Norvig’s Spell Checker Algorithm for Azerbaijani Language.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 139,
   "id": "a737cf0c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\004567\\Anaconda3\\envs\\root3\\lib\\site-packages\\IPython\\core\\magics\\pylab.py:160: UserWarning: pylab import has clobbered these variables: ['sample']\n",
      "`%matplotlib` prevents importing * from pylab and numpy\n",
      "  \"\\n`%matplotlib` prevents importing * from pylab and numpy\"\n"
     ]
    }
   ],
   "source": [
    "#import libraries\n",
    "\n",
    "from __future__ import division\n",
    "%pylab inline\n",
    "import re\n",
    "import math\n",
    "import string\n",
    "from collections import Counter\n",
    "import nltk\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5197709e",
   "metadata": {},
   "source": [
    "# 1.Import corpus\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "id": "fdf90ec6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "11098116"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "TEXT = open('allbook.txt',encoding='utf-8').read()\n",
    "len(TEXT)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'n orta\\nümumitəhsil məktəblərinin biologiya müəllimləri də faydalana bilər.\\nKitabın bütövlükdə və hisələrinin üzünün çıxarılması\\nqadağandır.\\n3\\nÖN SÖZ\\nAzərbaycan dilində embriologiya haqqında məlumatlar\\n1976-cı ildə Abdullayev M.S., Abiyev H.S. “Ümumi\\n'"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "TEXT[1000:1250]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "id": "5df1436e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def tokens(text):\n",
    "    #add Azerbaijani alphabet\n",
    "    return re.findall('[ABCÇDEƏFGĞHXIİJKQLMNOÖPRSŞTUÜVYZabcçdeəfgğhxıijkqlmnoöprsştuüvyz]+', text.lower()) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "id": "b6b2dfe7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1478667"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#tokenize corpus\n",
    "WORDS = tokens(TEXT)\n",
    "len(WORDS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['və',\n",
       " 'genetika',\n",
       " 'kafedrasının',\n",
       " 'baş',\n",
       " 'müəllimi',\n",
       " 'əsgərov',\n",
       " 'lənkaran',\n",
       " 'dövlət',\n",
       " 'universiteti',\n",
       " 'dosent',\n",
       " 'şahbazov',\n",
       " 'lənkaran',\n",
       " 'dövlət',\n",
       " 'universiteti',\n",
       " 'dosent',\n",
       " 'əvəzi',\n",
       " 'ağayeva',\n",
       " 'elmi',\n",
       " 'redaktor',\n",
       " 'biol',\n",
       " 'şəmmədov',\n",
       " 'hüseynov',\n",
       " 'quliyev',\n",
       " 'embriologiya',\n",
       " 'dərs',\n",
       " 'vəsaiti',\n",
       " 'bakı',\n",
       " 'səh',\n",
       " 'birinci',\n",
       " 'nəşr',\n",
       " 'dərs',\n",
       " 'vəsaitindən',\n",
       " 'universitetlərin',\n",
       " 'biologiya',\n",
       " 'fakültəsində',\n",
       " 'bakalavr',\n",
       " 'dərəcəsi',\n",
       " 'üzrə',\n",
       " 'təhsil',\n",
       " 'alan',\n",
       " 'tələbələr',\n",
       " 'tibb',\n",
       " 'universiteti',\n",
       " 'tibb',\n",
       " 'kolleci',\n",
       " 'tələbələri',\n",
       " 'embriologiya',\n",
       " 'sahəsində',\n",
       " 'tədqiqat',\n",
       " 'aparan']"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "WORDS = [i for i in WORDS if len(i)>1]\n",
    "\n",
    "WORDS[50:100]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab253af2",
   "metadata": {},
   "source": [
    "# 2. Models: Bag of Words\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "e5238522",
   "metadata": {},
   "outputs": [],
   "source": [
    "def sample(bag, n=10):\n",
    "    return ' '.join(random.choice(bag) for _ in range(n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "04cb0365",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'insanla vacib keçid belə göstərilməsi şərqi amili görünüşün və ay'"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample(WORDS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "id": "fb690b81",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('və', 31140), ('bir', 21109), ('bu', 16103), ('də', 14191), ('ki', 12363), ('ilə', 10145), ('da', 9110), ('üçün', 7021), ('isə', 6086), ('idi', 6014)]\n"
     ]
    }
   ],
   "source": [
    "#create dictionary \n",
    "COUNTS = Counter(WORDS)\n",
    "\n",
    "print(COUNTS.most_common(10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "id": "d6e039b3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "M = COUNTS['və']\n",
    "yscale('log'); xscale('log'); title('Frequency of n-th most frequent word and 1/n line.')\n",
    "plot([c for (w, c) in COUNTS.most_common()])\n",
    "plot([M/i for i in range(1, len(COUNTS)+1)]);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "298946c4",
   "metadata": {},
   "source": [
    "# 3. Task: Spelling Correction\n",
    "\n",
    "        \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "id": "247481b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "def correct(word):\n",
    "    # Prefer edit distance 0, then 1, then 2; otherwise default to word itself.\n",
    "    candidates = (known(edits0(word)) or \n",
    "                  known(edits1(word)) or \n",
    "                  known(edits2(word)) or \n",
    "                  [word])\n",
    "    return max(candidates, key=COUNTS.get)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "id": "13c40a23",
   "metadata": {},
   "outputs": [],
   "source": [
    "def known(words):\n",
    "    return {w for w in words if w in COUNTS}\n",
    "\n",
    "def edits0(word): \n",
    "    return {word}\n",
    "\n",
    "def edits2(word):\n",
    "    return {e2 for e1 in edits1(word) for e2 in edits1(e1)}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "id": "4934b766",
   "metadata": {},
   "outputs": [],
   "source": [
    "def edits1(word):\n",
    "    pairs      = splits(word)\n",
    "    deletes    = [a+b[1:]           for (a, b) in pairs if b]\n",
    "    transposes = [a+b[1]+b[0]+b[2:] for (a, b) in pairs if len(b) > 1]\n",
    "    replaces   = [a+c+b[1:]         for (a, b) in pairs for c in alphabet if b]\n",
    "    inserts    = [a+c+b             for (a, b) in pairs for c in alphabet]\n",
    "    return set(deletes + transposes + replaces + inserts)\n",
    "\n",
    "def splits(word):\n",
    "    return [(word[:i], word[i:]) \n",
    "            for i in range(len(word)+1)]\n",
    "\n",
    "alphabet = 'ABCÇDEƏFGĞHXIİJKQLMNOÖPRSŞTUÜVYZabcçdeəfgğhxıijkqlmnoöprsştuüvyz'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "id": "b7fcc53e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('', 'necesen'),\n",
       " ('n', 'ecesen'),\n",
       " ('ne', 'cesen'),\n",
       " ('nec', 'esen'),\n",
       " ('nece', 'sen'),\n",
       " ('neces', 'en'),\n",
       " ('necese', 'n'),\n",
       " ('necesen', '')]"
      ]
     },
     "execution_count": 136,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "splits('necesen')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "22056ff5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'neİcesen', 'necesenr', 'cecesen', 'necesIn', 'Gecesen', 'necZsen', 'neFcesen', 'nevcesen', 'necDsen', 'nemcesen', 'necOesen', 'neceFen', 'necesenƏ', 'nZcesen', 'necegsen', 'nIcesen', 'necGsen', 'necesçn', 'necelen', 'Unecesen', 'neceseh', 'Üecesen', 'nüecesen', 'necesQen', 'enecesen', 'necewen', 'nŞecesen', 'neceesn', 'necvsen', 'necbesen', 'npcesen', 'necesezn', 'necewsen', 'neceseYn', 'necTesen', 'necesenÖ', 'nAcesen', 'necesehn', 'lnecesen', 'neceOsen', 'neceseb', 'necesenI', 'nEecesen', 'nJecesen', 'necesZen', 'necjesen', 'nicesen', 'nxcesen', 'nVecesen', 'neycesen', 'necesenQ', 'necesenN', 'neceseL', 'necezsen', 'necYsen', 'neceŞen', 'npecesen', 'nccesen', 'necsen', 'nycesen', 'neGcesen', 'necesej', 'bnecesen', 'necesenv', 'neceusen', 'necFesen', 'nevesen', 'neceöen', 'neceseC', 'neResen', 'eecesen', 'nPcesen', 'neƏcesen', 'neceen', 'necYesen', 'neMesen', 'neceCsen', 'neceIen', 'negesen', 'neceDen', 'necZesen', 'necnesen', 'necesePn', 'neceAsen', 'neceTen', 'necesden', 'neQcesen', 'iecesen', 'necesEen', 'neceseç', 'necəsen', 'neceyen', 'necesxn', 'nÜcesen', 'tnecesen', 'necesegn', 'nucesen', 'neOcesen', 'neNcesen', 'necesUen', 'necesenl', 'necesenX', 'Tecesen', 'necezen', 'necesev', 'İnecesen', 'neceseIn', 'ğnecesen', 'necesğen', 'necesxen', 'neLcesen', 'neÇesen', 'znecesen', 'nHecesen', 'nQecesen', 'necCsen', 'neceskn', 'neDcesen', 'necesXen', 'necesenİ', 'necesenO', 'necesenp', 'Recesen', 'neceseen', 'neciesen', 'unecesen', 'nkecesen', 'nrcesen', 'nRecesen', 'neceİsen', 'nəcesen', 'necespn', 'nuecesen', 'nezesen', 'Şecesen', 'necepen', 'neceven', 'nnecesen', 'necesexn', 'neOesen', 'necejen', 'necesXn', 'neceNsen', 'necAsen', 'necisen', 'neceRen', 'neceksen', 'neceqsen', 'neceosen', 'necesenY', 'necUsen', 'nYcesen', 'nercesen', 'neceSsen', 'neceMen', 'neceseÖn', 'Mnecesen', 'Snecesen', 'necğsen', 'nrecesen', 'necestn', 'necesenə', 'necesenq', 'Şnecesen', 'necƏesen', 'nVcesen', 'neCesen', 'necIsen', 'neceseSn', 'neXcesen', 'neJesen', 'neceBen', 'nPecesen', 'necesAen', 'necesel', 'necefen', 'necösen', 'necesPn', 'neoesen', 'Yecesen', 'nwcesen', 'necesEn', 'necesCn', 'Qnecesen', 'Iecesen', 'hnecesen', 'şnecesen', 'necüesen', 'necesenT', 'necesenf', 'nfcesen', 'necesyn', 'neceUsen', 'necesein', 'vnecesen', 'necesOn', 'necEsen', 'necesCen', 'necesÜen', 'nXecesen', 'nebcesen', 'neceTsen', 'nyecesen', 'neceseZn', 'nxecesen', 'neceken', 'necesZn', 'necesMen', 'neıesen', 'neceisen', 'necRsen', 'necəesen', 'necEesen', 'necesenÜ', 'çnecesen', 'decesen', 'nhecesen', 'nejcesen', 'necAesen', 'necesejn', 'nOecesen', 'necesenx', 'neceseİ', 'neceysen', 'neceswen', 'niecesen', 'necHsen', 'nKecesen', 'şecesen', 'necsesen', 'nqecesen', 'necexsen', 'neesen', 'necesmen', 'nİecesen', 'nencesen', 'neöcesen', 'nƏecesen', 'necvesen', 'Ğecesen', 'neceqen', 'necesNn', 'necesena', 'önecesen', 'neceoen', 'necesşn', 'neXesen', 'neQesen', 'necesVn', 'neşesen', 'necesçen', 'necesfen', 'ncecesen', 'nçcesen', 'nYecesen', 'necedsen', 'necemen', 'neceseun', 'mnecesen', 'neĞcesen', 'wecesen', 'necesöen', 'necesBen', 'neceseJn', 'neeesen', 'necKsen', 'nÇcesen', 'neceıen', 'jnecesen', 'necesenn', 'neIesen', 'necÜsen', 'nzcesen', 'nAecesen', 'neĞesen', 'neceZen', 'Çecesen', 'necesŞn', 'qnecesen', 'necesenw', 'neceĞsen', 'neÇcesen', 'neceHen', 'neceHsen', 'necesOen', 'necesebn', 'neceseÜn', 'neceseH', 'neəcesen', 'necVesen', 'neceAen', 'necesnn', 'necesƏn', 'neceseN', 'anecesen', 'necesea', 'neceŞsen', 'neceDsen', 'neceseön', 'nMecesen', 'neecesen', 'necesoen', 'nehesen', 'Becesen', 'necesDn', 'neceseƏ', 'neçcesen', 'necersen', 'necesĞen', 'necKesen', 'neczsen', 'necMesen', 'nececen', 'neceseğ', 'necesei', 'neFesen', 'necesLen', 'neccsen', 'neÜcesen', 'nEcesen', 'neUesen', 'necesern', 'neceUen', 'jecesen', 'necqsen', 'ndecesen', 'neMcesen', 'neZesen', 'neceseMn', 'neceösen', 'necesRen', 'necesenö', 'necesSen', 'nebesen', 'necfsen', 'necBesen', 'necehsen', 'necesey', 'nfecesen', 'necesyen', 'ndcesen', 'necesen', 'neceeen', 'necehen', 'necesFn', 'necebsen', 'neacesen', 'newesen', 'nececsen', 'necesren', 'tecesen', 'neceEen', 'nepcesen', 'neceİen', 'necesben', 'neceIsen', 'neceseŞ', 'zecesen', 'dnecesen', 'neceseO', 'ynecesen', 'Onecesen', 'nefcesen', 'nezcesen', 'snecesen', 'necasen', 'nÖecesen', 'neceseÇ', 'neceseÇn', 'neceseOn', 'neceesen', 'necpsen', 'necresen', 'neceEsen', 'neceseln', 'nCcesen', 'neceQsen', 'neKesen', 'neceset', 'Mecesen', 'neceseŞn', 'nçecesen', 'Decesen', 'nBcesen', 'neceGsen', 'neceseD', 'neucesen', 'neceSen', 'necesJen', 'nXcesen', 'nekesen', 'nFecesen', 'nectesen', 'necgsen', 'ngecesen', 'necesun', 'neceasen', 'necetsen', 'necesenL', 'neğcesen', 'knecesen', 'neceben', 'nJcesen', 'nocesen', 'neceten', 'necesqen', 'necesenz', 'necesşen', 'üecesen', 'Aecesen', 'nmecesen', 'neYcesen', 'neceses', 'nöcesen', 'nTecesen', 'neceısen', 'neceÇen', 'necoesen', 'neSesen', 'necesdn', 'neceseQn', 'neEesen', 'neceseİn', 'necMsen', 'necaesen', 'necUesen', 'neScesen', 'necesepn', 'neceseJ', 'necesKn', 'wnecesen', 'neceƏsen', 'necesAn', 'necesenı', 'necesYen', 'nşecesen', 'necnsen', 'neDesen', 'necBsen', 'nUcesen', 'necJsen', 'neGesen', 'necOsen', 'neceden', 'nGecesen', 'necepsen', 'neceseNn', 'ənecesen', 'necesex', 'neceĞen', 'neNesen', 'necwsen', 'necesenu', 'necesRn', 'neceseK', 'nsecesen', 'neEcesen', 'neceVen', 'neceseö', 'necNsen', 'nlecesen', 'neceszen', 'neŞesen', 'necesən', 'necesenB', 'nşcesen', 'necesenA', 'İecesen', 'ncesen', 'neceseE', 'neÖesen', 'neceüsen', 'necesTn', 'necPsen', 'necöesen', 'necesne', 'nQcesen', 'Jecesen', 'nUecesen', 'necÇesen', 'neceXen', 'neYesen', 'neaesen', 'Ynecesen', 'ınecesen', 'neceseCn', 'necesfn', 'Vecesen', 'pnecesen', 'encesen', 'neceLen', 'necesDen', 'necesuen', 'Knecesen', 'nNecesen', 'neceÇsen', 'necesetn', 'necesenü', 'necesenk', 'nğcesen', 'aecesen', 'neüesen', 'neceKen', 'nehcesen', 'necespen', 'nBecesen', 'neceseR', 'neceğen', 'necSsen', 'necısen', 'necDesen', 'necesefn', 'Pecesen', 'nechsen', 'necİsen', 'necüsen', 'nechesen', 'neclesen', 'necyesen', 'neİesen', 'necesĞn', 'nePcesen', 'neceCen', 'necXsen', 'neıcesen', 'necQsen', 'ıecesen', 'necİesen', 'secesen', 'neceseğn', 'neVesen', 'pecesen', 'neceğsen', 'neBesen', 'necesİen', 'necesene', 'neceYen', 'necmsen', 'neceJen', 'neceseHn', 'necfesen', 'Cecesen', 'necgesen', 'nScesen', 'neceçen', 'njecesen', 'Əecesen', 'neceseLn', 'necbsen', 'neceseı', 'necesenR', 'neceseRn', 'neceseY', 'nĞecesen', 'Ğnecesen', 'necÇsen', 'Önecesen', 'yecesen', 'necesenF', 'necŞesen', 'lecesen', 'neÖcesen', 'neHcesen', 'nSecesen', 'necesUn', 'neVcesen', 'neceshn', 'necesJn', 'necesln', 'necesgen', 'neqesen', 'necksen', 'neceseü', 'nDecesen', 'necssen', 'necesIen', 'nwecesen', 'neceseon', 'nLecesen', 'necesenc', 'neceson', 'necesrn', 'necesenS', 'necesean', 'necesəen', 'necesenĞ', 'Gnecesen', 'necesıen', 'neceien', 'necResen', 'neÜesen', 'neceseng', 'necLsen', 'necesesn', 'Hecesen', 'neceseGn', 'neTcesen', 'nÖcesen', 'necesenJ', 'nDcesen', 'neceslen', 'nemesen', 'Pnecesen', 'nÇecesen', 'nscesen', 'ecesen', 'necesGn', 'neceseÖ', 'noecesen', 'nePesen', 'nMcesen', 'naecesen', 'necNesen', 'neceseĞn', 'njcesen', 'necıesen', 'necçsen', 'öecesen', 'necesÇn', 'nejesen', 'necejsen', 'necesmn', 'fecesen', 'neUcesen', 'necesvn', 'xecesen', 'neğesen', 'neiesen', 'neceseu', 'necesİn', 'neAcesen', 'necesek', 'nacesen', 'necesYn', 'neceseP', 'neceseZ', 'gnecesen', 'Kecesen', 'neceseĞ', 'necesBn', 'neceşsen', 'necesenV', 'nOcesen', 'necesgn', 'nkcesen', 'neceseX', 'neJcesen', 'nzecesen', 'necesenğ', 'Jnecesen', 'nRcesen', 'necesee', 'nGcesen', 'nlcesen', 'neRcesen', 'necesenş', 'neocesen', 'netesen', 'fnecesen', 'necesenP', 'necesenD', 'neceseT', 'necesew', 'neccesen', 'Dnecesen', 'necƏsen', 'necşsen', 'necesNen', 'neceseən', 'necesÖn', 'nelesen', 'kecesen', 'rnecesen', 'necesend', 'necesTen', 'necxesen', 'Oecesen', 'necesün', 'necÖsen', 'neczesen', 'necesSn', 'necesef', 'neŞcesen', 'neceuen', 'neöesen', 'nesesen', 'neceswn', 'neceÜsen', 'necesten', 'neceseI', 'nedcesen', 'necosen', 'neceseDn', 'negcesen', 'nescesen', 'necesan', 'necseen', 'nedesen', 'necĞsen', 'necesın', 'vecesen', 'necesecn', 'neceseVn', 'Ənecesen', 'neceseUn', 'nhcesen', 'neceseS', 'nIecesen', 'necLesen', 'necÖesen', 'necesenU', 'neceÖen', 'necesed', 'necesüen', 'nTcesen', 'necpesen', 'neceseg', 'necxsen', 'necesqn', 'xnecesen', 'necese', 'necesenŞ', 'neAesen', 'necXesen', 'necesnen', 'nmcesen', 'necdesen', 'mecesen', 'neceseFn', 'Zecesen', 'neçesen', 'neəesen', 'necqesen', 'neceseF', 'necCesen', 'neceseın', 'neceseün', 'neceseEn', 'neceMsen', 'necJesen', 'neceVsen', 'necusen', 'neyesen', 'neceseqn', 'neceseny', 'necesenç', 'nbcesen', 'necescn', 'Enecesen', 'necegen', 'necesekn', 'Uecesen', 'nenesen', 'necjsen', 'nücesen', 'neceren', 'necIesen', 'neceseTn', 'neceseAn', 'nNcesen', 'Lecesen', 'necescen', 'necrsen', 'necGesen', 'necesenb', 'neceəsen', 'necŞsen', 'neecsen', 'Hnecesen', 'neceKsen', 'neceseo', 'Xecesen', 'neceBsen', 'necesjen', 'nexcesen', 'necesğn', 'neceseBn', 'neceOen', 'Ünecesen', 'neceQen', 'necesKen', 'neceseXn', 'nƏcesen', 'necesLn', 'Rnecesen', 'necevsen', 'necesec', 'necesenm', 'neceəen', 'Tnecesen', 'necesQn', 'uecesen', 'necensen', 'neqcesen', 'çecesen', 'necesenj', 'neIcesen', 'neceszn', 'neceLsen', 'recesen', 'neceseA', 'nəecesen', 'nncesen', 'necVsen', 'Nnecesen', 'nHcesen', 'nKcesen', 'ngcesen', 'nıcesen', 'necesbn', 'necesien', 'nvecesen', 'necesjn', 'nŞcesen', 'necFsen', 'necçesen', 'Öecesen', 'nefesen', 'necesn', 'necesÖen', 'neceJsen', 'necesemn', 'Lnecesen', 'Necesen', 'neceseB', 'necesedn', 'neceser', 'necQesen', 'neicesen', 'necwesen', 'nceesen', 'necefsen', 'neCcesen', 'neceseçn', 'necesPen', 'Fecesen', 'necePsen', 'nexesen', 'necmesen', 'necesem', 'necesÜn', 'neceseKn', 'necTsen', 'nÜecesen', 'Bnecesen', 'nZecesen', 'nöecesen', 'nıecesen', 'necesGen', 'cnecesen', 'nğecesen', 'necessen', 'onecesen', 'necPesen', 'Çnecesen', 'nepesen', 'necesenG', 'neceGen', 'Eecesen', 'necessn', 'neceşen', 'neceseQ', 'necesenZ', 'necesVen', 'necesenh', 'necePen', 'necĞesen', 'neresen', 'neuesen', 'necesin', 'neckesen', 'ünecesen', 'neceshen', 'neLesen', 'neücesen', 'necesevn', 'necesken', 'ntcesen', 'nbecesen', 'necesŞen', 'Xnecesen', 'necesven', 'neceÜen', 'necesFen', 'nĞcesen', 'netcesen', 'neceçsen', 'necesHen', 'neceseÜ', 'necenen', 'nectsen', 'neceseV', 'oecesen', 'necSesen', 'neceseş', 'neceRsen', 'neceYsen', 'gecesen', 'neceZsen', 'necesön', 'nCecesen', 'necesens', 'neceseşn', 'neceseni', 'neceƏen', 'necexen', 'neceÖsen', 'Vnecesen', 'necesent', 'becesen', 'ntecesen', 'qecesen', 'necuesen', 'necelsen', 'Cnecesen', 'neKcesen', 'neBcesen', 'necesewn', 'neclsen', 'necesep', 'necesHn', 'necÜesen', 'necesenÇ', 'nFcesen', 'necesenC', 'nekcesen', 'nLcesen', 'neHesen', 'neceseə', 'Inecesen', 'inecesen', 'neceaen', 'Qecesen', 'neceseG', 'hecesen', 'Secesen', 'necHesen', 'Znecesen', 'necesez', 'neceFsen', 'Anecesen', 'neceseM', 'neZcesen', 'nİcesen', 'necemsen', 'neceüen', 'necşesen', 'neceseno', 'neƏesen', 'əecesen', 'neceseq', 'necesÇen', 'neceseyn', 'nelcesen', 'neceseƏn', 'necesenM', 'neTesen', 'necysen', 'newcesen', 'nvcesen', 'neşcesen', 'necesenE', 'neceNen', 'necğesen', 'necesMn', 'necesenK', 'necesenH', 'necesƏen', 'neceXsen', 'neceseU', 'Fnecesen', 'ğecesen', 'necesaen', 'nqcesen', 'necdsen'}\n"
     ]
    }
   ],
   "source": [
    "print(edits1('necesen'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "1a81fab0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "444194\n"
     ]
    }
   ],
   "source": [
    "print(len(edits2('sagolun')))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "16af699a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<map at 0x21d00bae088>"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "def case_of(text):\n",
    "    return (str.upper if text.isupper() else\n",
    "            str.lower if text.islower() else\n",
    "            str.title if text.istitle() else\n",
    "            str)\n",
    "map(case_of, ['UPPER', 'lower', 'Title', 'CamelCase'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "id": "87cd5606",
   "metadata": {},
   "outputs": [],
   "source": [
    "def correct_text(text):\n",
    "    return re.sub('[ABCÇDEƏFGĞHXIİJKQLMNOÖPRSŞTUÜVYZabcçdeəfgğhxıijkqlmnoöprsştuüvyz]+', correct_match, text)\n",
    "\n",
    "def correct_match(match):\n",
    "    word = match.group()\n",
    "\n",
    "    result = case_of(word)(correct(word.lower()))\n",
    "    return case_of(word)(correct(word.lower()))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "4a614333",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'salam həkim mənim sağ böyrək ağrıyır'"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "correct_text('salam hekim menim sag boyrek ağryyir')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'necəsən qardaş'"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "correct_text('necesen qardaw')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'metro haçan boş olur? son zamanlar biraz vaxt az olur '"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "correct_text('merto hacan bow olur? son zamanlr birax vxt axz olur ')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Building bigrams"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 296,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"allbook.txt\", \"r\", encoding='utf-8') as book:\n",
    "    lines = book.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 297,
   "metadata": {},
   "outputs": [],
   "source": [
    "def clean_list(element):\n",
    "    regex = re.compile('[^ABCÇDEƏFGĞHXIİJKQLMNOÖPRSŞTUÜVYZabcçdeəfgğhxıijkqlmnoöprsştuüvyz+]')\n",
    "    temp = regex.sub(' ', element.lower())\n",
    "    return \" \".join(temp.split())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "metadata": {},
   "outputs": [],
   "source": [
    "def delete_element(element):\n",
    "    return ' '.join([i for i in element.split() if len(i)>1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {},
   "outputs": [],
   "source": [
    "output_list = list(map(clean_list, lines))\n",
    "output_list = list(map(delete_element, output_list))\n",
    "output_list = [i for i in output_list if len(i)>1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 300,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['mehmanxana onların hər birində orta hesabla yer var',\n",
       "       'kondensasiyası ilə əlaqədar olaraq nüvə membranının onlarla əlaqəsi',\n",
       "       'var imiş da nə bilim gündə orda burda qızlarla kefə',\n",
       "       'istəsə də bacarmırdı sanki bütün xəyal arzu və istəkləri bir andaca',\n",
       "       'azərbaycanlını cəzalandıracağını söylədi ancaq xəfiyyə',\n",
       "       'mədəniyyət abidələrinin qorunub mühafizə olunmasını daha',\n",
       "       'adqeziya', 'cinsiyyəti idarə edən xüsusi genlər də mövcuddur',\n",
       "       'və yenidən düşünməyi tələb etmişdi amma heç bir faydası yox idi',\n",
       "       'hər nə qədər ümidsizliyə qapılmamaq üçün özümü'], dtype='<U451')"
      ]
     },
     "execution_count": 300,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random.choice(output_list,10)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "metadata": {},
   "outputs": [],
   "source": [
    "bigrams = [b for l in output_list for b in zip(l.split(\" \")[:-1], l.split(\" \")[1:])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('rus', 'və'),\n",
       " ('digər', 'dillərdən'),\n",
       " ('dillərdən', 'rus'),\n",
       " ('rus', 'dilinə'),\n",
       " ('dilinə', 'tərcümə'),\n",
       " ('tərcümə', 'olunmuş'),\n",
       " ('olunmuş', 'ədəbiyyatlardan'),\n",
       " ('ədəbiyyatlardan', 'istifadə'),\n",
       " ('istifadə', 'olunub'),\n",
       " ('olunub', 'ngilis'),\n",
       " ('ngilis', 'dilində'),\n",
       " ('dilində', 'çap'),\n",
       " ('çap', 'olunmuş'),\n",
       " ('olunmuş', 'ən'),\n",
       " ('ən', 'son')]"
      ]
     },
     "execution_count": 302,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bigrams[150:165]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "metadata": {},
   "outputs": [],
   "source": [
    "f = open('bigram_all_books.txt', 'w',encoding='utf-8')\n",
    "\n",
    "for t in bigrams:\n",
    "  f.write(' '.join(str(s) for s in t) + '\\n')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}