{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "3dc7fa1d",
   "metadata": {},
   "source": [
    "# Albatross — Expert notebook (WPS + programmatic usage)\n",
    "\n",
    "This notebook provides a practical, developer-oriented walkthrough for running **Albatross** via WPS and (optionally) locally."
   ]
  },
  {
   "cell_type": "code",
   "id": "79deadf3",
   "metadata": {
    "jupyter": {
     "is_executing": true
    }
   },
   "source": [
    "import urllib.parse\n",
    "import requests\n",
    "import re\n",
    "from pathlib import Path\n",
    "\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt"
   ],
   "outputs": [],
   "execution_count": null
  },
  {
   "cell_type": "markdown",
   "id": "a3a4fc42",
   "metadata": {},
   "source": [
    "## WPS call\n",
    "\n",
    "Set your endpoint URL and build an `Execute` request (KVP encoding)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ccbdaa91",
   "metadata": {},
   "outputs": [],
   "source": [
    "WPS_URL = \"https://<your-wps-host>/wps\"  # <-- change me\n",
    "INPUT_FILE_URL = \"https://raw.githubusercontent.com/climateintelligence/albatross/main/albatross/data/APGD_Como_ppt.txt\"\n",
    "\n",
    "params = {\n",
    "    \"service\": \"WPS\",\n",
    "    \"version\": \"1.0.0\",\n",
    "    \"request\": \"Execute\",\n",
    "    \"identifier\": \"Drought\",\n",
    "    \"datainputs\": \";\".join([\n",
    "        f\"input_file={INPUT_FILE_URL}\",\n",
    "        \"target_variable=ZRB\",\n",
    "        \"index=ONI\",\n",
    "        \"glo_var=SST\",\n",
    "        \"phase_mode=2\",\n",
    "        \"months=3\",\n",
    "        \"training_start_year=1951\",\n",
    "        \"training_end_year=2015\",\n",
    "        \"target_season=1\",\n",
    "    ]),\n",
    "}\n",
    "url = f\"{WPS_URL}?{urllib.parse.urlencode(params)}\"\n",
    "print(url)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "214ffffa",
   "metadata": {},
   "outputs": [],
   "source": [
    "resp = requests.get(url, timeout=300)\n",
    "print(\"HTTP status:\", resp.status_code)\n",
    "text = resp.text\n",
    "print(text[:800])\n",
    "\n",
    "if resp.status_code != 200 or \"ProcessSucceeded\" not in text:\n",
    "    m = re.search(r\"<(?:\\w+:)?ExceptionText>(.*?)</(?:\\w+:)?ExceptionText>\", text, re.DOTALL)\n",
    "    if m:\n",
    "        raise RuntimeError(\"WPS error: \" + re.sub(r\"\\s+\", \" \", m.group(1)).strip())\n",
    "    raise RuntimeError(\"WPS execution did not succeed (inspect response above).\")\n",
    "\n",
    "print(\"✅ WPS ProcessSucceeded\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11acd18e",
   "metadata": {},
   "source": [
    "## Extract output references (if any)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "32fa45da",
   "metadata": {},
   "outputs": [],
   "source": [
    "hrefs = re.findall(r'href=\"([^\"]+)\"', text)\n",
    "hrefs = [h for h in hrefs if h.startswith(\"http\")]\n",
    "print(\"Found hrefs:\", hrefs)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6ff79083",
   "metadata": {},
   "source": [
    "## Local (developer) use\n",
    "\n",
    "If you have installed the repo (`pip install -e .`), you can inspect the process class and its inputs/outputs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2d5aefae",
   "metadata": {},
   "outputs": [],
   "source": [
    "from albatross.processes.wps_drought import Drought\n",
    "\n",
    "proc = Drought()\n",
    "print(\"Process identifier:\", proc.identifier)\n",
    "print(\"Inputs:\", [i.identifier for i in proc.inputs])\n",
    "print(\"Outputs:\", [o.identifier for o in proc.outputs])\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "pygments_lexer": "ipython3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}