How to bypass commented lines in external text file using ifstream in C++

Question

I'm having trouble filtering the 'commented' lines in a text file. I want to filter the lines beginning with # or // in the text file.

Thanks!

string code, name, year, semester, value, data;

char delimiter = '|';

ifstream ifsUnits;
ifsUnits.open("./data/units.txt");

if (ifsUnits.fail())
    cout << "\nError reading from file <units.txt>.";
else
{
    while (!ifsUnits.eof())
    {
        getline(ifsUnits, data);
        stringstream ssData(data);
        while (ssData.good())
        {
            getline(ssData, code, delimiter);
            getline(ssData, name, delimiter);
            getline(ssData, year, delimiter);
            getline(ssData, semester, delimiter);
            getline(ssData, value, delimiter);

            lUnits.push_back(Unit(stoi(code), name, stoi(year), stoi(semester), stoi(value)));
        }
    }
}
ifsUnits.close();

Text file contents:

how about just getting the first segment `code`, detecting if it starts with `//` and then `continue`? — JHBonarius, Dec 22 '20 at 19:59
@JHBonarius there's probably no guaranteed `delimiter` in a commented line. But that's something OP should clarify — scohe001, Dec 22 '20 at 19:59
A separate issue: [Why it's bad to use feof() to control a loop](https://faq.cprogramming.com/cgi-bin/smartfaq.cgi?answer=1046476070&id=1043284351). The article is about C, but it applies here as well. — Michael Burr, Dec 22 '20 at 20:00
just suggesting. Could also just get the first two chars, and add them to `code` if they're not `//`. — JHBonarius, Dec 22 '20 at 20:01
One way: instead of calling `getline()` directly, write a wrapper that calls `getline()` and discards any that start with "//" or "#". Return the first line that doesn't. — Michael Burr, Dec 22 '20 at 20:02
Simply check if the first two characters of `data` contain `'#'` or `'/'`. — πάντα ῥεῖ, Dec 22 '20 at 20:02
Or better, use a regex that checks for any amount of leading whitespace followed by `#` or `"//"` and if it matches, simply discard the line by `continue;` in your read loop. (always control your read loop with the return from your read function and stream-state) — David C. Rankin, Dec 22 '20 at 20:03
I would have generally thought so too, but given their simple implementation, it's certainly an option. Of course `.find_first_of()` and `.find_first_not_of()` are also simple alternatives. — David C. Rankin, Dec 22 '20 at 20:05
[Watch out for `while (!ifsUnits.eof())`](https://stackoverflow.com/questions/5605125/why-is-iostreameof-inside-a-loop-condition-i-e-while-stream-eof-cons). `while (ssData.good())` will result in similar bugs.. — user4581301, Dec 22 '20 at 20:15

scohe001 · Accepted Answer · 2020-12-23T13:19:21.483

2

You're already pulling one line into a string at a time. Just check if that string starts with your comment characters and if so, continue:

while (getline(ifsUnits, data)) // more robust than eof check
{
    if ((data.size() > 1 && data[0] == '/' && data[1] == '/')
        || (data.size() > 0 && data[0] == '#')) { continue; }
    
    stringstream ssData(data);
    // ...

edited Dec 23 '20 at 13:19

answered Dec 22 '20 at 20:06

scohe001

13,879
2
28
47

while (getline(ifsUnits, data)) // more robust than eof check Why if this method more robust? Thanks for you answer. – zumuha Dec 22 '20 at 20:58
1

@zumuha [Why is iostream::eof inside a loop condition (i.e. `while (!stream.eof())`) considered wrong?](https://stackoverflow.com/q/5605125/2602718) – scohe001 Dec 22 '20 at 21:19

How to bypass commented lines in external text file using ifstream in C++

1 Answers1