Group
Extension

zhangbo-NLP-plugin_eng/lib/zhangbo/NLP/engine_rea_zhonto.pl

#!/usr/bin/perl
#use JSON::XS;
 use Clone qw(clone);
use MongoDB;
use Smart::Comments;
use lib "/home/wuyabo/shell/";
use Conn_mongo_jc;
use Data::Dumper;
use IO::File;
#use Add_info;

open(FDst,"stop.txt");
our $hash_st;
while(<FDst>)
{
	my $line;
	$line=$_;
	chomp($line);	
		$hash_st->{$line}=1;

}

# http://poe.perl.org/?POE_Cookbook/TCP_Servers


# Include POE and POE::Component::Server::TCP.
 our $hash=();
 $hash_mn=();


my $hash_sc=();
#i#close FD_log;
	my $w1=$ARGV[0];
	my $w2=$ARGV[1];
	my $f1=$ARGV[2];
	my $f2=$ARGV[3];
	my $t=search_article($f1);	
	$res=clone($t);
#		print Data::Dumper->Dump([$res]);

my	$t=search_article($f2);	
	$res2=clone($t);
	my $k,$v;
				my $tv1;
				my $tv2;
		while(($k,$v)=each(%{$res}))

		{
			if(($k eq $w1)||($k eq $w2))
			{next;}
			if(exists($res2->{$k}))	
			{ 
			

				$res3->{$k}=($res2->{$k})+($res->{$k});				
#				print "---same : $k  $res3->{$k}\n ";
			}
		}

my @key =sort {$res3->{$b} <=> $res3->{$a}} keys %{$res3};
my @re;
foreach my $ttk (@key)
{
print "$ttk-$res3->{$ttk}\n";
#system ("grep $ttk $f1");
}
#电影-25.02-
#美国-50.75-
#纳多-21.26-
#print FD_log @res;
#close(FD_log);
#print scalar(@res)," num\n";
	  #$session->yield("shutdown");

#, ClientFlushed => sub {
#           my $data_source = $_[HEAP]{file_handle};
#           my $read_count = sysread($data_source, my $buffer = "", 65536);
#           if ($read_count) {
#             $_[HEAP]{client}->put($buffer);
#           }
#           else {
#		print FD_log "------------------error\n";
#             $_[KERNEL]->yield("shutdown");
#           }
         #}


#$input="坦克,我的希望非常诱人电视里有主持人,那里有大熊猫";
#	$input="网络和网民的意见,现在正行驶在通州回北京的高速公路上,十评论员单仁平的文章,标题是“做大众政治焦点,茅于轼的选择”。这篇文章的核心意见是,茅于轼应该做中国社会团结的促进者,不应该";
#$input="致力于宣传 市场万能 剥削有理 汉奸人性 保钓无用 保粮错误替富人说话的茅于轼,今天下午两点在北京海淀翠宫饭店演讲顽强继续。未知海淀区委书记隋振江,宣传部长陈名杰是否到场。外媒问我是否到场,告曰:先参加央视《苦难辉煌》座谈会,或会晚到一会儿。今天要长见识了";
#	$input2="茅于轼刚刚吃完胡辣汤和烧饼夹猪头肉。现在正行驶在通州回北京的高速公路上,十五分钟以后要开始腾讯微访谈直播,我和方舟子拟就昨天的热点问题,回答网友的提问。敬请各位网友提示一下,昨天有什么热点问题需要谈一谈?";
#$file=$ARGV[0];
sub search_article
{
my $file=$_[0];
chomp($file);
if(! -e "$file")
{return "file is't exist\n"};
my %my_hash;
my %my_hash2;
my %thash;
my $j;

my $rand=rand();
#open(FD_hash,">$file.txt");
#print "openfile $file\n";
system("/home/wuyabo/xs/bin/scws -I -t 500  -E -N -d /home/lzj/shell/dict.xdb -c utf-8   $file -o /tmp/$rand.sctxt2");
my @all2=`cat /tmp/$rand.sctxt2`;
#20. 建中                   nz    13.84(1)
#21. 恒大                   nz    13.41(1)
unlink("/tmp/$rand.sctxt2");
my $hash3;
foreach $j (@all2)
{
#print "$j\n";
my $word;
 my $kv;
my $pos;
        if($j=~/\d(\d)\. (.*?)\s+(\w+?)\s+(\d.*?)\(/)
        {
        $word=$2;
        $kv=$4;
	$pos=$3;
	if(($pos=~/n.*/)||($pos=~/v.*/))
	{
 	 $hash3->{$word}=$kv;
	}
	}

}
	$hash3;
}


#	}
#	);
#$poe_kernel->run();

sub txt2arr
{
my	$input_t=$_[0];
my	$exp_h=$_[1];
my $tmp_hash=();
#		chomp($input_t);
my @list=split(" ",$input_t);
#print "txt2arr input =$input_t\n";
#$rand=rand();
#open(FD_r,">/tmp/$rand");
#print FD_log  "$input_t\n"; 
#close(FD_r);
#$dir="/home/wuyabo/stanford-parser-2012-07-09";
@res=();
#@res=`$dir/parser-10.pl /tmp/$rand 0 50`;
#$part=0;
#unlink("/tmp/$rand");
#print @res;
my $n_a=0;
my $n_b=0;
my $v_a=0;
my $v_b=0;
my $n1;
my $n2;
my $v1;
my $v2;
my $nline;
my $tmp_first=0;
foreach $tmp_res (@list)
{
		if ($tmp_first==0)
		{
			$nline=$tmp_res;	
			$tmp_first++;
			next;
		}
		
		if(exists($hash_st->{$tmp_res}))
		{next;}
			my $l=length($tmp_res);
#		if($l<=3)
#		{
				#if($hash_sc->{$tmp_res}->{'a'} !~/n|v|a/)	
				#{
				#print FD_log "seek $tmp_res\n";
				#next;}
				if($l<=3)
				{next;}
#主题放大
			if(exists($hash_sc->{$tmp_res}))
			{
					if($hash_sc->{$tmp_res}->{'a'}=~/n/)									
					{
						if($hash_sc->{$tmp_res}->{'h'}> $n_a)
						{
							$n_a=$hash_sc->{$tmp_res}->{'h'};
								$n1=$tmp_res;
				#	print FD_log  "放大主题-$tmp_res-","$n_a n\n";
								if($n_a >$n_b)
								{
									my	$swtmpn=$n_b;
									my	$swtmpnw=$n2;
										$n_b=$n_a;$n2=$n1;
										$n_a=$swtmpn;$n1=$swtmpnw;
								}
						}
						
					}
					elsif($hash_sc->{$tmp_res}->{'a'}=~/v/)
					{
						if($hash_sc->{$tmp_res}->{'h'}> $v_a)
                                                {
                                                        $v_a=$hash_sc->{$tmp_res}->{'h'};
                                                                $v1=$tmp_res;
					print FD_log  "放大主题-$tmp_res-","$v_a v \n";
                                                                if($v_a >$v_b)
                                                                {
                                                                        my      $swtmpn=$v_b;
                                                                        my      $swtmpnw=$v2;
                                                                                $v_b=$v_a;$v2=$v1;
                                                                                $v_a=$swtmpn;$v1=$swtmpnw;
                                                                }
                                                }
					}
			}

			#	}
#		}
	#	print $tmp_res,"---",length($tmp_res),"\n";
		
		#if (($part==1)&&($tmp_res=~/(.*?),(.*)/))
		#{
			#print "切词1=$1,2=$2\n";
			my $w=$tmp_res;
#			my $value=$2;
		#	if(exists($exp_h->{$w}))
		#	{
			#	print "$w\n";
		#	next;}
			if(exists($hash->{$w}))
			{
				my $t_n=0;

				while(($k,$v)=each($hash->{$w}))
				{
					if(exists($hash_st->{$k}))
       				         {next;}

					
						$tmp_hash->{$k}+=0.1;
					#$tmp_hash->{$k}=$v;
					#$tmp_hash->{$k}=$hash_sc->{$w}->{'h'};
                                                #print FD_log  "正常词 $w  转换",$hash_sc->{$w}->{'h'},"--",$tmp_hash->{$k},"   $k\n";
							       #正常词 交  转换0.00--0.00   相互
					#if(exists($sytmp->{$k}))
					#{
					#	print "归一 $k ",$sytmp->{$k},"\n";
				#		$k=$sytmp->{$k};
				#}
#						if($value>20)
		# $heap->{client}->put("$k $v\n");
				}
					#$tmp_hash->{$k}=$v;
			}
			else
			{
				if(exists($hash_sc->{$w}))
				{
                                                $tmp_hash->{$w}+=0.2;
	
				 #$tmp_hash->{$w}=$hash_sc->{$w}->{'h'};
				}
				else
				{

				 	$tmp_hash->{$w}=0.1;
				}
				
				#		print FD_log "no exists wordnet 切词 k=$w,v=",$hash_sc->{$w}->{'h'},"\n";
			}


}
#是否采用scws放大文章中心词
			while(($k,$v)=each(%{$tmp_hash}))
                                {
						if(($k eq $n1)||($k eq $n2)||($k eq $v1)||($k eq $v2))
						{$tmp_hash->{$k}=$v*2;
						}
						if(exists($exp_h->{$k}))
                                        {
                                                 $tmp_hash->{$k}+=0.8;
#					print  "自定义放大主题-$k-",$tmp_hash->{$k}," \n";
                                        }

				}
		#print Data::Dumper->Dump([%tmp_hash]);
	#return (%tmp_hash,$nline);
	return (\%{$tmp_hash},$nline);
}				

#print "output = $res\n";
#		 $heap->{client}->stop();
#	$session->stop();


# Start the server.

sub  comp_hash
{
my	$h1=@_[0];
my	$h2=@_[1];
my	$akv=0;
	my $all;
	my %all;
	my $vall=0.01;
	my $vall2=0.01;
my $k,$v;
	    while(($k,$v)=each(%{$h1}))
                                {
				$all->{$k}++;
				$vall+=$v;
 #                                      print FD_log " hash 1$k  $v\n";
                               }
my $k,$v;
	  while(($k,$v)=each(%{$h2}))
                                {
				$all->{$k}++;
		
#                                      print FD_log " hash 2$k  $v\n";
				$vall2+=$v;
                               }
my $k,$v;
	 while(($k,$v)=each(%{$all}))
                                {
					if($v >= 2)
					{
						$kv=($h1->{$k}+$h2->{$k})/($vall+$vall2);
						$kv2=((($h1->{$k})/$vall) + ($h2->{$k}/$vall2))/2;
						$kv3=($kv+$kv2)/2;
						#if($h1->{$k}<$h2->{$k})
						#{
								
						#}
#						print  FD_log "double $k $v kv $kv kv2 $kv2\n";
						$akv+=$kv3;
					}
                               }
					#	print "---------------\n";


	$akv;

}

sub ndate
{
        ($sec,$min,$hour,$mday,$mon,$year,$wday,$yday,$isdst)=localtime(time());
       $year=~s/1(.*)/20$1/;
       $mon+=1;
       if($mon=~/^\d$/)
       {
               $mon="0"."$mon";
       }
       if($mday=~/^\d$/)
       {
               $mday="0"."$mday";
       }
               if($hour=~/^\d$/)
       {
               $hour="0"."$hour";
       }

               if($min=~/^\d$/)
       {
               $min="0"."$min";
       }
               if($sec=~/^\d$/)
       {
               $sec="0"."$sec";
       }




my       $res=$year."-".$mon."-".$mday." ".$hour.":"."$min".":"."$sec";
#print $res,"\n";
        $res;

}

sub txt2arr_one
{
my        $w=$_[0];
	
my        $exp_h=$_[1];
$tmp_hash=();
$rand=rand();
@res=();
$part=0;
#print @res;

	my @w_arr=split(",",$w);
	
	foreach my $w_tmp (@w_arr)
	{
			if($w_tmp=~/(.*?) (.*)/)
			{
			$w=$1;
			$vh=$2;
			#print "$w -- $vh\n";
                        if(exists($hash->{$w}))
                        {
                                my $t_n=0;

                                while(($k,$v)=each($hash->{$w}))
                                {
					if($v==0)
					{
						$v=1;
					}
                                        $tmp_hash->{$k}=$vh;
                                }
                        }
				$tmp_hash->{$w}=$w;

			}
			else
			{
				print "input error\n";
			}
	}
                 #       while(($k,$v)=each(%{$tmp_hash}))
#
 #                               {
                                        #print "meaning  $k  $v\n";
  #                              }
#	}
        return (%{$tmp_hash});
}

sub search_article2
{
$file=$_[0];
chomp($file);
#print "file=$file\n";
$expword=$ARGV[1];
$exp_hash=();
@exp_w=split(",",$expword);
foreach $exp (@exp_w)
{
$exp_hash->{$exp}=1;
}
my $f=0;
my $n=0;
my $kk;
my $thash=();
my @input_ar=`cat $file`;
my %my_hash;
my %my_hash2;
my %thash;
my $j;
open(FD_arr,"<$file");
my $first=0;
my $rand=rand();
open(FD_aro,">/tmp/$rand.sctxt");
my $j=1;
while(<FD_arr>)
{
        my $in=$_;
        chomp($in);
                 print FD_aro "$j ",$in,"\n";
        $j++;
}
close FD_arr;
close FD_aro;
system("/home/wuyabo/xs/bin/scws -I  -E -N -d /home/wuyabo/xs/etc/dict.utf8.xdb:/home/wuyabo/stanford/dict_user.txt -c utf-8   /tmp/$rand.sctxt -o /tmp/$rand.sctxt2");
unlink("/tmp/$rand.sctxt");
#print "/tmp/$rand.sctxt\n";
open(FD_scws,"/tmp/$rand.sctxt2");
my $hashleft;
my $alln=1;
while(<FD_scws>)
{
        my $in=$_;
#print FD_log "line ttt $kk $in \n";
        chomp($in);
        #$input_ar[$f]=$in;

                my $n_line;
                my ($my_hash2,$n_line)=txt2arr($in,$exp_hash);
                $hashleft[$n_line]=clone($my_hash2);
        $alln++;
}
my $j=0;
my $cluster,$cln;
             for( 1 .. $alln-1)
                {
                        my $nn=$_;
                        for ($nn .. $alln-1)
                        {
                                $nn2=$_;
                                if($nn==$nn2)
                                {next;}
                                $n=comp_hash($hashleft[$nn],$hashleft[$nn2]);
                                $n=$n * 100;
                             #   print "$n $nn $nn2 \n";
                                $cluster->{"$nn-$nn2"}=$n;
                                $cln[$j]="$nn-$nn2";$j++;
#                                print  FD_log "$n $nn $nn2 \n";

                        }

                }
my @m_res;
my @key =sort {$cluster->{$b} <=> $cluster->{$a}} keys %{$cluster};
        for(0 .. (scalar(@key))-1)
        {
                my $itmp=$_;
#                print FD_log $key[$itmp],"-",$cluster->{$key[$itmp]},"\n";
                push @m_res,"$key[$itmp]-$cluster->{$key[$itmp]}\n";
        }
unlink("/tmp/$rand.sctxt2");
                close(FD_arr);
        @m_res;
}



Powered by Groonga
Maintained by Kenichi Ishigaki <ishigaki@cpan.org>. If you find anything, submit it on GitHub.