Read Text file using Boost mmap

问题

I am reading the following file:

File.txt

Y:\Test\DOCUMENTS\DOCUMENTS\Flux Assurance 2\multi\ACTEPROC_OK\018-1-R.xml Y:\Test\DOCUMENTS\DOCUMENTS\Flux Assurance 2\multi\ACTEPROC_OK\A018-2-R.xml Y:\Test\DOCUMENTS\DOCUMENTS\Flux Assurance 2\multi\ACTEPROC_OK\021-1-R.xml Y:\Test\DOCUMENTS\DOCUMENTS\Flux Assurance 2\multi\ACTEPROC_OK\A021-2-R.xml Y:\Test\DOCUMENTS\DOCUMENTS\Flux Assurance 2\multi\ACTEPROC_OK\022-1-R.xml Y:\Test\DOCUMENTS\DOCUMENTS\Flux Assurance 2\multi\ACTEPROC_OK\022-2-R.xml Y:\Test\DOCUMENTS\DOCUMENTS\Flux Assurance 2\multi\ACTEPROC_OK\025-1-R.xml

Source code:

#include <iostream>
#include <vector>
#include <boost/iostreams/device/mapped_file.hpp>
#include <boost/iostreams/stream.hpp>

std::vector<string> readFile(string);

int main()
{
    std::vector<string> res = readFile("file.txt");
    return 0;
}

std::vector<string> readFile(string f)
{
    boost::iostreams::stream<boost::iostreams::mapped_file_source> str(f);
    std::vector<string> app;
    for(string x; str >> x;)
    {
        app.push_back(x);
    }
    return app;
}//end

Problem:

Since There is a space Between Flux and Assurance 2, it considers space as a separator.

How can remove the separator, or explicitly identify what the separator is?

回答1:

This has nothing to with boost or memorymapping.

Firstly

str.unsetf(std::ios::skipws);

will prevent whitespace from being used as delimiter/skipped.

Secondly, I think (you don't mention any of that) you wanted to read by line:

for(std::string x; std::getline(str, x, '\n');)
{
    app.push_back(x);
}

As you can see you can already specify the delimiter.

Lastly, consider using a parser generator. See e.g. here:

How to parse space-separated floats in C++ quickly?

Which contains an example that uses Boost Spirit to parse from a memory-mapped file.

来源：https://stackoverflow.com/questions/21431518/read-text-file-using-boost-mmap

标签

file

boost

iostream

mmap