观察你的数据：
- 每个数据点都以以下模式存储：数据、空间。
- 如果数据点不存在，则用空格表示，除非它是所有其他输出被截断为换行符的最后一个不存在的数据点。
这就是我想出的：
#include <fstream>
#include <iostream>
#include <string>
#include <vector>
#include <cstdlib>
#include <sstream>
#include <iomanip>
#include <cctype>
using namespace std;
//note all the lines are stored WITH newlines at the end of them.
//This is merely an artifact of the methodology I am using,
//as the newline is a flag that truncates output (as per your problem)
vector<string> preparse_input(const std::string& filename) {
vector<string> lines;
ifstream ifile;
ifile.open(filename.c_str(), ios::in);
if (!ifile.is_open()) {
exit(1);
}
string temp, chars, line;
char ch;
while(getline(ifile, temp)) {
temp += "\n";//getline removes the newline: because we need it, reinsert it
istringstream iss(temp);
//first read in the line char by char
while(iss >> noskipws >> ch) {
chars += ch;
}
bool replaced_newline = false;
int nargs = 0;
//I could have used iterators here, but IMO, this way is easier to read. Modify if need be.
for (int i = 0; i < chars.size(); ++i) {
if (isdigit(chars[i]) && chars[i+1] == ' ') {
nargs += 1;
}
else if(isspace(chars[i]) && isspace(chars[i+1])) {
if (chars[i+1] == '\n') {
replaced_newline = true;
}
//this means that there is no value set
//hence, set the value to 0 for the value part:
chars[i+1] = '0';
line += chars[i];
++i;//now, skip to the next character since 1 is for spacing, the other is for the value
nargs += 1;
}
//now rebuild the line:
line += chars[i];
if(isdigit(chars[i]) && chars[i+1] == '\n') {
nargs += 1;
//check nargs:
for (int i = nargs; i < 5; ++i) {
line += " 0";
nargs += 1;
}
}
if (replaced_newline) {
line += '\n';
}
replaced_newline = false;
}
lines.push_back(line);
chars.clear();
line.clear();
}
ifile.close();
return lines;
}
//this way, it's much easier to adapt to any type of input that you may have
template <typename T>
vector< vector<T> > parse_input (const vector<string>& lines) {
vector< vector<T> > values;
T val = 0;
for(vector<string>::const_iterator it = lines.begin(); it != lines.end(); ++it) {
vector<T> line;
istringstream iss(*it);
string temp;
while(getline(iss, temp, ' ')) {
if (istringstream(temp) >> val) {
line.push_back(val);
}
else {
line.push_back(0);//this is the value that badly parsed values will be set to.
//you have the option of setting it to some sentinel value, say -1, so you can go back and correct it later on, if need be. Depending on how you want to treat this error - hard or soft (stop program execution vs adapt and continue parsing), then you can adapt it accordingly
//I opted to treat it as a soft error but without a sentinel value - so I set it to 0 (-1 as that is probably more applicable in a general case), and informed the user that an error occurred
//The flipside of that is that I could have treated this as a hard error and have `exit(2)` (or whatever error code you wish to set).
cerr << "There was a problem storing:\"" << temp << "\"\n";
}
}
values.push_back(line);
}
return values;
}
int main() {
string filename = "data.dat";
vector<string> lines = preparse_input(filename);
vector < vector<double> > values = parse_input<double>(lines);
for (int i = 0; i < values.size(); ++i) {
for (int j = 0; j < values[i].size(); ++j) {
cout << values[i][j] << " ";
}
cout << endl;
}
return 0;
}
总而言之，我通过逐个字符读取每一行来分解字符串，然后通过替换空格来重建每一行，0以便于解析。为什么？因为没有这样的值，就无法判断哪个参数被存储或跳过（使用默认ifstream_object >> type方法）。
这样，如果我然后使用stringstream对象来解析输入，我可以正确地确定设置了哪个参数，或者没有设置；然后，存储结果，一切都很好。这就是你想要的。
并且，在以下数据上使用它：
1.1 1.2 1.3 1.4 1.5
1.6 1.7 1.8 1.9
2.0
2.0
2.1 2.2 2.3 2.4 2.5
2.1 2.4
给你输出：
1.1 1.2 1.3 1.4 1.5
1.6 0 1.7 1.8 1.9
2 0 0 0 0
2 0 0 0 0
2.1 2.2 2.3 2.4 2.5
2.1 0 0 2.4 0
注意：第 3 行有 8 个空格（1 个无数据，1 个空格）。第 4 行是来自原始数据的行。第 6 行包含 5 个空格（遵循引用的模式）。
最后，让我说，这是迄今为止我遇到过的最疯狂的数据存储方法之一。