File stops being read after a newline character

Question

I'm trying to create a spam filter. I need to train the model first. I read the words from a text file which has the word "spam" or "ham" as the first word of a paragraph, and then the words in the mail and number of its occurrences just after the word. There are paragraphs in the file. My program is able to read the first paragraph that is the words and their number of occurrences.

The problem is that, the file stops reading after encountering the newline that and doesn't read the next paragraph. Although I have a feeling that the way I am checking for a newline character that is the end of a paragraph is not entirely correct.

I have given two paragraphs so you just get the idea of the train text. Train text file.

/000/003 ham need 1 fw 1 35 2 39 1 thanks 1 thread 2 40 1 copy 1 else 1 correlator 1 under 1 companies 1 25 1 he 2 26 2 168 1 29 2 content 4 1 1 6 1 5 1 4 1 review 2 we 1 john 3 17 1 use 1 15 1 20 1 classes 1 may 1 a 1 back 1 l 1 01 1 produced 1 i 1 yes 1 10 2 713 2 v6 1 p 1 original 2

/000/031 ham don 1 kim 5 dave 1 39 1 customer 1 38 2 thanks 1 over 1 thread 2 year 1 correlator 1 under 1 williams 1 mon 2 number 2 kitchen 1 168 1 29 1 content 4 3 2 2 6 system 2 1 2 7 1 6 1 5 2 4 1 9 1 each 1 8 1 view 2

#include <iostream>
#include <fstream>
#include <string>
using namespace std;

int main()
{
    int V = 0; // Total number of words

    ifstream fin;
    fin.open("train", ios::in);
    string word;
    int wordnum;
    int N[2] = {0};
    char c, skip;
    for (int i = 0; i < 8; i++) fin >> skip; // There are 8 characters before the first word of the paragraph
    while (!fin.fail())
    {
        fin >> word;
        if (word == "spam") N[0]++;
        else if (word == "ham") N[1]++;
        else
        {
            V++;
            fin >> wordnum;
        }
        int p = fin.tellg();
        fin >> c; //To check for newline. If its there, we skip the first eight characters of the new paragraph because those characters aren't supposed to be read
        if (c == '\n')
        {
            for (int i = 0; i < 8; i++) fin >> skip;
        }
        else fin.seekg(p);
    }

    cout << "\nSpam: " << N[0];
    cout << "\nHam :" << N[1];
    cout << "\nVocab: " << V;

    fin.close();

    return 0;
}

`while(!fin.fail()) { /* ... */ }` might not be much better than stated in [here](http://stackoverflow.com/questions/5605125/why-is-iostreameof-inside-a-loop-condition-considered-wrong). Though here's [some starter](http://stackoverflow.com/questions/24504582/how-to-test-whether-stringstream-operator-has-parsed-a-bad-type-and-skip-it), to armour you with some techniques, that could be useful to solve your problem. — πάντα ῥεῖ, Sep 25 '14 at 17:48
Could you provide a sample `train` file (by editing your question)? — Michael Petch, Sep 25 '14 at 17:53
Gosh _who_ is teaching everybody all these incorrect ways of using streams?! — Lightness Races in Orbit, Sep 25 '14 at 17:57
Consider using `istream::peek` (http://www.cplusplus.com/reference/istream/istream/peek/) rather than the `tellg()` `seekg()` combo you are doing. — Mad Physicist, Sep 25 '14 at 18:01
@LightnessRacesinOrbit Hope your _"lightning"_ wasn't smashed against me here (_duck and cover_) ;-) ... — πάντα ῥεῖ, Sep 25 '14 at 18:01
@MadPhysicist _"Whoever put in the close vote"_ It was me, and I well read the question. The [MCVE](http://stackoverflow.com/help/mcve) policy isn't fulfilled, and the exemplary input is missing. — πάντα ῥεῖ, Sep 25 '14 at 18:05
@MadPhysicist: No, they did not read the testcase, and that's because none was provided. — Lightness Races in Orbit, Sep 25 '14 at 18:05

score 1 · Accepted Answer · edited Sep 25 '14 at 18:43

1

std::ifstream::operator>>() doesn't read \n in the variable; it drops it. If you need to manipulate with whitespaces and \n symbols, you can use std::ifstream::get()

edited Sep 25 '14 at 18:43

Keith Pinson

7,162
5
54
97

answered Sep 25 '14 at 18:02

Kastaneda

679
1
7
14

Your English is fine. Thanks for the answer. – Mad Physicist Sep 25 '14 at 18:03
also you can read all line (through std::getline()) in std::string and after split it by words. – Kastaneda Sep 25 '14 at 18:06

File stops being read after a newline character

1 Answers1