"""The stage cache, from the side that needs a database.

The pipeline half — what a hit restores and what a key is made of — is in
`tests/flow/test_runs.py`, which runs without one.
"""

import json
from datetime import UTC, datetime, timedelta

import pytest
from sqlmodel import Session, col, select

from fluksio.core.config import settings
from fluksio.core.db import engine as db_engine
from fluksio.flow.artifacts import ArtifactStore
from fluksio.flow.messages import DType, MessageSpec
from fluksio.flow.pipeline import NodeOutcome
from fluksio.flow.runs import (
    OUTPUT_CAP,
    RunCache,
    RunRejected,
    _cacheable,
    new_run_id,
    resolve_references,
    seed_values,
)
from fluksio.flow.schemas import FlowDef, FlowInput, NodeDef
from fluksio.models import Run, RunArtifact, RunMetric, RunNode


def test_a_run_cache_finds_what_an_earlier_run_recorded(tmp_path):
    """The run history is the cache; there is no second store to keep."""
    store = ArtifactStore(tmp_path / "artifacts")
    reference = store.put([b"payload"], name="data.bin")
    plain, with_artifact, collected = "k-plain", "k-artifact", "k-collected"
    with Session(db_engine) as session:
        session.add(
            RunNode(
                run_id="cache-1",
                node="study.a",
                status="ok",
                cache_key=plain,
                outputs=json.dumps({"study.loss": 1.5}),
            )
        )
        session.add(
            RunNode(
                run_id="cache-2",
                node="study.b",
                status="ok",
                cache_key=with_artifact,
                outputs=json.dumps({"study.data": reference}),
            )
        )
        session.add(
            RunNode(
                run_id="cache-3",
                node="study.c",
                status="ok",
                cache_key=collected,
                outputs=json.dumps(
                    {"study.data": {**reference, "digest": "sha256:" + "1" * 64}}
                ),
            )
        )
        session.commit()

    cache = RunCache(store)
    hit = cache.lookup(plain)
    assert hit is not None and hit.outputs == {"study.loss": 1.5}
    # Where its series is, so a run reusing it can read the curve back.
    assert hit.metrics_run == "cache-1"
    found = cache.lookup(with_artifact)
    assert found is not None and found.outputs == {"study.data": reference}
    # Its bytes have gone from the store, so the reference names nothing a
    # restored run could open. That is a miss, not a broken run.
    assert cache.lookup(collected) is None
    assert cache.lookup("never-seen") is None
    assert cache.lookup("") is None


def test_what_may_be_stored_as_a_cache_entry():
    """A row carries a key and its outputs together, or neither."""
    ok = NodeOutcome(
        node="study.a", ok=True, cache_key="k", output_values={"study.loss": 1.0}
    )
    assert _cacheable(ok) == '{"study.loss":1.0}'
    # A node that published nothing is still an answer worth reusing.
    assert _cacheable(ok.model_copy(update={"output_values": None})) == "null"
    # Not cacheable: it failed, it has no key, or it returned too much.
    assert _cacheable(ok.model_copy(update={"ok": False})) is None
    assert _cacheable(ok.model_copy(update={"cache_key": ""})) is None
    big = {"study.data": "x" * (OUTPUT_CAP + 1)}
    assert _cacheable(ok.model_copy(update={"output_values": big})) is None


# -----------------------------------------------------------------------------
# Naming an artifact from outside the process that made it
#
# A python caller passes the reference it holds. A shell holds nothing, so the
# same input also takes `@run:<id>.<output>` or a bare digest, resolved here
# rather than in each client.
# -----------------------------------------------------------------------------


def artifact_flow() -> FlowDef:
    """A flow taking a dataset somebody else's run produced."""
    dataset = MessageSpec(name="dataset", dtype=DType.ARTIFACT)
    return FlowDef(
        name="study",
        mode="batch",
        inputs=[FlowInput(spec=dataset)],
        nodes=[NodeDef(id="train", requires=[dataset])],
    )


@pytest.fixture
def made_artifact():
    """A finished run with one artifact, as a later run would find it."""
    digest = "sha256:" + "a1" * 32
    reference = {
        "digest": digest,
        "size": 12,
        "media_type": "text/csv",
        "name": "cities.csv",
    }
    run_id = new_run_id()
    with Session(db_engine) as session:
        session.add(
            Run(
                id=run_id,
                flow="prepare",
                status="ok",
                result={"dataset": reference},
                created_at=datetime.now(UTC),
            )
        )
        session.add(
            RunArtifact(
                run_id=run_id,
                name="prepare.dataset",
                node="load",
                digest=digest,
                size=12,
            )
        )
        session.commit()
    yield run_id, reference
    with Session(db_engine) as session:
        session.delete(session.get(RunArtifact, (run_id, "prepare.dataset")))
        session.delete(session.get(Run, run_id))
        session.commit()


def test_a_run_reference_resolves_to_what_that_run_produced(made_artifact):
    run_id, reference = made_artifact
    resolved = resolve_references(
        artifact_flow(), {"dataset": f"@run:{run_id}.dataset"}
    )

    # The producer's own reference, file name and all — not one rebuilt from
    # the row, which carries the message name instead.
    assert resolved["dataset"] == reference


def test_a_bare_digest_resolves_to_the_bytes_under_it(made_artifact):
    _run_id, reference = made_artifact
    resolved = resolve_references(artifact_flow(), {"dataset": reference["digest"]})

    assert resolved["dataset"]["digest"] == reference["digest"]
    assert resolved["dataset"]["size"] == 12


def test_a_resolved_reference_passes_the_input_check(made_artifact):
    run_id, _reference = made_artifact
    flow = artifact_flow()
    resolved = resolve_references(flow, {"dataset": f"@run:{run_id}.dataset"})

    assert "study.dataset" in seed_values(flow, resolved)


def test_an_output_a_run_never_made_says_what_it_did(made_artifact):
    run_id, _reference = made_artifact
    with pytest.raises(RunRejected, match="prepare.dataset"):
        resolve_references(artifact_flow(), {"dataset": f"@run:{run_id}.weights"})


def test_a_reference_to_no_run_at_all_is_refused():
    with pytest.raises(RunRejected, match="no run"):
        resolve_references(artifact_flow(), {"dataset": "@run:nothing.dataset"})


def test_an_unknown_digest_is_refused():
    with pytest.raises(RunRejected, match="nothing here"):
        resolve_references(artifact_flow(), {"dataset": "sha256:" + "b2" * 32})


def test_a_reference_passed_whole_is_left_alone(made_artifact):
    """A python caller already has the object, and hands it over as one."""
    _run_id, reference = made_artifact
    assert resolve_references(artifact_flow(), {"dataset": reference}) == {
        "dataset": reference
    }


def chaining_flow() -> FlowDef:
    """A flow taking a json config and a label another run worked out."""
    meta = MessageSpec(name="meta", dtype=DType.JSON)
    label = MessageSpec(name="label", dtype=DType.STR)
    return FlowDef(
        name="study",
        mode="batch",
        inputs=[FlowInput(spec=meta), FlowInput(spec=label)],
        nodes=[NodeDef(id="train", requires=[meta, label])],
    )


@pytest.fixture
def made_config():
    """A finished run whose result is an object, not bytes."""
    meta = {"rows": 256, "source": "builtin"}
    run_id = new_run_id()
    with Session(db_engine) as session:
        session.add(
            Run(
                id=run_id,
                flow="generate",
                status="ok",
                result={"meta": meta, "label": "run-7"},
                created_at=datetime.now(UTC),
            )
        )
        session.commit()
    yield run_id, meta
    with Session(db_engine) as session:
        session.delete(session.get(Run, run_id))
        session.commit()


def test_a_json_input_may_name_a_run_s_output(made_config):
    """The gap this closes: chaining without pasting the object into a shell."""
    run_id, meta = made_config
    flow = chaining_flow()
    resolved = resolve_references(flow, {"meta": f"@run:{run_id}.meta"})

    assert resolved["meta"] == meta
    # And it is the value's own type from here on, so the input check passes.
    assert seed_values(flow, resolved)["study.meta"] == meta


def test_the_spelling_is_reserved_on_a_text_input_too(made_config):
    run_id, _meta = made_config
    resolved = resolve_references(chaining_flow(), {"label": f"@run:{run_id}.label"})

    assert resolved["label"] == "run-7"


def test_text_that_names_nothing_is_still_left_alone(made_config):
    """Only the two spellings are read as names; everything else is a value."""
    _run_id, _meta = made_config
    params = {"label": "@run-of-the-mill", "meta": {"rows": 1}}
    assert resolve_references(chaining_flow(), params) == params


def test_the_overview_counts_a_flow_the_list_page_would_not_reach(
    client, superuser_token_headers
):
    """The list caps at 500 newest; the flow rail needs whole counts."""
    with Session(db_engine) as session:
        for index in range(3):
            session.add(
                Run(
                    id=f"ov-{index}",
                    flow="overviewed",
                    status="ok" if index else "running",
                    created_at=datetime(2026, 1, 1 + index, tzinfo=UTC),
                )
            )
        session.commit()

    rows = client.get(
        f"{settings.API_V1_STR}/runs/overview", headers=superuser_token_headers
    ).json()
    row = next(r for r in rows if r["flow"] == "overviewed")

    assert (row["runs"], row["running"], row["queued"]) == (3, 1, 0)


def test_a_running_run_reports_how_long_it_has_been_going(
    client, superuser_token_headers
):
    """A duration is only written at the end; until then, time since it began."""
    with Session(db_engine) as session:
        session.add(
            Run(
                id="in-flight",
                flow="timed",
                status="running",
                created_at=datetime.now(UTC),
                started_at=datetime.now(UTC) - timedelta(seconds=30),
            )
        )
        session.commit()

    rows = client.get(
        f"{settings.API_V1_STR}/runs",
        headers=superuser_token_headers,
        params={"flow": "timed"},
    ).json()

    assert rows[0]["duration_ms"] >= 30_000


def test_a_run_records_which_caller_asked_for_it(
    client, superuser_token_headers, monkeypatch
):
    """The dashboard, the CLI and the SDK are told apart by what they send.

    Client-supplied, so the vocabulary is closed: a column nobody can write
    free text into is one a table can group by.
    """
    seen: dict[str, object] = {}

    class Recorder:
        def submit(self, name, **kwargs):
            seen.update(kwargs)
            return Run(
                id="cause-1",
                flow=name,
                cause=str(kwargs["cause"]),
                created_at=datetime.now(UTC),
            )

    monkeypatch.setattr(client.app.state, "run_service", Recorder())
    url = f"{settings.API_V1_STR}/runs/flows/demo"

    answer = client.post(url, headers=superuser_token_headers, json={"cause": "cli"})
    assert answer.status_code == 202
    assert (seen["cause"], answer.json()["cause"]) == ("cli", "cli")

    # Nothing said still means the dashboard, which is the only caller that
    # does not name itself.
    client.post(url, headers=superuser_token_headers, json={})
    assert seen["cause"] == "api"

    refused = client.post(
        url, headers=superuser_token_headers, json={"cause": "somewhere else"}
    )
    assert refused.status_code == 422


def test_overview_is_not_read_as_a_run_id(client, superuser_token_headers):
    """`/overview` is declared before `/{run_id}`, which would swallow it."""
    answer = client.get(
        f"{settings.API_V1_STR}/runs/overview", headers=superuser_token_headers
    )

    assert answer.status_code == 200
    assert isinstance(answer.json(), list)


# -----------------------------------------------------------------------------
# A cached node's curve
#
# A hit replays no emissions, so the series stays in the run that recorded it
# and the run reusing it points there. Reading either one answers the same.
# -----------------------------------------------------------------------------


@pytest.fixture
def reused_run():
    """A run of `quick` whose node was restored from a run of `train`."""
    with Session(db_engine) as session:
        made = datetime.now(UTC)
        session.add(Run(id="src-1", flow="train", status="ok", created_at=made))
        session.add(Run(id="reuse-1", flow="quick", status="ok", created_at=made))
        session.add(RunNode(run_id="src-1", node="train.fit", status="ok"))
        session.add(
            RunNode(
                run_id="reuse-1",
                node="quick.fit",
                status="cached",
                cached_from="src-1",
            )
        )
        for step, value in enumerate([3.0, 2.0, 1.0]):
            session.add(
                RunMetric(
                    run_id="src-1",
                    name="train.loss",
                    step=step,
                    node="train.fit",
                    value=value,
                )
            )
        session.commit()
    yield
    with Session(db_engine) as session:
        for row in session.exec(select(RunMetric)).all():
            session.delete(row)
        for row in session.exec(select(RunNode)).all():
            session.delete(row)
        for run_id in ("src-1", "reuse-1"):
            run = session.get(Run, run_id)
            if run is not None:
                session.delete(run)
        session.commit()


@pytest.mark.usefixtures("reused_run")
def test_a_cached_node_answers_with_the_curve_it_was_restored_from(
    client, superuser_token_headers
):
    points = client.get(
        f"{settings.API_V1_STR}/runs/reuse-1/metrics",
        headers=superuser_token_headers,
    ).json()

    # Named for the flow that asked, not the one that recorded it.
    assert [point["name"] for point in points] == ["quick.loss"] * 3
    assert [point["value"] for point in points] == [3.0, 2.0, 1.0]

    named = client.get(
        f"{settings.API_V1_STR}/runs/reuse-1/metrics",
        params={"name": "quick.loss"},
        headers=superuser_token_headers,
    ).json()
    assert len(named) == 3


@pytest.mark.usefixtures("reused_run")
def test_a_curve_whose_run_is_gone_is_empty_rather_than_an_error(
    client, superuser_token_headers
):
    """Deleting a flow deletes its runs; what pointed at one is left holding it."""
    with Session(db_engine) as session:
        session.delete(session.get(Run, "src-1"))
        session.commit()

    answer = client.get(
        f"{settings.API_V1_STR}/runs/reuse-1/metrics",
        headers=superuser_token_headers,
    )
    assert answer.status_code == 200
    assert answer.json() == []


def _metric(run_id: str, name: str, step: int, value: float, ts: float) -> RunMetric:
    return RunMetric(run_id=run_id, name=name, step=step, value=value, ts=ts)


@pytest.fixture
def plotted():
    """A run with a loss curve and an epoch counter beside it."""
    run_id = new_run_id()
    with Session(db_engine) as session:
        session.add(
            Run(id=run_id, flow="study", status="ok", created_at=datetime.now(UTC))
        )
        for step, (loss, epoch) in enumerate([(1.0, 10.0), (0.5, 20.0), (0.25, 30.0)]):
            session.add(_metric(run_id, "study.loss", step, loss, 100.0 + step * 5))
            session.add(_metric(run_id, "study.epoch", step, epoch, 100.0 + step * 5))
        session.commit()
    yield run_id
    with Session(db_engine) as session:
        for row in session.exec(
            select(RunMetric).where(col(RunMetric.run_id) == run_id)
        ).all():
            session.delete(row)
        session.delete(session.get(Run, run_id))
        session.commit()


def test_a_comparison_is_plotted_against_the_step_by_default(
    client, superuser_token_headers, plotted
):
    answer = client.get(
        f"{settings.API_V1_STR}/runs/series/compare",
        params={"ids": plotted, "metric": "study.loss"},
        headers=superuser_token_headers,
    ).json()

    assert answer["x"] == "step"
    assert answer["lines"][0]["points"] == [[0.0, 1.0], [1.0, 0.5], [2.0, 0.25]]


def test_time_is_measured_from_this_runs_own_first_reading(
    client, superuser_token_headers, plotted
):
    """Runs started hours apart still lie on top of each other."""
    answer = client.get(
        f"{settings.API_V1_STR}/runs/series/compare",
        params={"ids": plotted, "metric": "study.loss", "x": "time"},
        headers=superuser_token_headers,
    ).json()

    assert answer["x"] == "time"
    assert answer["lines"][0]["points"] == [[0.0, 1.0], [5.0, 0.5], [10.0, 0.25]]


def test_one_metric_can_be_plotted_against_another(
    client, superuser_token_headers, plotted
):
    answer = client.get(
        f"{settings.API_V1_STR}/runs/series/compare",
        params={"ids": plotted, "metric": "study.loss", "x": "study.epoch"},
        headers=superuser_token_headers,
    ).json()

    assert answer["lines"][0]["points"] == [[10.0, 1.0], [20.0, 0.5], [30.0, 0.25]]


def test_a_step_the_x_metric_never_reached_is_left_out(
    client, superuser_token_headers, plotted
):
    """The join is on the step, which is the only thing two series share."""
    with Session(db_engine) as session:
        session.add(_metric(plotted, "study.loss", 3, 0.1, 120.0))
        session.commit()

    answer = client.get(
        f"{settings.API_V1_STR}/runs/series/compare",
        params={"ids": plotted, "metric": "study.loss", "x": "study.epoch"},
        headers=superuser_token_headers,
    ).json()

    assert [point[0] for point in answer["lines"][0]["points"]] == [10.0, 20.0, 30.0]