PERL homework #1

#!/usr/bin/perl
use locale;
use utf8;
use encoding 'utf8';

####################### config variables
$operation;
$input;
$output;
$minCharCount;
$minFrequency;
####################### procedures and functions
sub readConfig
{
        my $source = shift(@_);
        if(open(SOURCE,$source))
        {
                while ($row = <SOURCE>)
                {
                chomp($row);
                push(@arrayRow,$row);
                }
        $operation              = uc($arrayRow[0]);
        $input                  = $arrayRow[1];
        $output                 = $arrayRow[2];
        $minCharCount           = $arrayRow[3];
        $minFrequency           = $arrayRow[4];
        }
        else
        {
                print "FATAL ERROR: unable to read configuration\n";
                exit();
        }
        if($operation ne "TP")
        {
         if($operation ne "TF")
         {
          if($operation ne "TF_IDF")
           {
                print "FATAL ERROR: unknown operation!\n";
                print "HINT: use tp, tf or tf_idf only.\n";
                exit();
        }}}
}

sub writeConfig()
{
        print "##       Configuration file: $ARGV[0]    ##\n";
        print "Operation:                       $operation\n";
        print "Input file:                      $input\n";
        print "Output file:                     $output\n";
        print "Minimal char count:              $minCharCount\n";
        print "Minimal Frequency:               $minFrequency\n";
        print "###########################################\n";
}

sub getWords
{
                open(DATA, $input);#otevre soubor zadany argumentem
                @radky=<DATA>; #rozdeleno po radcich
        $pocetSouboru=0; #pocitadlo
        for $lala (@radky)              #pro kazde slovo na predhozenem radku
        {
                $lala =~s/^.+\t//;              #odstraneni tridy
                $lala =~s/\n//; #odstraneni konce radku
                $lala =~s/<(.*)>//; #odstraneni tagu
                push(@slova,  split(/\s+/, $lala)); #pridani do vektoru slov
                $pocetSouboru++; #pocet souboru +1
        }
        $pocetSouboru--;#odecteni posledniho prazdneho radku
        $counter=0;
        for $lala (@slova) #prazdnych slov
        {
                $lala=lc($lala);#lower case
                $lala =~s/\s//; #prazdne znaky
                $lala =~s/[^a-zA-Z]//;#jiny znaky nez pismena
                if($lala eq "")
                {
                        splice(@slova, $counter, 1);#odstraneni prazdnych veci
                };
                $counter++;
        }
        $counter=0;

        if($minFrequency > 1) #ma smysl jen pokud minimalni vyskyt je vetsi nez 1
        {
        #print "-> minFrequency greater than 1\n";
                for($counter=0;$counter<=$#slova;$counter++)    #minimalni vyskyt
                {
                        $i=0;
                        #print"-> testuji slovo: ( $slova[$i] ) |>  $slova[$counter]\n";
                        $frequency=0; #nulovy vyskyt
                        for($i=0;$i<=$#slova;$i++)
                        {
                                #print" $slova[$counter] -> $slova[$i] \n";
                                if($slova[$i] eq $slova[$counter])
                                {
                                        $frequency++;
                                        #print"Stejna slova! F: $frequency\n";
                                }
                        }
                        if($frequency < $minFrequency)
                        {
                                #print "odstranuji slovo: $slova[$counter]\n";
                                splice(@slova, $counter, 1);

                        }
                #print"####################\n\n";
                }
        }
        $counter=0;
        for $lala (@slova)#odstraneni duplicit
        {
                for($i=0;$i<$counter;$i++)
                {
                        if($slova[$i] eq $slova[$counter])
                        {
                                splice(@slova, $i, 1);
                                $i=0;#nutny reset
                        }
                }
                $counter++;
        }
        $counter = 0;
        for $lala(@slova)#ostrani slova kratsi nez pozadovane
        {
                if(length($lala) <= $minCharCount)
                {
                        #print "nalezeno slovo: $lala\n";
                        #print length($lala);
                        #print "\n";
                        splice(@slova, $counter, 1);
                }
                $counter++;
        }
return @slova;  #vraci vektor cistych slov
}

sub writeWords
{
        print "\n##     Vectore of clear word   ##\n";
        for $lala (@words)
        {
                $lala=uc($lala);
                $lala =~s/\s//;
                print  $lala;
                print " ";
        }
        print "\n##     End of vector   ##\n\n";
}

sub TP
{
        open(DATA, $input); #otevre soubor zadany argumentem
        @radky=<DATA>; #rozdeleno po radcich

        for $lala (@slova)
        {
                $lala=uc($lala);
                $lala =~s/\s//;
                print  $lala;
                print "   ";
        }
        print "_CLASS_\n";

        for $radek (@radky)
        {
                $class = $radek;
                $radek =~s/\n//;
                $class =~s/\t.*\n//;
                $radek =~s/^.*\t//;
                $radek=uc($radek);
                for $word (@words)
                {
                        $word=uc($word);
                        if(index(" ".$radek." "," ".$word." ") != -1)
                        {
                                print "1";
                                for($i=0;$i<(length($word)+2);$i++)
                                {
                                        print " ";
                                }


                        }
                        else
                        {
                                print "0";
                                for($i=0;$i<(length($word)+2);$i++)
                                {
                                        print " ";
                                }
                        }
                }
                print "$class\n";
        }
}


sub TF
{
        open(DATA, $input); #otevre soubor zadany argumentem
        @radky=<DATA>; #rozdeleno po radcich

        for $lala (@slova)
        {
                $lala=uc($lala);
                $lala =~s/\s//;
                print  $lala;
                print "   ";
        }
        print "_CLASS_\n";
        for $radek (@radky)
        {
                $class = $radek;
                $radek =~s/\n//;
                $class =~s/\t.*\n//;
                $radek =~s/^.*\t//;
                $radek=uc($radek);
                @wordsOnRow = split(' ',$radek);

                for $word (@words)
                {
                        $word=uc($word);
                        $sum=0;
                        for $wordOnRow (@wordsOnRow)
                        {
                                if($wordOnRow eq $word)
                                {
                                        $sum++;
                                }
                        }
                        print "$sum";
                        for($i=0;$i<(length($word)+2);$i++)
                        {
                                print " ";
                        }

                }
                print "$class\n";
        }
}


sub TF_IDF
{
        open(DATA, $input); #otevre soubor zadany argumentem
        @radky=<DATA>; #rozdeleno po radcich

        $docCount = @radky;     # N -> log(N) -> pocet dokumentu
        $docIncCount = 0;       # n(t_i) -> pocet dokumentu obsahujici vyraz
        $countInDoc = 0;        # n_ij -> pocet vyskytu v dokumentu
        $sumInDoc = 0;          # suma n_ij -> pocet termu v danem dokumentu

                ##      TF-IDF = TF * IDF =  n_ij / suma n_ij  *  log N / n(t_i)

        for $lala (@slova)
        {
                $lala=uc($lala);
                $lala =~s/\s//;
                print  $lala;
                print "      ";
        }
        print "_CLASS_\n";

        @pomRows = @radky;
        for $radek (@radky)
        {
                $class = $radek;
                $radek =~s/\n//;
                $class =~s/\t.*\n//;
                $radek =~s/^.*\t//;
                $radek=uc($radek);
                @wordsOnRow = split(' ',$radek);

                for $word (@words)
                {
                        $word=uc($word);
                        $countInDoc=0; #pocet vyhovujicich
                        $sumInDoc=0;   #celkovy pocet
                        $docInCount=0;
                        for $wordOnRow (@wordsOnRow)
                        {
                                for $row (@pomRows)
                                {
                                        $row =~s/\n//;
                                        $row =~s/^.*\t//;
                                        $row=uc($row);
                                        @wordsOnPomRow = split(' ',$row);
                                        for $wordOnPomRow (@wordsOnPomRow)
                                        {
                                                if($wordOnPomRow eq $word)
                                                {
                                                        $docInCount++;
                                                        last;
                                                }
                                        }
                                        }

                                if($wordOnRow eq $word)
                                {
                                        $countInDoc++;
                                }
                                $sumInDoc++;
                        }
                        # TF-IDF = TF * IDF =  n_ij / suma n_ij  *  log N / n(t_i)
                        $TF = $countInDoc / $sumInDoc;
                        if($countInDoc <= 0)
                        {
                                $result=0;
                        }
                        else
                        {
                                $IDF = log($docCount)/$countInDoc;
                                $result = $TF*$IDF;
                        }
                        if($result == 0) {print "$result   ";}
                        else {  printf("%.2f", $result);}
                        for($i=0;$i<(length($word)+2);$i++)
                        {
                                print " ";
                        }

                }
                print "$class\n";
        }
}


####################### main program
readConfig($ARGV[0]);
writeConfig();
@words = getWords();
writeWords();
if($operation eq "TP")
{
        TP();
}
elsif($operation eq "TF")
{
        TF();
}
elsif($operation eq "TF_IDF")
{
        TF_IDF();
}
else
{
        print"FATAL ERROR: unknown operation\n";
}