#!/bin/sh
# Materia Supervisor - Continuous pipeline orchestration
# Inspired by TigerBeetle's CFO supervisor: simple, resilient, easy to understand
# https://github.com/tigerbeetle/tigerbeetle/blob/main/src/scripts/cfo_supervisor.sh
#
# Environment variables (set in systemd EnvironmentFile):
#   LANDING_DIR        — local path for extracted landing data
#   DUCKDB_PATH        — path to DuckDB lakehouse file
#   ALERT_WEBHOOK_URL  — optional ntfy.sh / Slack / Telegram webhook for failure alerts

set -eu

readonly REPO_DIR="/opt/materia"

while true
do
    (
        # Clone repo if missing
        if ! [ -d "$REPO_DIR/.git" ]
        then
            echo "Repository not found, bootstrap required!"
            exit 1
        fi

        cd "$REPO_DIR"

        # Update code from git
        git fetch origin master
        git switch --discard-changes --detach origin/master
        uv sync

        # Extract all data sources
        LANDING_DIR="${LANDING_DIR:-/data/materia/landing}" \
        DUCKDB_PATH="${DUCKDB_PATH:-/data/materia/lakehouse.duckdb}" \
            uv run materia pipeline run extract

        LANDING_DIR="${LANDING_DIR:-/data/materia/landing}" \
        DUCKDB_PATH="${DUCKDB_PATH:-/data/materia/lakehouse.duckdb}" \
            uv run materia pipeline run extract_cot

        LANDING_DIR="${LANDING_DIR:-/data/materia/landing}" \
        DUCKDB_PATH="${DUCKDB_PATH:-/data/materia/lakehouse.duckdb}" \
            uv run materia pipeline run extract_prices

        LANDING_DIR="${LANDING_DIR:-/data/materia/landing}" \
        DUCKDB_PATH="${DUCKDB_PATH:-/data/materia/lakehouse.duckdb}" \
            uv run materia pipeline run extract_ice

        # Transform all data sources
        LANDING_DIR="${LANDING_DIR:-/data/materia/landing}" \
        DUCKDB_PATH="${DUCKDB_PATH:-/data/materia/lakehouse.duckdb}" \
            uv run materia pipeline run transform

    ) || {
        # Notify on failure if webhook is configured, then sleep to avoid busy-loop
        if [ -n "${ALERT_WEBHOOK_URL:-}" ]; then
            curl -s -d "Materia pipeline failed at $(date)" "$ALERT_WEBHOOK_URL" 2>/dev/null || true
        fi
        sleep 600  # Sleep 10 min on failure
    }
done