URL相似度计算的思考

#!/usr/bin/perl
use strict;
use warnings;
 
my $url1 = $ARGV[0];
my $url2 = $ARGV[1];
 
my @array1 = split( /\//, $url1 );
my @array2 = split( /\//, $url2 );
 
 
#特征1：目录数
my $path1 = @array1;
my $path2 = @array2;
 
#print $path1, $path2;
#特征2：是否目录结尾
my $lastispath1 = 0;
my $lastispath2 = 0;
 
if( $url1 =~ /\/$/ )
{
	$lastispath1 = 1;
}
if( $url1 =~ /\/$/ )
{
	$lastispath2 = 1;
}
#特征3：最后一级是否有后缀(htm,html,shtml等)
my $len;
my $hassuffix1 = 0;
my $hassuffix2 = 0;
my $suffixstr;
my $laststr1 = $array1[$path1 - 1];
my $laststr2 = $array2[$path2 - 1];
my $issuffixsame = 0;
 
if( $lastispath1 == 0 )
{
	my @suffix1 = split( /\./, $array1[$path1 - 1]);
	if( @suffix1 >= 2 )
	{
		$len = rindex( $suffix1[@suffix1 - 1]."\$", "\$");
		if( $len <= 5 )
		{
			$hassuffix1 = 1;
			$suffixstr = $suffix1[@suffix1 - 1];
			my $tmplen = rindex( $array1[@array1 - 1]."\$", "\$");
			$laststr1 = substr( $array1[@array1 - 1], 0, $tmplen-$len-1 );
		}
	}
}
if( $lastispath2 == 0 )
{
	my @suffix2 = split( /\./, $array2[$path2 - 1]);
	if( @suffix2 >= 2 )
	{
		$len = rindex( $suffix2[@suffix2 - 1]."\$", "\$");
		if( $len <= 5 )
		{
			$hassuffix2 = 1;
			if($suffixstr eq $suffix2[@suffix2 - 1])
			{
				$issuffixsame = 1;
			}
			my $tmplen = rindex( $array2[@array2 - 1]."\$", "\$");
			$laststr2 = substr( $array2[@array2 - 1], 0, $tmplen-$len-1 );
		}
	}
}
 
#特征3：最后一级几个分隔符(通过特征匹配计算laststr1和laststr2相似度，如果仅计算字符串相似度，可以用编辑长度)
my @area1 = split(/-/, $laststr1);
my @area2 = split(/-/, $laststr2);
my $i;
my $j;
my $totalarea1=0;
my $totalarea2=0;
my @patternarray1={0};
my @patternarray2={0};
my @splitarray1={0};
my @splitarray2={0};
#my $numarea1 = @area2;
 
#print $laststr1," ",$laststr2,"\n",$numarea1,"\n";
for ( $i = 1; $i<=@area1; $i++ )
{
	my @tmp1 = split( /_/, $area1[$i-1]);
 
	for( $j = 0; $j<@tmp1; $j++)
	{
		if( $tmp1[$j] =~ /^\d+$/ )	
		{
			$patternarray1[$totalarea1] = 1; #数字pattern
		}
		elsif( $tmp1[$j] =~ /^[a-zA-Z]+$/)
		{
			$patternarray1[$totalarea1] = 2; #纯字母pattern
		}
		elsif( $tmp1[$j] =~ /^[a-zA-Z]+[0-9]+$/)
		{
			$patternarray1[$totalarea1] = 3; #先字母后数字pattern
		}
		elsif( $tmp1[$j] =~ /^[0-9]+[a-zA-Z]+$/)
		{
			$patternarray1[$totalarea1] = 4; #先数字后字母pattern
		}
		else 
		{
			$patternarray1[$totalarea1] = 5; #其他pattern
		}
 
		if( $j == 0 )
		{
			$splitarray1[$totalarea1]=1;
		}
		else
		{
			$splitarray1[$totalarea1]=2;
		}
		$totalarea1 ++;
 
	}	
}
 
for ( $i = 1; $i<=@area2; $i++ )
{
	my @tmp2 = split( /_/, $area2[$i-1]);
	for( $j = 0; $j<@tmp2; $j++)
	{
		if( $tmp2[$j] =~ /^\d+$/ )	
		{
			$patternarray2[$totalarea2] = 1; #数字pattern
		}
		elsif( $tmp2[$j] =~ /^[a-zA-Z]+$/)
		{
			$patternarray2[$totalarea2] = 2; #纯字母pattern
		}
		elsif( $tmp2[$j] =~ /^[a-zA-Z]+[0-9]+$/)
		{
			$patternarray2[$totalarea2] = 3; #先字母后数字pattern
		}
		elsif( $tmp2[$j] =~ /^[0-9]+[a-zA-Z]+$/)
		{
			$patternarray2[$totalarea2] = 4; #先数字后字母pattern
		}
		else 
		{
			$patternarray2[$totalarea2] = 5; #其他pattern
		}
 
		if( $j == 0 )
		{
			$splitarray2[$totalarea2]=1;
		}
		else
		{
			$splitarray2[$totalarea2]=2;
		}
		$totalarea2 ++;
 
	}	
}
 
print $path1," ",$lastispath1," ",$hassuffix1," ",$issuffixsame," ",$totalarea1;
for( $i = 0; $i<$totalarea1; $i++)
{
	print " ",$splitarray1[$i]," ",$patternarray1[$i];
}
print "\n";
print $path2," ",$lastispath2," ",$hassuffix1," ",$issuffixsame," ",$totalarea2;
for( $i = 0; $i<$totalarea2; $i++)
{
	print " ",$splitarray2[$i]," ",$patternarray2[$i];
}
 
print "\n";
#print @array1;