import os
import json
import yaml
from datetime import datetime
from fastapi import APIRouter, status
from app.core.config import settings
from app.schemas.sitomat import (
    BundleItem,
    BundleListResponse,
    SitomatProvisionRequest,
    SitomatProvisionResponse,
    SitomatTriggerRequest,
    SitomatTriggerResponse,
    SitomatLedgerCallback,
)

router = APIRouter(prefix="/sitomat", tags=["sitomat Ingestion Governance"])

@router.get("/health")
def sitomat_health():
    """Health status of sitomat control plane integration bridge."""
    return {
        "status": "online",
        "service": "MERICA sitomat Governance Bridge",
        "supported_bundles": ["dlh-core-ingestion", "bpkp-perizinan"]
    }

@router.get("/bundles", response_model=BundleListResponse)
def list_dag_bundles():
    """
    List configured Airflow 3 GitDagBundles available for pipeline binding.
    Matches AIRFLOW__DAG_PROCESSOR__DAG_BUNDLE_CONFIG_LIST specifications.
    """
    return {
        "bundles": [
            {
                "name": "dlh-core-ingestion",
                "tracking_ref": "development",
                "repo_url": "https://repo.sicabe.com/sekota/merica-beta.git",
                "subdir": "airflow/configs",
                "type": "git"
            },
            {
                "name": "bpkp-perizinan",
                "tracking_ref": "development",
                "repo_url": "https://repo.sicabe.com/sekota/merica-beta.git",
                "subdir": "airflow/configs",
                "type": "git"
            }
        ]
    }

@router.post("/pipelines/provision", response_model=SitomatProvisionResponse, status_code=status.HTTP_201_CREATED)
def provision_sitomat_pipeline(req: SitomatProvisionRequest):
    """
    Control Plane Handshake (Gate-1):
    Register metadata, compile initial YAML config for Airflow Git DAG Bundle,
    and create initial semantic version snapshot (v1.0.0).
    """
    clean_id = req.dataset_id.lower().replace(" ", "_")
    bronze_dag_id = f"ingestion_{clean_id}"
    transform_dag_id = f"transform_{clean_id}"
    bundle_name = req.bundle_name or "dlh-core-ingestion"
    initial_version = "v1.0.0"
    now_iso = datetime.utcnow().isoformat() + "Z"
    
    # 1. Canonical YAML specification for Airflow DAG builder
    yaml_config = {
        "pipeline_id": bronze_dag_id,
        "dataset_id": req.dataset_id,
        "bundle_name": bundle_name,
        "version": initial_version,
        "schedule": req.schedule or "0 2 * * *",
        "mode": "medallion",
        "bronze": {
            "source_type": req.source_type,
            "connection_id": req.connection_id,
            "source_config": req.source_config,
            "target": req.bronze_target or {
                "s3_bucket": "dlh-landing",
                "s3_prefix": f"bronze/raw/{clean_id}/"
            }
        },
        "silver": {
            "transformation": {
                "select_models": req.dbt_models.get("silver", f"stg_{clean_id}") if req.dbt_models else f"stg_{clean_id}"
            }
        },
        "gold": {
            "transformation": {
                "select_models": req.dbt_models.get("gold", f"fct_{clean_id}") if req.dbt_models else f"fct_{clean_id}"
            }
        }
    }
    
    os.makedirs(settings.CONFIGS_DIR, exist_ok=True)
    target_meta_dir = getattr(settings, "METADATA_DIR", settings.CONFIGS_DIR)
    os.makedirs(target_meta_dir, exist_ok=True)

    yaml_filename = f"{bronze_dag_id}.yaml"
    yaml_path = os.path.join(settings.CONFIGS_DIR, yaml_filename)
    yaml_header = (
        f"# ============================================================\n"
        f"# MERICA Studio Data Lakehouse Pipeline\n"
        f"# Airflow 3 DAG Bundle: {bundle_name} | Dataset: {req.dataset_id}\n"
        f"# Initial Version: {initial_version} | Provisioned: {now_iso}\n"
        f"# ============================================================\n"
    )
    with open(yaml_path, "w", encoding="utf-8") as f:
        f.write(yaml_header + yaml.dump(yaml_config, sort_keys=False))

    # 2. JSON metadata for UI Canvas load
    json_path = os.path.join(target_meta_dir, f"{bronze_dag_id}.json")
    canvas_metadata = {
        "id": bronze_dag_id,
        "dataset_id": req.dataset_id,
        "bundle_name": bundle_name,
        "version": initial_version,
        "changelog": "Initial provisioning via sitomat Control Plane",
        "published_at": now_iso,
        "is_draft": False,
        "status": "deployed",
        "nodes": [],
        "edges": []
    }
    with open(json_path, "w", encoding="utf-8") as f:
        json.dump(canvas_metadata, f, indent=2)

    # 3. Create initial version snapshot file for Time-Machine
    snapshot_filename = f"{bronze_dag_id}.version.{initial_version}.json"
    snapshot_path = os.path.join(target_meta_dir, snapshot_filename)
    with open(snapshot_path, "w", encoding="utf-8") as f:
        json.dump(canvas_metadata, f, indent=2)
        
    return SitomatProvisionResponse(
        status="provisioned",
        dataset_id=req.dataset_id,
        bundle_name=bundle_name,
        bronze_dag_id=bronze_dag_id,
        transform_dag_id=transform_dag_id,
        initial_version=initial_version,
        git_commit_hash="local-dev",
        yaml_config_path=yaml_path
    )

@router.post("/runs/trigger", response_model=SitomatTriggerResponse)
def trigger_sitomat_run(req: SitomatTriggerRequest):
    """
    Data Plane Trigger:
    sitomat triggers an ingestion execution with correlated ledger run ID and pinned version.
    """
    clean_id = req.dataset_id.lower().replace(" ", "_")
    dag_id = f"ingestion_{clean_id}"
    dag_run_id = f"sitomat_run__{req.correlation_id}"
    pinned_version = req.target_dag_version or "latest"
    
    return SitomatTriggerResponse(
        status="triggered",
        correlation_id=req.correlation_id,
        dag_id=dag_id,
        dag_run_id=dag_run_id,
        pinned_version=pinned_version
    )

@router.post("/runs/callback")
def sitomat_run_callback(payload: SitomatLedgerCallback):
    """
    Ingestion Run Ledger Callback:
    Airflow task notifies sitomat Control Plane when Bronze extraction completes.
    """
    return {
        "status": "received",
        "correlation_id": payload.correlation_id,
        "dataset_id": payload.dataset_id,
        "message": f"Ledger updated for run '{payload.dag_run_id}' with status '{payload.status}'."
    }