portfolio-optimiser-claude/tests/test_ingest_adoption.py

"""Adoption seam: Door A ingest is the shared llm-ingestion-okf library (§11).

This repo's local ingest implementation was replaced by the shared library
(first consumer adoption). These load-bearing tests bind the adapter seam:

- Delegation — ``ingest.materialize`` IS a call into the library (RED if a
  local reimplementation sneaks back in).
- Offline invariant — the adapter NEVER passes the per-run network opt-in:
  an http-source manifest is refused at the library's network gate (RED if
  the adapter starts granting network access).
- Error contract — the consumer-facing error types ARE the library's typed
  hierarchy rooted in ``IngestError`` (RED if the seam re-wraps or forks).
"""

from __future__ import annotations

import json
from pathlib import Path
from typing import Any

import pytest

import llm_ingestion_okf
from portfolio_optimiser_claude import ingest

INGESTED_AT = "2026-07-16T12:00:00Z"


def _http_manifest() -> dict[str, Any]:
    return {
        "manifest_version": 1,
        "source": {"type": "http", "id": "api", "base_url": "https://example.invalid"},
        "bundle_summary": "s",
        "extractions": [
            {"id": "e", "title": "T", "query": "rows", "okf_type": "dataset", "max_rows": 1}
        ],
    }


class TestDelegation:
    """Seam: the consumer entry points are the library's (RED if detached)."""

    def test_load_manifest_is_the_library_entry_point(self) -> None:
        assert ingest.load_manifest is llm_ingestion_okf.load_manifest

    def test_materialize_delegates_with_the_offline_default(
        self, monkeypatch: pytest.MonkeyPatch, tmp_path: Path
    ) -> None:
        calls: dict[str, Any] = {}

        def fake(
            manifest_path: Path, bundle_dir: Path, ingested_at: str, **kwargs: Any
        ) -> llm_ingestion_okf.IngestResult:
            calls["args"] = (manifest_path, bundle_dir, ingested_at)
            calls["kwargs"] = kwargs
            return llm_ingestion_okf.IngestResult(
                written=(tmp_path / "ingest-e.md",), stamp="e@0123456789abcdef"
            )

        monkeypatch.setattr(ingest, "materialize_bundle", fake)
        out = ingest.materialize(tmp_path / "manifest.json", tmp_path / "bundle", INGESTED_AT)
        assert out == [tmp_path / "ingest-e.md"]  # IngestResult.written → list, order kept
        assert calls["args"] == (tmp_path / "manifest.json", tmp_path / "bundle", INGESTED_AT)
        # The offline invariant at the seam: allow_network/http_get are NEVER
        # passed — the library's local-only default stays in force.
        assert calls["kwargs"] == {}


class TestOfflineInvariant:
    """Seam: the adapter cannot grant network access (RED if it opts in)."""

    def test_http_source_is_refused_at_the_network_gate(self, tmp_path: Path) -> None:
        manifest = tmp_path / "manifest.json"
        manifest.write_text(json.dumps(_http_manifest()), encoding="utf-8")
        bundle = tmp_path / "bundle"
        with pytest.raises(ingest.NetworkGateError):
            ingest.materialize(manifest, bundle, INGESTED_AT)
        assert not bundle.exists() or not any(bundle.iterdir())  # gate fires before any write


class TestErrorContract:
    """Seam: consumer-facing errors ARE the library hierarchy (RED if forked)."""

    def test_error_types_are_the_library_types(self) -> None:
        for name in (
            "IngestError",
            "ManifestError",
            "MaterializationError",
            "NetworkGateError",
            "RenderError",
            "SourceError",
        ):
            assert getattr(ingest, name) is getattr(llm_ingestion_okf, name)

    def test_every_error_roots_in_ingest_error(self) -> None:
        for exc in (
            ingest.ManifestError,
            ingest.MaterializationError,
            ingest.NetworkGateError,
            ingest.RenderError,
            ingest.SourceError,
        ):
            assert issubclass(exc, ingest.IngestError)


class TestLibraryGuarantees:
    """K2.9 (R-3): library-side guarantees this consumer relies on, bound at the seam.

    Pre-adoption the local connector crashed mid-materialization on an empty
    CSV and left a PARTIAL bundle on disk (run-proven R-3). The library stages
    in memory, so the typed ``SourceError`` fires BEFORE the disk phase. Bound
    THROUGH the consumer entry point so a pin bump can never silently regress
    either guarantee.
    """

    def test_empty_csv_fails_typed_before_any_disk_write(self, tmp_path: Path) -> None:
        case = tmp_path / "case"
        fixture = case / "fixture"
        fixture.mkdir(parents=True)
        (fixture / "e.csv").write_text("", encoding="utf-8")
        manifest = {
            "manifest_version": 1,
            "source": {"type": "file", "id": "arkiv", "root": "fixture"},
            "bundle_summary": "s",
            "extractions": [
                {"id": "e", "title": "T", "query": "e.csv", "okf_type": "dataset", "max_rows": 5}
            ],
        }
        manifest_path = case / "manifest.json"
        manifest_path.write_text(json.dumps(manifest), encoding="utf-8")
        bundle = tmp_path / "bundle"
        with pytest.raises(ingest.SourceError):
            ingest.materialize(manifest_path, bundle, INGESTED_AT)
        assert not bundle.exists()  # never a partial bundle (in-memory staging)

    def test_non_select_sql_fails_typed_with_no_partial_bundle(
        self, tmp_path: Path, monkeypatch: pytest.MonkeyPatch
    ) -> None:
        import sqlite3

        db = tmp_path / "src.sqlite"
        con = sqlite3.connect(db)
        con.execute("CREATE TABLE t (a INTEGER)")
        con.commit()
        con.close()
        monkeypatch.setenv("SRC_DSN", str(db))
        manifest = {
            "manifest_version": 1,
            "source": {"type": "sql", "id": "db", "connection_ref": "SRC_DSN"},
            "bundle_summary": "s",
            "extractions": [
                {"id": "e", "title": "T", "query": "BEGIN", "okf_type": "dataset", "max_rows": 5}
            ],
        }
        manifest_path = tmp_path / "manifest.json"
        manifest_path.write_text(json.dumps(manifest), encoding="utf-8")
        bundle = tmp_path / "bundle"
        with pytest.raises(ingest.SourceError) as exc:
            ingest.materialize(manifest_path, bundle, INGESTED_AT)
        assert exc.value.code == "sql_no_columns"
        assert not bundle.exists()