Finds orphaned quantities in farmOS - not referenced by any current log revision

find_orphaned_quantities.ipynb

{
  "metadata": {
    "kernelspec": {
      "name": "python",
      "display_name": "Python (Pyodide)",
      "language": "python"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "python",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8"
    }
  },
  "nbformat_minor": 5,
  "nbformat": 4,
  "cells": [
    {
      "id": "952a6f9d-30db-4a1c-898e-f098c2b58d54",
      "cell_type": "code",
      "source": "from pyodide.http import pyfetch\nfrom js import location\nimport json\n\nasync def get_all_entities(base_url):\n    next_page = base_url\n    while next_page is not None:\n        resp = await pyfetch(next_page, method='GET')\n\n        data = await resp.json()\n\n        for entity in data.get('data', []):\n            yield entity\n\n        next_page = data.get('links', {}).get('next', {}).get('href', None)\n",
      "metadata": {
        "trusted": true
      },
      "outputs": [],
      "execution_count": 1
    },
    {
      "id": "3132a1de-2720-4d00-ba73-7903dc8b0666",
      "cell_type": "code",
      "source": "all_quantity_types = [q['attributes']['drupal_internal__id'] async for q in get_all_entities(location.origin + '/api/quantity_type/quantity_type')]\n\nprint(all_quantity_types)",
      "metadata": {
        "trusted": true
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": "['material', 'standard']\n"
        }
      ],
      "execution_count": 2
    },
    {
      "id": "32a3b012-2503-4962-a977-f5e9ae9eaaae",
      "cell_type": "code",
      "source": "all_quantity_ids = {q['id'] for quantity_type in all_quantity_types async for q in get_all_entities(f'{location.origin}/api/quantity/{quantity_type}?fields[quantity--{quantity_type}]=id&sort=drupal_internal__id')}\n\nlen(all_quantity_ids)",
      "metadata": {
        "trusted": true
      },
      "outputs": [
        {
          "execution_count": 14,
          "output_type": "execute_result",
          "data": {
            "text/plain": "5538"
          },
          "metadata": {}
        }
      ],
      "execution_count": 14
    },
    {
      "id": "4034aa2c-0952-4b61-b17c-f88b43e4fe57",
      "cell_type": "code",
      "source": "all_log_types = [q['attributes']['drupal_internal__id'] async for q in get_all_entities(location.origin + '/api/log_type/log_type')]\n\nprint(all_log_types)",
      "metadata": {
        "trusted": true
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": "['activity', 'birth', 'harvest', 'input', 'maintenance', 'observation', 'seeding']\n"
        }
      ],
      "execution_count": 5
    },
    {
      "id": "62b7d49e-a1d7-412c-ab56-54c47510c0a2",
      "cell_type": "code",
      "source": "# Kind of a weird function: iterate through the pages, but yield the included entities\nasync def get_all_entities_includes(base_url):\n    next_page = base_url\n    while next_page is not None:\n        resp = await pyfetch(next_page, method='GET')\n\n        data = await resp.json()\n\n        for entity in data.get('included', []):\n            yield entity\n\n        next_page = data.get('links', {}).get('next', {}).get('href', None)\n\nquantity_fields_query_part = '&'.join([f'fields[quantity--{quantity_type}]=id' for quantity_type in all_quantity_types])\n\nlogs_with_quantities_filter = 'filter[a][condition][path]=quantity.id&filter[a][condition][operator]=IS NOT NULL'\n\nbase_query = lambda log_type: f'{location.origin}/api/log/{log_type}?{logs_with_quantities_filter}&sort=drupal_internal__id&fields[log--{log_type}]=&include=quantity&{quantity_fields_query_part}'\n\nall_referenced_quantity_ids = {q['id'] for log_type in all_log_types async for q in get_all_entities_includes(base_query(log_type))}\n\nlen(all_referenced_quantity_ids)",
      "metadata": {
        "trusted": true
      },
      "outputs": [
        {
          "execution_count": 10,
          "output_type": "execute_result",
          "data": {
            "text/plain": "5514"
          },
          "metadata": {}
        }
      ],
      "execution_count": 10
    },
    {
      "id": "91763140-6784-49c5-991e-d4e8911488d7",
      "cell_type": "code",
      "source": "len(all_quantity_ids - all_referenced_quantity_ids)",
      "metadata": {
        "trusted": true
      },
      "outputs": [
        {
          "execution_count": 15,
          "output_type": "execute_result",
          "data": {
            "text/plain": "24"
          },
          "metadata": {}
        }
      ],
      "execution_count": 15
    },
    {
      "id": "89c882f9-313d-4e87-885e-35241bbef6c9",
      "cell_type": "code",
      "source": "len(all_referenced_quantity_ids - all_quantity_ids)",
      "metadata": {
        "trusted": true
      },
      "outputs": [
        {
          "execution_count": 16,
          "output_type": "execute_result",
          "data": {
            "text/plain": "0"
          },
          "metadata": {}
        }
      ],
      "execution_count": 16
    }
  ]
}