dags.py

import json

import numpy as np
from csst_dfs_client import plan, level0

from ._base_dag import BaseDAG
from ..csst import csst


# DAG_DETECTOR_NAMES = {
#     "csst-msc-l1-mbi": {"detector": csst["msc"]["mbi"].effective_detector_names},
#     "csst-msc-l1-ast": {"detector": csst["msc"]["mbi"].effective_detector_names},
#     "csst-msc-l1-sls": {"detector": csst["msc"]["sls"].effective_detector_names},
#     "csst-msc-l1-qc0": {"detector": csst["msc"].effective_detector_names},
#     "csst-msc-l1-ooc": {"detector": csst["msc"].effective_detector_names},
#     "csst-mci-l1": {"detector": csst["mci"].effective_detector_names},
#     "csst-ifs-l1-rss": {"detector": csst["ifs"].effective_detector_names},
#     "csst-cpic-l1": {"detector": csst["cpic"].effective_detector_names},
#     "csst-cpic-l1-qc0": {"detector": csst["cpic"].effective_detector_names},
#     "csst-hstdm-l1": {"detector": csst["cpic"].effective_detector_names},
# }


def get_detector_names_via_dag(dag: str) -> list[str]:
    if "msc" in dag:
        if "mbi" in dag:
            return csst["msc"]["mbi"].effective_detector_names
        elif "sls" in dag:
            return csst["msc"]["sls"].effective_detector_names
        else:
            return csst["msc"].effective_detector_names
    elif "mci" in dag:
        return csst["mci"].effective_detector_names
    elif "ifs" in dag:
        return csst["ifs"].effective_detector_names
    elif "cpic" in dag:
        return csst["cpic"].effective_detector_names
    elif "hstdm" in dag:
        return csst["hstdm"].effective_detector_names


class GeneralDAGViaObsid(BaseDAG):

    def __init__(self, dag_group: str, dag: str, use_detector: bool = False):
        super().__init__(dag_group=dag_group, dag=dag, use_detector=use_detector)
        # set effective detector names
        self.effective_detector_names = get_detector_names_via_dag(dag)

    def schedule(
        self,
        batch_id: str | None = "-",
        priority: int = 1,
        dataset: str = "csst-msc-c9-25sqdeg-v3",
        obs_type: str = "WIDE",
        obs_group: str = "W1",
        pmapname: str = "csst_000001.map",
        initial_prc_status: int = -1024,  # level0 prc_status level1
        final_prc_status: int = -2,
        demo=True,
    ):
        # no need to query plan
        # plan.write_file(local_path="plan.json")
        # plan.find(
        #     instrument="MSC",
        #     dataset=dataset,
        #     obs_type=obs_type,
        #     project_id=project_id,
        # )

        # generate a dag_group_run
        dag_group_run = self.gen_dag_group_run(
            dag_group=self.dag_group,
            batch_id=batch_id,
            priority=priority,
        )
        dag_run_list = []

        # find level0 data records
        recs = level0.find(
            instrument=self.instrument,
            dataset=dataset,
            obs_type=obs_type,
            obs_group=obs_group,
            prc_status=initial_prc_status,
        )
        assert recs.success, recs
        print(f"{len(recs.data)} records -> ", end="")

        if self.use_detector:
            # generate DAG messages via obs_id-detector
            for this_rec in recs.data:

                # filter level0 data records: detector in expected list
                if this_rec["detector"] in self.effective_detector_names:
                    # generate a DAG message if is_selected
                    this_dag_run = self.gen_dag_run(
                        dag_group_run=dag_group_run,
                        dag_run=self.generate_sha1(),
                        batch_id=batch_id,
                        pmapname=pmapname,
                        dataset=dataset,
                        obs_type=obs_type,
                        obs_group=obs_group,
                        obs_id=this_rec["obs_id"],
                        detector=this_rec["detector"],
                    )
                    dag_run_list.append(this_dag_run)
                    # update level0 prc_status
                    if not demo:
                        this_update = level0.update_prc_status(
                            level0_id=this_rec["level0_id"],
                            dag_run=this_dag_run["dag_run"],
                            prc_status=final_prc_status,
                            dataset=dataset,
                        )
                        assert this_update.success, this_update.message

            # generate DAG messages via obs_id
            else:
                u_obsid, c_obsid = np.unique(
                    [this_rec["obs_id"] for this_rec in recs.data],
                    return_counts=True,
                )
                # select those obs_ids with `counts == effective detector number`
                u_obsid_selected = u_obsid[
                    c_obsid == len(self.effective_detector_names)
                ]
                for this_obsid in u_obsid[u_obsid_selected]:
                    # generate a DAG message if is_selected
                    this_dag_run = self.gen_dag_run(
                        dag_group_run=dag_group_run,
                        dag_run=self.generate_sha1(),
                        batch_id=batch_id,
                        pmapname=pmapname,
                        dataset=dataset,
                        obs_type=obs_type,
                        obs_group=obs_group,
                        obs_id=this_obsid,
                    )
                    dag_run_list.append(this_dag_run)

        if not demo:
            # push and update
            res_push = self.push_dag_group_run(dag_group_run, dag_run_list)
            print(
                f"{len(dag_run_list)} DAG runs -> "
                f"{json.dumps(dag_group_run, indent=None, separators=(',', ':'))} -> "
                f"{res_push}"
            )
            assert res_push.success, res_push.message
        else:
            # no push
            print(
                f"{len(dag_run_list)} DAG runs -> "
                f"{json.dumps(dag_group_run, indent=None, separators=(',', ':'))}"
            )
        # TODO: `dag_group_run` and `dag_run_list` should be dumped to a text file in the future
        return dict(
            dag_group_run=dag_group_run,
            dag_run_list=dag_run_list,
        )


class GeneralDAGViaObsgroup(BaseDAG):

    def __init__(self, dag_group: str, dag: str, use_detector: bool = False):
        super().__init__(dag_group=dag_group, dag=dag, use_detector=use_detector)
        # set effective detector names
        self.effective_detector_names = get_detector_names_via_dag(dag)

    def schedule(
        self,
        batch_id: str | None = "-",
        priority: int = 1,
        dataset: str = "csst-msc-c9-25sqdeg-v3",
        obs_type: str = "WIDE",
        obs_group: str = "W1",
        pmapname: str = "csst_000001.map",
        initial_prc_status: int = -1024,  # level0 prc_status level1
        final_prc_status: int = -2,
        demo=True,
    ):
        # generate a dag_group_run
        dag_group_run = self.gen_dag_group_run(
            dag_group=self.dag_group,
            batch_id=batch_id,
            priority=priority,
        )
        dag_run_list = []

        # find plan with compact mode
        plan.count_plan_level0(
            instrument=self.instrument,
            obs_type=obs_type,
            obs_group=obs_group,
            dataset=dataset,
            prc_status=initial_prc_status,
        )
        res_plan_level0 = plan.count_plan_level0(
            instrument="MSC",
            obs_type="WIDE",
            obs_group="W1",
            dataset="csst-msc-c9-25sqdeg-v3",
            prc_status=-1024,
        )
        assert res_plan_level0.success, res_plan_level0
        n_plan = res_plan_level0.data["plan_count"]
        n_level0 = res_plan_level0.data["level0_count"]

        # find level0 records
        if n_plan == 0:
            print(f"No plan found for {obs_type} {obs_group} {dataset}")
        if n_level0 < n_plan * self.n_effective_detector:
            print(
                f"Plan {obs_type} {obs_group} {dataset} has {n_plan} plans, "
                f"but {n_level0} level0 records found"
            )
        if n_plan == n_level0 * self.n_effective_detector:
            print(
                f"Plan {obs_type} {obs_group} {dataset} has {n_plan} plans, "
                f"and {n_level0} level0 records found"
            )

            # generate DAG run list
            if not self.use_detector:
                # generate a DAG run via obs_group
                this_dag_run = self.gen_dag_run(
                    dag_group_run=dag_group_run,
                    batch_id=batch_id,
                    dag_run=self.generate_sha1(),
                    dataset=dataset,
                    obs_type=obs_type,
                    obs_group=obs_group,
                    pmapname=pmapname,
                )
                dag_run_list.append(this_dag_run)
            else:
                # generate DAG runs via obs_group-detectors
                for this_detector in self.effective_detector_names:

                    this_dag_run = self.gen_dag_run(
                        dag_group_run=dag_group_run,
                        batch_id=batch_id,
                        dag_run=self.generate_sha1(),
                        dataset=dataset,
                        obs_type=obs_type,
                        obs_group=obs_group,
                        detector=this_detector,
                        pmapname=pmapname,
                    )
                    dag_run_list.append(this_dag_run)
            if not demo:
                # push and update
                res_push = self.push_dag_group_run(dag_group_run, dag_run_list)
                print(
                    f"{len(dag_run_list)} DAG runs -> "
                    f"{json.dumps(dag_group_run, indent=None, separators=(',', ':'))} -> "
                    f"{res_push}"
                )

        return dict(
            dag_group_run=dag_group_run,
            dag_run_list=dag_run_list,
        )