Perl和C/C++比较——文件读取写入，模式匹配

Perl语言相当的灵活，写起来比较快（我就比较慢了），并且相比之下代码长度也很短。

这是一个曾经写的Perl程序：

#! /usr/bin/perl -w
use strict;

open(F,"/ifs1/PAP/liuyanbo/Cryptorchidism/SNP_result/common_filter") or die"open F error:$!";
open(W,">/ifs1/PAP/liuyanbo/Cryptorchidism/SNP_result/new_filter_again") or die"open W error:$!";

while(my $string=<F>)
{
        chomp $string;
        my @array=split //s+/,$string;
        my $length=scalar@array;
        my $B=0;
        my $F=0;
        for(my $i=3;$i<$length;++$i)
        {
                if($array[$i]=~/B/){$B=1;}
                if($array[$i]=~/F/){$F=1;}
        }
        if($B==1 and $F==1){print W "$string/n";}
}
close F;close W;

昨天试着用C++写这个程序，弄了好长时间才勉强弄出来：

#include<iostream>
#include<string>
#include<boost/regex.hpp>
#include<fstream>
#include<cstdlib>
#include<vector>
using namespace std;

void StringSplit(string s,char splitchar,vector<string>& vec)
{
if(vec.size()>0)//..vec...
   vec.clear();
int length = s.length();
int start=0;
for(int i=0;i<length;i++)
{
   if(s[i] == splitchar && i == 0)//.........
   {
    start += 1;
   }
   else if(s[i] == splitchar)
   {
    vec.push_back(s.substr(start,i - start));
    start = i+1;
   }
   else if(i == length-1)//....
   {
    vec.push_back(s.substr(start,i+1 - start));
   }
}
}

int main()
{
        const char *Reg_B="B//d+";
        boost::regex reg_b(Reg_B);
        const char *Reg_F="F//d+";
        boost::regex reg_f(Reg_F);
        ifstream inFile;
        string filename="/ifs1/PAP/liuyanbo/Cryptorchidism/SNP_result/common_filter";//要读取的文件
        inFile.open(filename.c_str());

        ofstream outFile;
        char* outname="/ifs1/PAP/liuyanbo/Cryptorchidism/SNP_result/test_C/data.txt";//写进的文件
        outFile.open(outname,ostream::app);

        string line;
        while(!inFile.eof())
        {
                getline(inFile,line,'/n');//得到文件的一行
                int b_number=0;
                int f_number=0;
                vector<string> vec;
                StringSplit(line,'/t',vec);//按"/t"分割，赋给vec
                for(int i=0;i<vec.size();i++)
                {
                        bool b=boost::regex_match(vec[i],reg_b);
                        if(b){b_number++;}
                        bool f=boost::regex_match(vec[i],reg_f);
                        if(f){f_number++;}
                 }
                if(b_number>0&&f_number>0){outFile<<line<<endl;}
        }

        return 0;
}

这两个程序肯定还可以优化（水平有限），但这里要看的是Perl风格和C风格的差异。