portfolio-optimiser/tests/test_exploration_artefact_approaches_loadbearing.py

"""Fiks-ordre 20260904T070930Z, punkt 2 — ``{run_id}-exploration.json`` must carry the approaches
the loop actually SHAPED, not just the mandate announcement on stdout.

**Measured** on K2's syretest (``docs/2026-09-04-syretest-s7b-k2.md § 3.6``): the artefact had
``rounds``/``tool_calls``/``plan_reviews``/``quick_validations``, while which approaches the loop
minted stood only in ``mandate.announce``'s stdout block — printed by ``run.py``, never written.
An operator reading the artefact days later (the whole point of the async U12 door) could not
recover what the run had decided to evaluate without also having kept the terminal's stdout.

The fix widens ``explore.trace_payload`` with a required ``mandate`` keyword: it renders
``mandate.approaches`` under an ``"approaches"`` key, using the SAME fields ``mandate.announce``
reads off ``Approach`` (id, label). This file proves the artefact and the stdout announcement
agree — not just that a key exists.
"""

from __future__ import annotations

import json
from collections.abc import Callable
from pathlib import Path
from typing import Any

from agent_framework import BaseChatClient

from portfolio_optimiser import explore, run
from portfolio_optimiser.explore import ExplorationTrace
from portfolio_optimiser.simulation import ScriptedChatClient

_BUNDLE_DIR = Path(__file__).resolve().parents[1] / "shared" / "examples" / "bygg-energi-mikro"
_PID = "BYGG-KONTOR-NORD"

_LABEL = "SENTINEL-APPROACHES-9f2a"
_RATIONALE = "because the fittings are old (sentinel rationale)"

_ENERGY_REPLY = (
    '{"measure":"LED-retrofit av kontorbelysning","affected_items":'
    '[{"code":"ENERGI-TOTAL-EL","quantity":300000,"unit_cost":1.0}],"claimed_saving_nok":30000}'
)

_CONTRACT_JSON: dict[str, Any] = {
    "max_rounds": 4,
    "max_tokens": 100_000,
    "max_stall_count": 2,
    "max_reset_count": 1,
    "max_plan_revisions": 0,
    "enable_plan_review": False,
}


def _ledger_json(*, satisfied: bool, speaker: str = "hypothesiser") -> str:
    return json.dumps(
        {
            "is_request_satisfied": {"reason": "r", "answer": satisfied},
            "is_in_loop": {"reason": "r", "answer": False},
            "is_progress_being_made": {"reason": "r", "answer": True},
            "next_speaker": {"reason": "r", "answer": speaker},
            "instruction_or_question": {"reason": "r", "answer": "Shape one hypothesis."},
        }
    )


def _manager_script(ledgers: list[str]) -> Callable[[str, str], str]:
    def _select(blob: str, _role: str) -> str:
        if "provide the final answer" in blob:
            return "FINAL: exploration done."
        if "pure JSON format" in blob:
            return ledgers.pop(0) if ledgers else _ledger_json(satisfied=True)
        if "went wrong on this last run" in blob:
            return "PLAN-UPDATE: revised plan."
        if "rewrite the following fact sheet" in blob:
            return "FACTS-UPDATE: revised facts."
        if "bullet-point plan" in blob:
            return "PLAN: - ask the hypothesiser"
        if "pre-survey" in blob:
            return "FACTS: the bundle is anchored."
        return "{}"

    return _select


def _hypothesis_line(label: str, rationale: str) -> str:
    return f"{explore.HYPOTHESIS_MARKER} " + json.dumps({"label": label, "rationale": rationale})


def _factory(*, ledgers: list[str], hypothesiser: list[str]) -> Callable[[str], BaseChatClient]:
    def factory(role: str) -> BaseChatClient:
        if role == explore.MANAGER_ROLE:
            return ScriptedChatClient(reply_selector=_manager_script(ledgers), role=role)
        if role == explore.HYPOTHESISER_ROLE:
            replies = list(hypothesiser)

            def _hyp(_blob: str, _role: str) -> str:
                return replies.pop(0) if replies else "nothing further."

            return ScriptedChatClient(reply_selector=_hyp, role=role)
        if role == explore.NAVIGATOR_ROLE:
            return ScriptedChatClient("NAVIGATOR: index read.", role=role)
        return ScriptedChatClient(_ENERGY_REPLY, role=role)

    return factory


def _config_file(tmp_path: Path) -> str:
    path = tmp_path / "exploration.json"
    path.write_text(json.dumps(_CONTRACT_JSON), encoding="utf-8")
    return str(path)


def _base_argv(tmp_path: Path) -> list[str]:
    return [
        _PID,
        "--docs-dir",
        str(_BUNDLE_DIR),
        "--bundle-dir",
        str(_BUNDLE_DIR),
        "--explore",
        "Find the cheapest saving.",
        "--explore-config",
        _config_file(tmp_path),
    ]


def test_the_artefact_names_the_same_approach_the_announcement_did(
    tmp_path, capsys, monkeypatch
) -> None:
    """Goal test: the label ``mandate.announce`` printed on stdout is the SAME label the artefact
    carries, under the SAME minted id — read back from disk, not scraped off stdout.

    Detach point: drop the ``"approaches"`` key from ``trace_payload`` → ``KeyError`` here.
    Detach point: render an empty list regardless of ``mandate`` → the id/label assertions go RED.
    """
    factory = _factory(
        ledgers=[_ledger_json(satisfied=False), _ledger_json(satisfied=True)],
        hypothesiser=[_hypothesis_line(_LABEL, _RATIONALE)],
    )
    monkeypatch.setattr("portfolio_optimiser.run._default_factory", lambda profile: factory)

    outbox = tmp_path / "outbox"
    rc = run.main(_base_argv(tmp_path) + ["--outbox-dir", str(outbox), "--run-id", "r3"])
    assert rc == 0, capsys.readouterr().err

    out = capsys.readouterr().out
    assert f"1. hypothesis-1 — {_LABEL}" in out, "the stdout announcement is the fixture's oracle"

    payload = json.loads((outbox / "r3-exploration.json").read_text(encoding="utf-8"))
    assert payload["approaches"] == [
        {
            "id": "hypothesis-1",
            "label": _LABEL,
            "description": _RATIONALE,
            "affected_codes": [],
            "claimed_saving_nok": None,
            "bundle_id": "bygg-energi-mikro",
        }
    ], (
        "the artefact must name the same approach the terminal announced, or a reader without the terminal learns nothing"
    )


def test_a_mandate_with_no_commissioned_approaches_renders_an_empty_but_present_list() -> None:
    """Control: ``mandate.approaches == ()`` (pure own-proposals) is a legitimate state and must
    render as ``[]``, not be confused with "no mandate was ever formed" — that absence is what
    ``completed=False`` already says (``trace_payload``'s own docstring)."""
    trace = ExplorationTrace()
    from portfolio_optimiser.mandate import Mandate

    empty_mandate = Mandate(objective="find something", approaches=())
    payload = explore.trace_payload(trace, stop=None, completed=True, mandate=empty_mandate)
    assert payload["approaches"] == []

    payload_no_mandate = explore.trace_payload(trace, stop=None, completed=False, mandate=None)
    assert payload_no_mandate["approaches"] == []