from __future__ import annotations

import json
from pathlib import Path

from k1link.compute.e44_data_amplification_audit import (
    build_e44_data_amplification_audit,
)
from k1link.compute.e50_content_reference_index import (
    build_e50_content_reference_index,
)
from k1link.web.artifact_health_api import ArtifactHealthService


def _write(path: Path, payload: bytes) -> None:
    path.parent.mkdir(parents=True, exist_ok=True)
    path.write_bytes(payload)


def test_artifact_health_reports_live_inventory_and_exact_audit(tmp_path: Path) -> None:
    runtime = tmp_path / ".runtime"
    first = runtime / "compute-experiments" / "e30" / "result-a"
    second = runtime / "compute-experiments" / "e40" / "result-b"
    _write(first / "shared.bin", b"shared")
    _write(first / "only-a.bin", b"a")
    _write(second / "shared.bin", b"shared")
    _write(second / "only-b.bin", b"bb")
    e44_results = runtime / "compute-experiments" / "e44" / "results"
    e44 = build_e44_data_amplification_audit(
        artifact_roots={"e30": first, "e40": second},
        output_root=e44_results,
    )
    e50_results = runtime / "compute-experiments" / "e50" / "results"
    e50 = build_e50_content_reference_index(
        e44_result_root=e44.result_root,
        artifact_roots={"e30": first, "e40": second},
        output_root=e50_results,
    )

    service = ArtifactHealthService(
        lambda: runtime,
        lambda: e44_results,
        e50_results_root_provider=lambda: e50_results,
        cache_seconds=0,
    )
    document = service.snapshot()

    assert document.inventory.state == "live"
    assert document.inventory.file_count >= 6
    assert document.inventory.logical_bytes > 0
    assert document.exact_audit.state == "exact-current"
    assert document.exact_audit.duplicate_bytes == len(b"shared")
    assert document.content_references.state == "indexed"
    assert document.content_references.result_id == e50.result_id
    assert document.content_references.logical_reference_count == 4
    assert document.content_references.canonical_content_count == 3
    assert document.content_references.duplicate_reference_count == 1
    assert document.content_references.exact_duplicate_bytes == len(b"shared")
    assert document.content_references.physical_reclamation_applied is False
    assert document.content_references.storage_migration_authorized is False


def test_artifact_health_marks_exact_audit_stale_after_source_change(
    tmp_path: Path,
) -> None:
    runtime = tmp_path / ".runtime"
    first = runtime / "e30" / "result-a"
    second = runtime / "e40" / "result-b"
    _write(first / "a.bin", b"a")
    _write(second / "b.bin", b"b")
    e44_results = runtime / "e44" / "results"
    audit = build_e44_data_amplification_audit(
        artifact_roots={"e30": first, "e40": second},
        output_root=e44_results,
    )
    manifest = json.loads((audit.result_root / "manifest.json").read_text())
    first_name = manifest["identity"]["artifact_roots"][0]["root_name"]
    _write(first / "new.bin", b"new")

    document = ArtifactHealthService(
        lambda: runtime,
        lambda: e44_results,
        cache_seconds=0,
    ).snapshot()

    assert document.exact_audit.state == "exact-stale"
    assert first_name in document.exact_audit.changed_roots


def test_artifact_health_reports_unavailable_runtime(tmp_path: Path) -> None:
    document = ArtifactHealthService(
        lambda: tmp_path / "missing",
        lambda: tmp_path / "missing-e44",
        cache_seconds=0,
    ).snapshot()

    assert document.overall_state == "unavailable"
    assert document.inventory.state == "unavailable"
    assert document.exact_audit.state == "unavailable"
    assert document.content_references.state == "not-indexed"