Regular expression, not finding match

Question

I have code that goes through a file to find dates but its not returning that it found a match to my regular expression.

CODE:

std::string s(line);
std::smatch m;
std::regex e("^[0,1]?\d{1}\/(([0-2]?\d{1})|([3][0,1]{1}))\/(([1]{1}[9]{1}[9]{1}\d{1})|([2-9]{1}\d{3}))$");   
std::cout << "Target sequence: " << s << std::endl;
std::cout << "Regular expression: ^[0,1]?\d{1}\/(([0-2]?\d{1})|([3][0,1]{1}))\/(([1]{1}[9]{1}[9]{1}\d{1})|([2-9]{1}\d{3}))$" << std::endl;
std::cout << "The following matches and submatches were found:" << std::endl;

while (std::regex_search(s, m, e)) {
    for (auto x : m) std::cout << x << " ";
    std::cout << std::endl;
    s = m.suffix().str();
}

OUTPUT:

Success
Target sequence: 12/28/2002     2   15   38   43   50
Regular expression: ^[0,1]?d{1}/(([0-2]?d{1})|([3][0,1]{1}))/(([1]{1}[9]{1}[9]{1
}d{1})|([2-9]{1}d{3}))$
The following matches and submatches were found:
Enter q to quit:

Is my regular expression not correct or is it something else?

Answer 1

The cause is in your regex and in how you specify the string literal:

• Before we make any fix to your regex, try printing out the string literal to the console:

  std::cout << "^[0,1]?\d{1}\/(([0-2]?\d{1})|([3][0,1]{1}))\/(([1]{1}[9]{1}[9]{1}\d{1})|([2-9]{1}\d{3}))$";
  

  You will see that \ are missing, and <regex> can't see them.

  To specify \ in a string, you need to escape it, \\.

  By the way, printing the string is one of the debugging step in languages where there is no dedicated RegExp literal and the construction must be done via string.

• You are anchoring your search with ^ and $. It will only find a match if the date is on its own in a line, and there must not even be leading or trailing spaces.

• You are having a lot of redundant syntax, e.g. [1]{1} or . A character class with single character (which is not a special character in regex) can be taken out, i.e. 1{1}. And {1} is always redundant, i.e. [1]{1} can be shortened to 1.

• / doesn't need escaping, either in the string literal or regex.

• Fixing the syntactic problems above and remove ^ and $:

  "[0,1]?\\d/(([0-2]?\\d)|(3[0,1]))/((199\\d)|([2-9]\\d{3}))"
  

• By [0,1], you probably want [01]. When you want to match either character A or B, just place them next to each other in character class [AB]. Your [0,1] will also match a comma ,.

• You can drop the () in ([0-2]?\\d), (3[0,1]). The same for year portion. The outer capturing group is enough.

• Applying the 2 points above:

  "[01]?\\d/([0-2]?\\d|3[01])/(199\\d|[2-9]\\d{3})"
  

The regex should now work when you want to extract data, but not so nice if you want to use it to validate. I don't know why you restrict the year to 1990 to 9999, but it's probably your business logic.