{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "7db4c5ed",
   "metadata": {},
   "source": [
    "# Read past an incorrect PDF text layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "182296e7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-11-01T06:46:00.672837Z",
     "iopub.status.busy": "2026-11-01T06:46:00.672662Z",
     "iopub.status.idle": "2026-11-01T06:46:11.075443Z",
     "shell.execute_reply": "2026-11-01T06:46:11.072186Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Loaded DBNet detector (CPU, threads=8)\n",
      "Loaded SVTR recognizer (vocab=185 chars, CPU, threads=8)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[occular] LM (pure Python): compact_lm.npz (268 MB), unigrams=2010937, alpha=0.5 beta=1.0 beam=100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "EXISTING TEXT LAYER\n",
      "OLD INCORRECT TEXT LAYER\n",
      "\n",
      "FRESH OCR\n",
      "INVOICE147\n",
      "Paper:2packs\n",
      "Total:200.00\n"
     ]
    }
   ],
   "source": [
    "from occular import OCRPipeline\n",
    "\n",
    "reader = OCRPipeline()\n",
    "existing = reader.get_text(\"old-ocr.pdf\", workers=1)\n",
    "fresh = reader.get_text(\"old-ocr.pdf\", force_ocr=True, workers=1)\n",
    "print(\"EXISTING TEXT LAYER\")\n",
    "print(existing[0])\n",
    "print(\"\\nFRESH OCR\")\n",
    "print(fresh[0])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}