implementation"
        return "web_wasm_build_or_test"
    if root_kind == "doc":
        return "documentation"
    if root_kind in {"test", "unit_test"}:
        return "test"
    if root_kind == "config":
        return "simulation_config"
    if path.name.endswith(".comp"):
        return "hal_component_source"
    return "source"


def iter_files(repo_root: Path) -> Iterable[tuple[str, Path]]:
    for root_kind, relative_root in ROOTS:
        root = repo_root / relative_root
        if not root.is_dir():
            raise FileNotFoundError(f"required inventory root is missing: {relative_root}")
        for path in sorted(path for path in root.rglob("*") if path.is_file()):
            yield root_kind, path
    for relative_path in BUILD_FILES:
        path = repo_root / relative_path
        if not path.is_file():
            raise FileNotFoundError(f"required build source is missing: {relative_path}")
        yield "build", path


def make_entry(repo_root: Path, root_kind: str, path: Path) -> dict[str, object]:
    relative = path.relative_to(repo_root).as_posix()
    source_id = f"{root_kind}:{relative}"
    return {
        "source_id": source_id,
        "path": relative,
        "kind": source_kind(root_kind, path),
        "bytes": path.stat().st_size,
        "sha256": sha256(path),
    }


def make_manifest_entry(source: dict[str, object]) -> dict[str, object]:
    source_id = str(source["source_id"])
    return {
        "id": f"seed:{source_id}",
        "feature_id": None,
        "source": source_id,
        "source_sha": source["sha256"],
        "domain": "unassigned",
        "kind": source["kind"],
        "adaptation": None,
        "dependencies": [],
        "inputs": [source["path"]],
        "outputs": [],
        "errors": [],
        "sim_status": "planned",
        "web_status": "planned",
        "test_id": None,
        "oracle": None,
        "golden": None,
        "difference": None,
        "coverage_dimensions": [],
        "context_id": None,
        "exclusion_reason": None,
        "mapping_status": "needs_owner",
    }


def write_jsonl(path: Path, entries: Iterable[dict[str, object]]) -> None:
    path.parent.mkdir(parents=True, exist_ok=True)
    with path.open("w", encoding="utf-8", newline="\n") as handle:
        for entry in entries:
            handle.write(json.dumps(entry, ensure_ascii=False, sort_keys=True, separators=(",", ":")))
            handle.write("\n")


def generate(repo_root: Path, source_index: Path, manifest: Path) -> tuple[int, int]:
    sources = [make_entry(repo_root, kind, path) for kind, path in iter_files(repo_root)]
    sources.sort(key=lambda item: str(item["source_id"]))
    write_jsonl(source_index, sources)
    write_jsonl(manifest, (make_manifest_entry(source) for source in sources))
    return len(sources), len(sources)


def main() -> int:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("--repo-root", type=Path, default=Path(__file__).resolve().parents[1])
    parser.add_argument("--source-index", type=Path, default=Path("coverage/source-index.jsonl"))
    parser.add_argument("--manifest", type=Path, default=Path("coverage/manifest.jsonl"))
    args = parser.parse_args()
    source_count, manifest_count = generate(
        args.repo_root.resolve(),
        args.repo_root.resolve() / args.source_index,
        args.repo_root.resolve() / args.manifest,
    )
    print(f"source_index={source_count} manifest_seed={manifest_count}")
    return 0


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