{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "6cb8906f-d211-4bfe-8f2f-809879ae9408",
   "metadata": {},
   "source": [
    "# Create the Controlled Vocabulary of the DKRZ Data Catalog"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "82ce7dbe-451d-467d-8cc2-84efb69b253e",
   "metadata": {},
   "source": [
    "# Hardcode"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e9b12066-5de2-40ca-bc07-8e8a00a1bdfd",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "cv=dict()\n",
    "cv[\"required_metadata\"]=[\n",
    "    \"relation\",\n",
    "    \"account\"\n",
    "]\n",
    "cv[\"relation_types\"]=[\n",
    "    \"institution\",\n",
    "    \"project\",\n",
    "    \"model\"\n",
    "]\n",
    "cv[\"relation\"]={\n",
    "    \"mpi-m\":{\n",
    "        \"type\":\"institution\",\n",
    "        \"description\":\"Max-Planck-Institut for Meteorology\"\n",
    "    },\n",
    "    \"icon\":{\n",
    "        \"type\":\"model\",\n",
    "        \"description\":\"Earth System Model\"\n",
    "    },\n",
    "    \"nextgems\":{\n",
    "        \"type\":\"project\",\n",
    "        \"description\":\"Next Generation Earth System Models\"\n",
    "    },\n",
    "}\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ffa708ca-d4ff-4bcd-a795-ddd19667aaa3",
   "metadata": {},
   "source": [
    "# Add relations from users"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7a136684-9e4d-43d0-845b-8dce5ac09e18",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import glob\n",
    "import os\n",
    "all_lakes=[\n",
    "    relation.split('/')[-1]\n",
    "    for relation in glob.glob(\"../lake/*\")\n",
    "    if not relation.endswith(\"sources\") and not os.path.isdir(relation)\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e9121392-49a6-436d-aa0d-e8abec1a2be3",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "default_relation={\n",
    "    \"type\":\"unkown\",\n",
    "    \"description\":\"unknown\"\n",
    "}\n",
    "for relation in all_lakes:\n",
    "    if not relation in cv[\"relation\"]:\n",
    "        cv[\"relation\"][relation]=default_relation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "edfacec4-8bda-4608-9689-8e77ef03a266",
   "metadata": {},
   "source": [
    "# Add accounts from work"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "776c4575-f01a-4432-9cac-a655a01bfa70",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "cv[\"account\"]=sorted(\n",
    "    os.listdir(\"/work\")\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7b38edd5-36ab-47ff-a8df-8a289a8ad472",
   "metadata": {},
   "source": [
    "# Write CV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c8d708b3-96c5-496d-aacf-b12fdba78d9d",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a52e5904-b7bf-43ed-b27d-31c9628e59f8",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "with open(\"../metadata_cv.json\", \"w\") as f:\n",
    "    f.write(\n",
    "        json.dumps(cv,indent=4)\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e1b07d9f-e1cf-435e-86f1-1eeabd4dc830",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "dkrzcatalog",
   "language": "python",
   "name": "dkrzcatalog"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}