06: PDF_to_Excel_using_Paddle_OCR.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/gulabpatel/Table_Detection/blob/main/06%3A%20PDF_to_Excel_using_Paddle_OCR.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Code walk through video : https://www.youtube.com/watch?v=HZh31OGiQRQ"
      ],
      "metadata": {
        "id": "LI1avyCTIDAr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from PIL import Image\n",
        "import cv2\n",
        "import numpy as np\n",
        "import pandas as pd \n",
        "import tensorflow as tf"
      ],
      "metadata": {
        "id": "mhEjVUbiUytV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dvKiMyj2feUe"
      },
      "source": [
        "# pdf2image"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Installation"
      ],
      "metadata": {
        "id": "aIz56VvlWkf-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install pdf2image\n",
        "!apt-get install poppler-utils"
      ],
      "metadata": {
        "id": "biSgyAkWnUD3"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Conversion"
      ],
      "metadata": {
        "id": "pLaguCz1WmF9"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from pdf2image import convert_from_path"
      ],
      "metadata": {
        "id": "HI1dzCGmkeWI"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "id": "6tdK6DMgTOHI",
        "outputId": "ebf01007-7f23-42ed-af77-b7b977f0a82b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!wget \"https://onedrive.live.com/download?cid=DBD24D3DC9712F7A&resid=DBD24D3DC9712F7A%2110213&authkey=AKAJrobqemGhan8\" -O data.zip\n",
        "!unzip -qq data.zip"
      ],
      "metadata": {
        "id": "bCnthQoZLfFk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "images = convert_from_path('/content/drive/MyDrive/documents/BScI.pdf')\n",
        "# images = cv2.imread(\"/content/pages/page2.jpg\")"
      ],
      "metadata": {
        "id": "xyhjFgqjkgG7"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!mkdir pages"
      ],
      "metadata": {
        "id": "M5hBWxpskgJW"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for i in range(len(images)):\n",
        "  images[i].save('pages/page'+str(i)+'.jpg', 'JPEG')"
      ],
      "metadata": {
        "id": "7mlH9ltkkgMB"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Layout"
      ],
      "metadata": {
        "id": "Urwb7Q_OZC8t"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Installation"
      ],
      "metadata": {
        "id": "vTkwQbF4FZPN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!python3 -m pip install paddlepaddle-gpu\n",
        "!pip install \"paddleocr>=2.0.1\"\n",
        "!pip install protobuf==3.20.0\n",
        "!git clone https://github.com/PaddlePaddle/PaddleOCR.git"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "R0lRpQLkIpbZ",
        "outputId": "03d95448-bc5f-4188-cb02-74ff75aa3c45"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "remote: Total 38584 (delta 337), reused 480 (delta 296), pack-reused 38024\u001b[K\n",
            "Receiving objects: 100% (38584/38584), 320.51 MiB | 32.24 MiB/s, done.\n",
            "Resolving deltas: 100% (26847/26847), done.\n",
            "Checking out files: 100% (1277/1277), done.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!wget https://paddleocr.bj.bcebos.com/whl/layoutparser-0.0.0-py3-none-any.whl\n",
        "!pip install -U layoutparser-0.0.0-py3-none-any.whl"
      ],
      "metadata": {
        "id": "8gWFgGl5CXu6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Table Extraction"
      ],
      "metadata": {
        "id": "w5MA08E0F8aU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import cv2\n",
        "import layoutparser as lp\n",
        "image = cv2.imread(\"/content/pages/page0.jpg\")\n",
        "image = image[..., ::-1]\n",
        "\n",
        "# load model\n",
        "model = lp.PaddleDetectionLayoutModel(config_path=\"lp://PubLayNet/ppyolov2_r50vd_dcn_365e_publaynet/config\",\n",
        "                                threshold=0.5,\n",
        "                                label_map={0: \"Text\", 1: \"Title\", 2: \"List\", 3:\"Table\", 4:\"Figure\"},\n",
        "                                enforce_cpu=False,\n",
        "                                enable_mkldnn=True)#math kernel library\n",
        "# detect\n",
        "layout = model.detect(image)"
      ],
      "metadata": {
        "id": "bw9SFYnMCX0E"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "layout"
      ],
      "metadata": {
        "id": "G76oO2WXCX6v"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for l in layout:\n",
        "  # if l.type == 'Table':\n",
        "    x_1 = int(l.block.x_1)\n",
        "    y_1 = int(l.block.y_1)\n",
        "    x_2 = int(l.block.x_2)\n",
        "    y_2 = int(l.block.y_2)\n",
        "    \n",
        "    break"
      ],
      "metadata": {
        "id": "y3h0kCz-CX_U"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(x_1,y_1,x_2,y_2)"
      ],
      "metadata": {
        "id": "cdiYPeCJCYIg"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "im = cv2.imread('/content/pages/page0.jpg')"
      ],
      "metadata": {
        "id": "F39kJV3hCYLV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "cv2.imwrite('ext_im.jpg', im[y_1:y_2,x_1:x_2])"
      ],
      "metadata": {
        "id": "EQDXSNijCYPs"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EGwGhHnd8i_h"
      },
      "source": [
        "# Text Detection and Recognition"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from paddleocr import PaddleOCR, draw_ocr"
      ],
      "metadata": {
        "id": "N6WQZXhLLDWk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "ocr = PaddleOCR(lang='en')\n",
        "image_path = '/content/ext_im.jpg'\n",
        "image_cv = cv2.imread(image_path)\n",
        "image_height = image_cv.shape[0]\n",
        "image_width = image_cv.shape[1]\n",
        "output = ocr.ocr(image_path)"
      ],
      "metadata": {
        "id": "A8bCZ9AULDZF"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(output)"
      ],
      "metadata": {
        "id": "VBIDZA0XLDeA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "boxes = [line[0] for line in output]\n",
        "texts = [line[1][0] for line in output]\n",
        "probabilities = [line[1][1] for line in output]"
      ],
      "metadata": {
        "id": "nNMBAQ78LDgG"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "image_boxes = image_cv.copy()"
      ],
      "metadata": {
        "id": "uukcg4SWV_dg"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for box,text in zip(boxes,texts):\n",
        "  cv2.rectangle(image_boxes, (int(box[0][0]),int(box[0][1])), (int(box[2][0]),int(box[2][1])),(0,0,255),1)\n",
        "  cv2.putText(image_boxes, text,(int(box[0][0]),int(box[0][1])),cv2.FONT_HERSHEY_SIMPLEX,1,(222,0,0),1)"
      ],
      "metadata": {
        "id": "l_HzbiA7V_fw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "cv2.imwrite('detections.jpg', image_boxes)"
      ],
      "metadata": {
        "id": "PfUG9mcgV_iJ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kYWt0lzDHZNp"
      },
      "source": [
        "# Reconstruction"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Get Horizontal and Vertical Lines"
      ],
      "metadata": {
        "id": "ruzifYJz4H6y"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "im = image_cv.copy()"
      ],
      "metadata": {
        "id": "YLIoKedcqby_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "horiz_boxes = []\n",
        "vert_boxes = []\n",
        "\n",
        "for box in boxes:\n",
        "  x_h, x_v = 0,int(box[0][0])\n",
        "  y_h, y_v = int(box[0][1]),0\n",
        "  width_h,width_v = image_width, int(box[2][0]-box[0][0])\n",
        "  height_h,height_v = int(box[2][1]-box[0][1]),image_height\n",
        "\n",
        "  horiz_boxes.append([x_h,y_h,x_h+width_h,y_h+height_h])\n",
        "  vert_boxes.append([x_v,y_v,x_v+width_v,y_v+height_v])\n",
        "\n",
        "  cv2.rectangle(im,(x_h,y_h), (x_h+width_h,y_h+height_h),(0,0,255),1)\n",
        "  cv2.rectangle(im,(x_v,y_v), (x_v+width_v,y_v+height_v),(0,255,0),1)\n",
        "  "
      ],
      "metadata": {
        "id": "GwcAAe-wccnF"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "cv2.imwrite('horiz_vert.jpg',im)"
      ],
      "metadata": {
        "id": "7UxFGhMkccph"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Non-Max Suppression"
      ],
      "metadata": {
        "id": "ekVFvJrM4ROL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "horiz_out = tf.image.non_max_suppression(\n",
        "    horiz_boxes,\n",
        "    probabilities,\n",
        "    max_output_size = 1000,\n",
        "    iou_threshold=0.1,\n",
        "    score_threshold=float('-inf'),\n",
        "    name=None\n",
        ")"
      ],
      "metadata": {
        "id": "4LVSSB2fcoe7"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "horiz_lines = np.sort(np.array(horiz_out))\n",
        "print(horiz_lines)"
      ],
      "metadata": {
        "id": "pOboYpGnccr2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "im_nms = image_cv.copy()"
      ],
      "metadata": {
        "id": "pfxHrn3iccyK"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for val in horiz_lines:\n",
        "  cv2.rectangle(im_nms, (int(horiz_boxes[val][0]),int(horiz_boxes[val][1])), (int(horiz_boxes[val][2]),int(horiz_boxes[val][3])),(0,0,255),1)\n",
        "  "
      ],
      "metadata": {
        "id": "68PCHfmZcc0L"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "cv2.imwrite('im_nms.jpg',im_nms)"
      ],
      "metadata": {
        "id": "Z8r9qpiAcc2X"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "vert_out = tf.image.non_max_suppression(\n",
        "    vert_boxes,\n",
        "    probabilities,\n",
        "    max_output_size = 1000,\n",
        "    iou_threshold=0.1,\n",
        "    score_threshold=float('-inf'),\n",
        "    name=None\n",
        ")"
      ],
      "metadata": {
        "id": "mKgPuh7rcc4s"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(vert_out)"
      ],
      "metadata": {
        "id": "GBpKsImVcc6p"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "vert_lines = np.sort(np.array(vert_out))\n",
        "print(vert_lines)"
      ],
      "metadata": {
        "id": "K0lBh-yz5YLp"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for val in vert_lines:\n",
        "  cv2.rectangle(im_nms, (int(vert_boxes[val][0]),int(vert_boxes[val][1])), (int(vert_boxes[val][2]),int(vert_boxes[val][3])),(255,0,0),1)\n",
        "  "
      ],
      "metadata": {
        "id": "WqsLm0L_cc84"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "cv2.imwrite('im_nms.jpg',im_nms)"
      ],
      "metadata": {
        "id": "xZOEa7lpcdGM"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Convert to CSV"
      ],
      "metadata": {
        "id": "116eBUrO93-i"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "out_array = [[\"\" for i in range(len(vert_lines))] for j in range(len(horiz_lines))]\n",
        "print(np.array(out_array).shape)\n",
        "print(out_array)"
      ],
      "metadata": {
        "id": "HRQzwVUTcdIq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "unordered_boxes = []\n",
        "\n",
        "for i in vert_lines:\n",
        "  print(vert_boxes[i])\n",
        "  unordered_boxes.append(vert_boxes[i][0])"
      ],
      "metadata": {
        "id": "sSrupaRZIAk_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "ordered_boxes = np.argsort(unordered_boxes)\n",
        "print(ordered_boxes)"
      ],
      "metadata": {
        "id": "lRMlVNh_HuJV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def intersection(box_1, box_2):\n",
        "  return [box_2[0], box_1[1],box_2[2], box_1[3]]"
      ],
      "metadata": {
        "id": "AHHaxKUuC6jQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def iou(box_1, box_2):\n",
        "\n",
        "  x_1 = max(box_1[0], box_2[0])\n",
        "  y_1 = max(box_1[1], box_2[1])\n",
        "  x_2 = min(box_1[2], box_2[2])\n",
        "  y_2 = min(box_1[3], box_2[3])\n",
        "\n",
        "  inter = abs(max((x_2 - x_1, 0)) * max((y_2 - y_1), 0))\n",
        "  if inter == 0:\n",
        "      return 0\n",
        "      \n",
        "  box_1_area = abs((box_1[2] - box_1[0]) * (box_1[3] - box_1[1]))\n",
        "  box_2_area = abs((box_2[2] - box_2[0]) * (box_2[3] - box_2[1]))\n",
        "  \n",
        "  return inter / float(box_1_area + box_2_area - inter)"
      ],
      "metadata": {
        "id": "fDVb0DkxJSIf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for i in range(len(horiz_lines)):\n",
        "  for j in range(len(vert_lines)):\n",
        "    resultant = intersection(horiz_boxes[horiz_lines[i]], vert_boxes[vert_lines[ordered_boxes[j]]] )\n",
        "\n",
        "    for b in range(len(boxes)):\n",
        "      the_box = [boxes[b][0][0],boxes[b][0][1],boxes[b][2][0],boxes[b][2][1]]\n",
        "      if(iou(resultant,the_box)>0.1):\n",
        "        out_array[i][j] = texts[b]"
      ],
      "metadata": {
        "id": "LWGhCwg6BIoL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "out_array"
      ],
      "metadata": {
        "id": "UFu9-SzXBI7B"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "out_array = np.array(out_array)"
      ],
      "metadata": {
        "id": "c4tEY9LGNIM9"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "pd.DataFrame(out_array).to_csv('sample.csv')"
      ],
      "metadata": {
        "id": "D8UdX80wBI9V"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Convert to owl format"
      ],
      "metadata": {
        "id": "M9rH5mX9Mx7g"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "owl_head = [\n",
        "  '@prefix : <http://federated.amecse-conferences.org/foodOntology#> .',\n",
        "  '@prefix owl: <http://www.w3.org/2002/07/owl#> .',\n",
        "  '@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .',\n",
        "  '@prefix xml: <http://www.w3.org/XML/1998/namespace> .',\n",
        "  '@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .',\n",
        "  '@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .',\n",
        "  '@base <http://federated.amecse-conferences.org/foodOntology> .',\n",
        "  '',\n",
        "  '<http://federated.amecse-conferences.org/foodOntology> rdf:type owl:Ontology .',\n",
        "  '',\n",
        "  '',\n",
        "  ]\n"
      ],
      "metadata": {
        "id": "sS1YYdXlTmpR"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "owl_tail = [\n",
        "  '\\n#################################################################',\n",
        "  '#    Individuals',\n",
        "  '#################################################################',\n",
        "  '',\n",
        "  '###  http://federated.amecse-conferences.org/foodOntology#Azanzi',\n",
        "  ':Azanzi rdf:type owl:NamedIndividual ,',\n",
        "  '                 :Person ;',\n",
        "  '        :name \"Azanzi Jiomekong\"^^xsd:string .',\n",
        "  '',\n",
        "  '',\n",
        "  '###  http://federated.amecse-conferences.org/foodOntology#Azza',\n",
        "  ':Azza rdf:type owl:NamedIndividual ,',\n",
        "  '               :Person ;',\n",
        "  '      :name \"Azza\"^^xsd:string .',\n",
        "  '',\n",
        "  '',\n",
        "  '###  http://federated.amecse-conferences.org/foodOntology#Etoga',\n",
        "  ':Etoga rdf:type owl:NamedIndividual ,',\n",
        "  '                :Person ;',\n",
        "  '       :name \"Etoga Cosmas\"^^xsd:string .',\n",
        "  '',\n",
        "  '',\n",
        "  '###  http://federated.amecse-conferences.org/foodOntology#Folefac',\n",
        "  ':Folefac rdf:type owl:NamedIndividual ,',\n",
        "  '                  :Person ;',\n",
        "  '         :name \"Folefac Martins\"^^xsd:string .',\n",
        "  '',\n",
        "  '',\n",
        "  '###  Generated by the OWL API (version 4.5.9.2019-02-01T07:24:44Z) https://github.com/owlcs/owlapi',\n",
        "  '',\n",
        "  ]"
      ],
      "metadata": {
        "id": "m7MYVwh9XtTi"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "owl_object_properties_head = [\n",
        "                        \n",
        "  '#################################################################',\n",
        "  '#    Object properties',\n",
        "  '#################################################################',\n",
        "  '',\n",
        "  '',\n",
        "  ]"
      ],
      "metadata": {
        "id": "52GUFI93agF2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "owl_object_properties_body = []\n",
        "for i in range(1,len(out_array)):\n",
        "  owl_object_properties_body.append([\n",
        "    '###  http://federated.amecse-conferences.org/foodOntology#'+out_array[i][0].replace(' ','').replace(',',''),\n",
        "    ':'+out_array[i][0].replace(' ','').replace(',','')+' rdf:type owl:ObjectProperty ;',\n",
        "    '     rdfs:domain :'+ out_array[i][0].replace(' ','').replace(',','')+' ;',\n",
        "    '     rdfs:range :'+ out_array[i][1].replace(' ','').replace(',','')+' .',\n",
        "    '','',\n",
        "  ])"
      ],
      "metadata": {
        "id": "IYDe69Ysb0wc"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "owl_classes_head = [\n",
        "                        \n",
        "  '\\n#################################################################',\n",
        "  '#    Classes',\n",
        "  '#################################################################',\n",
        "  '',\n",
        "  '',\n",
        "  ]"
      ],
      "metadata": {
        "id": "PkWUJtSliRiE"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "owl_classes_body = [\n",
        "  '###  http://federated.amecse-conferences.org/foodOntology#FoodComposition',\n",
        "  ':FoodComposition rdf:type owl:Class .',\n",
        "  '',\n",
        "  '\\n',\n",
        "]"
      ],
      "metadata": {
        "id": "zGq-qVX7mFAk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "owl_sub_classes_body = []\n",
        "for i in range(1,len(out_array)):\n",
        "  owl_sub_classes_body.append([\n",
        "    '###  http://federated.amecse-conferences.org/foodOntology#'+out_array[i][0].replace(' ','').replace(',',''),\n",
        "    ':'+out_array[i][0].replace(' ','').replace(',','')+' rdf:type owl:Class ;',\n",
        "    '           rdfs:subClassOf :FoodComposition .',\n",
        "    '','',                      \n",
        "  ])"
      ],
      "metadata": {
        "id": "w_qpcHZXic8j"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "with open('/content/template.owl', 'a') as f:\n",
        "    f.writelines('\\n'.join(owl_head))\n",
        "    f.writelines('\\n'.join(owl_object_properties_head))\n",
        "    f.writelines('\\n'.join(list(np.concatenate(owl_object_properties_body).flat)))\n",
        "    f.writelines('\\n'.join(owl_classes_head))\n",
        "    f.writelines('\\n'.join(owl_classes_body))\n",
        "    f.writelines('\\n'.join(list(np.concatenate(owl_sub_classes_body).flat)))\n",
        "    f.writelines('\\n'.join(owl_tail))"
      ],
      "metadata": {
        "id": "wezAMcT9Tmty"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "collapsed_sections": [
        "aIz56VvlWkf-",
        "EGwGhHnd8i_h",
        "xifVUkvP9HS4",
        "kYWt0lzDHZNp",
        "Uq0h938DH3th",
        "bU_QoW0keVpC"
      ],
      "name": "06: PDF to Excel using Paddle OCR.ipynb",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    },
    "gpuClass": "standard"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}