from __future__ import annotations

import importlib.util
import json
import sys
from pathlib import Path

ROOT = Path(__file__).resolve().parents[1]
for candidate in (ROOT / 'src', ROOT):
    if str(candidate) not in sys.path:
        sys.path.insert(0, str(candidate))

from app_kit.tracing import write_trace_artifact


def _load_share_module():
    script_path = ROOT / 'scripts' / 'share_traces_to_hf_dataset.py'
    spec = importlib.util.spec_from_file_location('share_traces_to_hf_dataset', script_path)
    if spec is None or spec.loader is None:
        raise RuntimeError(f'Cannot import {script_path}')
    module = importlib.util.module_from_spec(spec)
    spec.loader.exec_module(module)
    return module


def test_share_traces_materializes_schema_and_local_artifacts(tmp_path):
    repo_root = tmp_path / 'repo'
    artifact_dir = repo_root / 'artifacts'
    trace_path = write_trace_artifact(
        artifact_dir,
        {
            'kind': 'eval',
            'project': 'p1',
            'pack_id': 'sample-pack',
            'pack_path': 'data/demo_packs/sample-pack',
            'result': {'parsed': 'output'},
            'model_name': 'test-model',
        },
    )
    assert trace_path.exists()

    module = _load_share_module()
    output_dir = module.default_output_dir(
        repo_root=repo_root,
        repo_name=ROOT.name,
        today='2026-06-10',
    )
    materialized = module.materialize_dataset(
        traces_dir=artifact_dir / 'traces',
        output_dir=output_dir,
        repo_root=repo_root,
        repo_name=ROOT.name,
        today='2026-06-10',
    )

    assert materialized['records_written'] == 1
    assert materialized['output_dir'] == output_dir
    assert materialized['dataset_jsonl'].exists()
    assert materialized['metadata_json'].exists()

    row = json.loads(materialized['dataset_jsonl'].read_text(encoding='utf-8').strip())
    assert {'timestamp', 'inputs', 'parsed_outputs', 'model_name'}.issubset(row)
    assert row['model_name'] == 'test-model'
    assert row['parsed_outputs'] == {'parsed': 'output'}

    metadata = json.loads(materialized['metadata_json'].read_text(encoding='utf-8'))
    assert metadata['repo_name'] == ROOT.name
    assert metadata['schema_fields'] == ['timestamp', 'inputs', 'parsed_outputs', 'model_name']
    assert metadata['source_trace_count'] == 1
    assert str(materialized['output_dir']).endswith(f"artifacts/verification/2026-06-10/sharing_is_caring/{ROOT.name}")