Extracting specific tokens with stringstream

Question

I am extracting exactly 14 tokens in each line on a csv file, and my problem is that in any token if there are spaces in it, I want to preserve it, but I am using stringstream so when I try and extract the data, I'm excluding spaces apart of the same token and it's not inputting correctly and will mess up all data after that. I need a way to use stringstream but still be able to preserve spaces in the tokens. I am looking to extract exactly the third token(in my case), but it would be nice to know in general how can I for any token

Ex: The csv file would be:

PHIL, 100, Intro to Philosophy, 6, Bobby,... (all on one line)

I have a constructor which takes a reference to stringstream object which is supposed to initialize my data members. When I get to a token which commonly has spaces, it will stop at a space and in this example it would only take, "Intro" and then store that into my classTitle variable, but I want the entire, "Intro to Philosophy". I also have a field for a teacher's name, but there's almost no spaces in a person's first name. But in general, how do I preserve the spacing?

This is my constructor:

Course( stringstream &ss ){


    ss >> dept >> courseNumber >> courseTitle >> numAs >> numBs >> numCs >> numDs >> numFs >> numIs >> numNRs
    >> numSs  >> numUs >> numWs >> instructor;

    dump();         
}

This is in main where I read the file, dump is just a function I am using to output the data that I get to verify if it's correct:

string dummyLine; // string meant to be used to skip over first line
    getline(gradeFile, dummyLine);
    while ( gradeFile.good() ){ // read until the end of the file

        stringstream ss;
        getline(gradeFile, input); // r)ead a line of input, stop at a comma
        replace(input.begin(), input.end() , ',', ' ');

        ss << input;
        Course course(ss);


        if ( count == 0 ){
            //course.dump();
        }

        count++;

        break;
    }

But actually in my output I'm getting a bunch of 0's after 'Intro' in my real data even though the numbers in the csv file aren't 0, so that's also puzzling to me because I thought that each token after that would get the value that the token that I wanted to input would've got, unless it's because the other fields up to 13 were ints.

You might consider a different technique. https://stackoverflow.com/questions/1120140/how-can-i-read-and-parse-csv-files-in-c — Retired Ninja, Sep 08 '18 at 04:52

score 0 · Answer 1 · answered Sep 08 '18 at 10:56

Streams is not really a good fit for what you are trying to solve. While possible in a few different ways, you would have to go through a lot of hassle to get it working. Instead, it is much easier to just read the entire the line/whatever and then tokenize the string - use of a regular expression makes perfect sense here i think.

Extracting specific tokens with stringstream

1 Answers1