当你使用boost::spirit这样漂亮的东西时，你应该感到自豪

这里我的一个解析器的尝试(几乎)符合这个链接的CSV规范(我不需要在字段中换行)。逗号周围的空格也被省略了)。

在你克服了编译这段代码需要等待10秒的令人震惊的经历之后:)，你就可以坐下来享受了。

// csvparser.cpp
#include <boost/spirit/include/qi.hpp>
#include <boost/spirit/include/phoenix_operator.hpp>

#include <iostream>
#include <string>

namespace qi = boost::spirit::qi;
namespace bascii = boost::spirit::ascii;

template <typename Iterator>
struct csv_parser : qi::grammar<Iterator, std::vector<std::string>(), 
    bascii::space_type>
{
    qi::rule<Iterator, char()                                           > COMMA;
    qi::rule<Iterator, char()                                           > DDQUOTE;
    qi::rule<Iterator, std::string(),               bascii::space_type  > non_escaped;
    qi::rule<Iterator, std::string(),               bascii::space_type  > escaped;
    qi::rule<Iterator, std::string(),               bascii::space_type  > field;
    qi::rule<Iterator, std::vector<std::string>(),  bascii::space_type  > start;

    csv_parser() : csv_parser::base_type(start)
    {
        using namespace qi;
        using qi::lit;
        using qi::lexeme;
        using bascii::char_;

        start       = field % ',';
        field       = escaped | non_escaped;
        escaped     = lexeme['"' >> *( char_ -(char_('"') | ',') | COMMA | DDQUOTE)  >> '"'];
        non_escaped = lexeme[       *( char_ -(char_('"') | ',')                  )        ];
        DDQUOTE     = lit("\"\"")       [_val = '"'];
        COMMA       = lit(",")          [_val = ','];
    }

};

int main()
{
    std::cout << "Enter CSV lines [empty] to quit\n";

    using bascii::space;
    typedef std::string::const_iterator iterator_type;
    typedef csv_parser<iterator_type> csv_parser;

    csv_parser grammar;
    std::string str;
    int fid;
    while (getline(std::cin, str))
    {
        fid = 0;

        if (str.empty())
            break;

        std::vector<std::string> csv;
        std::string::const_iterator it_beg = str.begin();
        std::string::const_iterator it_end = str.end();
        bool r = phrase_parse(it_beg, it_end, grammar, space, csv);

        if (r && it_beg == it_end)
        {
            std::cout << "Parsing succeeded\n";
            for (auto& field: csv)
            {
                std::cout << "field " << ++fid << ": " << field << std::endl;
            }
        }
        else
        {
            std::cout << "Parsing failed\n";
        }
    }

    return 0;
}

编译:

make csvparser

测试(例子摘自维基百科):

./csvparser
Enter CSV lines [empty] to quit

1999,Chevy,"Venture ""Extended Edition, Very Large""",,5000.00
Parsing succeeded
field 1: 1999
field 2: Chevy
field 3: Venture "Extended Edition, Very Large"
field 4: 
field 5: 5000.00

1999,Chevy,"Venture ""Extended Edition, Very Large""",,5000.00"
Parsing failed

如果你不关心转义逗号和换行符， 并且你不能在引号中嵌入逗号和换行符(如果你不能转义那么…) 那么它只有大约三行代码(好的14 ->，但它只有15读取整个文件)。

std::vector<std::string> getNextLineAndSplitIntoTokens(std::istream& str)
{
    std::vector<std::string>   result;
    std::string                line;
    std::getline(str,line);

    std::stringstream          lineStream(line);
    std::string                cell;

    while(std::getline(lineStream,cell, ','))
    {
        result.push_back(cell);
    }
    // This checks for a trailing comma with no data after it.
    if (!lineStream && cell.empty())
    {
        // If there was a trailing comma then add an empty element.
        result.push_back("");
    }
    return result;
}

我只需要创建一个表示一行的类。 然后流到该对象:

#include <iterator>
#include <iostream>
#include <fstream>
#include <sstream>
#include <vector>
#include <string>

class CSVRow
{
    public:
        std::string_view operator[](std::size_t index) const
        {
            return std::string_view(&m_line[m_data[index] + 1], m_data[index + 1] -  (m_data[index] + 1));
        }
        std::size_t size() const
        {
            return m_data.size() - 1;
        }
        void readNextRow(std::istream& str)
        {
            std::getline(str, m_line);

            m_data.clear();
            m_data.emplace_back(-1);
            std::string::size_type pos = 0;
            while((pos = m_line.find(',', pos)) != std::string::npos)
            {
                m_data.emplace_back(pos);
                ++pos;
            }
            // This checks for a trailing comma with no data after it.
            pos   = m_line.size();
            m_data.emplace_back(pos);
        }
    private:
        std::string         m_line;
        std::vector<int>    m_data;
};

std::istream& operator>>(std::istream& str, CSVRow& data)
{
    data.readNextRow(str);
    return str;
}   
int main()
{
    std::ifstream       file("plop.csv");

    CSVRow              row;
    while(file >> row)
    {
        std::cout << "4th Element(" << row[3] << ")\n";
    }
}

但只要做一点工作，我们就可以在技术上创建一个迭代器:

class CSVIterator
{   
    public:
        typedef std::input_iterator_tag     iterator_category;
        typedef CSVRow                      value_type;
        typedef std::size_t                 difference_type;
        typedef CSVRow*                     pointer;
        typedef CSVRow&                     reference;

        CSVIterator(std::istream& str)  :m_str(str.good()?&str:nullptr) { ++(*this); }
        CSVIterator()                   :m_str(nullptr) {}

        // Pre Increment
        CSVIterator& operator++()               {if (m_str) { if (!((*m_str) >> m_row)){m_str = nullptr;}}return *this;}
        // Post increment
        CSVIterator operator++(int)             {CSVIterator    tmp(*this);++(*this);return tmp;}
        CSVRow const& operator*()   const       {return m_row;}
        CSVRow const* operator->()  const       {return &m_row;}

        bool operator==(CSVIterator const& rhs) {return ((this == &rhs) || ((this->m_str == nullptr) && (rhs.m_str == nullptr)));}
        bool operator!=(CSVIterator const& rhs) {return !((*this) == rhs);}
    private:
        std::istream*       m_str;
        CSVRow              m_row;
};


int main()
{
    std::ifstream       file("plop.csv");

    for(CSVIterator loop(file); loop != CSVIterator(); ++loop)
    {
        std::cout << "4th Element(" << (*loop)[3] << ")\n";
    }
}

现在我们已经到了2020年，让我们添加一个CSVRange对象:

class CSVRange
{
    std::istream&   stream;
    public:
        CSVRange(std::istream& str)
            : stream(str)
        {}
        CSVIterator begin() const {return CSVIterator{stream};}
        CSVIterator end()   const {return CSVIterator{};}
};

int main()
{
    std::ifstream       file("plop.csv");

    for(auto& row: CSVRange(file))
    {
        std::cout << "4th Element(" << row[3] << ")\n";
    }
}

如果你确实关心正确解析CSV，这将做它…相对较慢，因为它一次只处理一个字符。

 void ParseCSV(const string& csvSource, vector<vector<string> >& lines)
    {
       bool inQuote(false);
       bool newLine(false);
       string field;
       lines.clear();
       vector<string> line;

       string::const_iterator aChar = csvSource.begin();
       while (aChar != csvSource.end())
       {
          switch (*aChar)
          {
          case '"':
             newLine = false;
             inQuote = !inQuote;
             break;

          case ',':
             newLine = false;
             if (inQuote == true)
             {
                field += *aChar;
             }
             else
             {
                line.push_back(field);
                field.clear();
             }
             break;

          case '\n':
          case '\r':
             if (inQuote == true)
             {
                field += *aChar;
             }
             else
             {
                if (newLine == false)
                {
                   line.push_back(field);
                   lines.push_back(line);
                   field.clear();
                   line.clear();
                   newLine = true;
                }
             }
             break;

          default:
             newLine = false;
             field.push_back(*aChar);
             break;
          }

          aChar++;
       }

       if (field.size())
          line.push_back(field);

       if (line.size())
          lines.push_back(line);
    }

我写了一个很好的解析CSV文件的方法，我认为我应该把它作为一个答案:

#include <algorithm>
#include <fstream>
#include <iostream>
#include <stdlib.h>
#include <stdio.h>

struct CSVDict
{
  std::vector< std::string > inputImages;
  std::vector< double > inputLabels;
};

/**
\brief Splits the string

\param str String to split
\param delim Delimiter on the basis of which splitting is to be done
\return results Output in the form of vector of strings
*/
std::vector<std::string> stringSplit( const std::string &str, const std::string &delim )
{
  std::vector<std::string> results;

  for (size_t i = 0; i < str.length(); i++)
  {
    std::string tempString = "";
    while ((str[i] != *delim.c_str()) && (i < str.length()))
    {
      tempString += str[i];
      i++;
    }
    results.push_back(tempString);
  }

  return results;
}

/**
\brief Parse the supplied CSV File and obtain Row and Column information. 

Assumptions:
1. Header information is in first row
2. Delimiters are only used to differentiate cell members

\param csvFileName The full path of the file to parse
\param inputColumns The string of input columns which contain the data to be used for further processing
\param inputLabels The string of input labels based on which further processing is to be done
\param delim The delimiters used in inputColumns and inputLabels
\return Vector of Vector of strings: Collection of rows and columns
*/
std::vector< CSVDict > parseCSVFile( const std::string &csvFileName, const std::string &inputColumns, const std::string &inputLabels, const std::string &delim )
{
  std::vector< CSVDict > return_CSVDict;
  std::vector< std::string > inputColumnsVec = stringSplit(inputColumns, delim), inputLabelsVec = stringSplit(inputLabels, delim);
  std::vector< std::vector< std::string > > returnVector;
  std::ifstream inFile(csvFileName.c_str());
  int row = 0;
  std::vector< size_t > inputColumnIndeces, inputLabelIndeces;
  for (std::string line; std::getline(inFile, line, '\n');)
  {
    CSVDict tempDict;
    std::vector< std::string > rowVec;
    line.erase(std::remove(line.begin(), line.end(), '"'), line.end());
    rowVec = stringSplit(line, delim);

    // for the first row, record the indeces of the inputColumns and inputLabels
    if (row == 0)
    {
      for (size_t i = 0; i < rowVec.size(); i++)
      {
        for (size_t j = 0; j < inputColumnsVec.size(); j++)
        {
          if (rowVec[i] == inputColumnsVec[j])
          {
            inputColumnIndeces.push_back(i);
          }
        }
        for (size_t j = 0; j < inputLabelsVec.size(); j++)
        {
          if (rowVec[i] == inputLabelsVec[j])
          {
            inputLabelIndeces.push_back(i);
          }
        }
      }
    }
    else
    {
      for (size_t i = 0; i < inputColumnIndeces.size(); i++)
      {
        tempDict.inputImages.push_back(rowVec[inputColumnIndeces[i]]);
      }
      for (size_t i = 0; i < inputLabelIndeces.size(); i++)
      {
        double test = std::atof(rowVec[inputLabelIndeces[i]].c_str());
        tempDict.inputLabels.push_back(std::atof(rowVec[inputLabelIndeces[i]].c_str()));
      }
      return_CSVDict.push_back(tempDict);
    }
    row++;
  }

  return return_CSVDict;
}

当对CSV文件使用Boost Tokenizer escaped_list_separator时，应该注意以下几点:

它需要一个转义字符(默认的反斜杠- \) 它需要一个分割符/分隔符-字符(默认逗号-，) 它需要一个引号字符(默认的引号- ")

wiki指定的CSV格式规定数据字段可以包含引号分隔符(支持):

1997年，福特E350，“超级豪华卡车”

由wiki指定的CSV格式规定单引号应该用双引号处理(escaped_list_separator将剥离所有引号字符):

1997年，福特E350，“超级”“豪华”“卡车”

CSV格式没有指定应该删除任何反斜杠字符(escaped_list_separator将删除所有转义字符)。

修复boost escaped_list_separator的默认行为的一个可能的变通方法:

首先将所有反斜杠字符(\)替换为两个反斜杠字符(\\)，这样它们就不会被剥离。 其次，将所有双引号("")替换为一个反斜杠字符和一个引号(\")

这种变通方法有一个副作用，即由双引号表示的空数据字段将被转换为单引号标记。在遍历令牌时，必须检查令牌是否是单引号，并将其视为空字符串。

不漂亮，但它工作，只要在引号中没有换行。

@sastanin的解决方案的一个小版本，以便它可以处理引号中的换行。

std::vector<std::vector<std::string>> readCSV(std::istream &in) {
    std::vector<std::vector<std::string>> table;

    while (!in.eof()) {
        CSVState state = CSVState::UnquotedField;
        std::vector<std::string> fields {""};
        size_t i = 0; // index of the current field
        for (char c : row) {
            switch (state) {
                case CSVState::UnquotedField:
                    switch (c) {
                        case ',': // end of field
                                  fields.push_back(""); i++;
                                  break;
                        case '"': state = CSVState::QuotedField;
                                  break;
                        default:  fields[i].push_back(c);
                                  break; }
                    break;
                case CSVState::QuotedField:
                    switch (c) {
                        case '"': state = CSVState::QuotedQuote;
                                  break;
                        default:  fields[i].push_back(c);
                                  break; }
                    break;
                case CSVState::QuotedQuote:
                    switch (c) {
                        case ',': // , after closing quote
                                  fields.push_back(""); i++;
                                  state = CSVState::UnquotedField;
                                  break;
                        case '"': // "" -> "
                                  fields[i].push_back('"');
                                  state = CSVState::QuotedField;
                                  break;
                        case '\n': // newline
                                  table.push_back(fields);
                                  state = CSVState::UnquotedField;
                                  fields = vector<string>{""};
                                  i = 0;
                        default:  // end of quote
                                  state = CSVState::UnquotedField;
                                  break; }
                    break;
            }
        }
    }
    return table;
}