Thread: Regex match ISBN number

I want to regex match an ISBN number of the format: 1234-5678-9x ie. 4 digits [hyphen] 4 digits [hyphen] 1 digit 1 alphanum but the following program rejects valid output. Any help much appreciated (also, if not too much trouble, please tell me where and how I'm wrong as well):

Code:

#include <iostream>
#include <string>
#include <regex>
int main()
{
    std::regex ISBN ("[0-9][{4}][-][0-9][{4}][-][0-9][::alnum::]");
    std::cout << "Enter ISBN number \n";
    std::string input{};
    getline(std::cin, input);
    if (std::regex_match(input, ISBN))
    {
        std::cout << "Success \n";
    }
    else
    {
        std::cout << "Fail \n";
    }
}

Thank you

There is a [::digit::] class too in POSIX.

regex assumes you will use ECMAScript grammar, try constructing like this:

Code:

std::regex isbn ("[::digit::][{4}][-][::digit::][{4}][-][::digit::][::alnum::]", std::regex::basic);

std::basic_regex - cppreference.com

Also don't forget that the last character may only be an 'x', 'X', or any valid digit.

Jim

I feel bad for giving such a broken example. Regex does support escaped character classes though, as the docs in Salem's link explains.

This does match ISBNs:

Code:

        string re = "(\\d{4})-(\\d{4})-((\\d)(\\d|X))";
        if (regex_match(tests[i], regex(re)))

Many thanks to all for your replies, reverse engineering can see I was going wrong in:
(a) putting brackets around the preceding atom match quantifiers
(b) not putting brackets around class names
(c) putting an extra : in front and behind :alnum:
(d) also brackets around the hyphen character seems optional
whiteflags: I tweaked your code slightly to take into a/c all words including the 'X' from my OP example
So putting everything together, here are a few that work and I'm sure there are several others:

Code:

std::regex ISBN ("[0-9]{4}[-][0-9]{4}[-][0-9][[:alnum:]]");
std::regex ISBN ("(\\d{4})-(\\d{4})-((\\d)(\\d|[[:alpha:]]))"); (whiteflags, modified)
std::regex ISBN ("(\\d{4})-(\\d{4})-((\\d)(\\d|\\w))"); (whiteflags, modified)
std::regex ISBN ("[0-9]{4}[-][0-9]{4}[-][0-9][[:alnum:]]"); (Salem)
std::regex ISBN ("[[:d:]]{4}[-][[:d:]]{4}[-][[:d:]][[:alnum:]]");
std::regex ISBN ("[\\d]{4}[-][\\d]{4}[-][\\d][\\w]");

Some people, when confronted with a problem, think "I know, I'll use regular expressions." Now they have two problems.

I couldn't resist :P

(a) putting brackets around the preceding atom match quantifiers
(b) not putting brackets around class names
(c) putting an extra : in front and behind :alnum:
(d) also brackets around the hyphen character seems optional
whiteflags: I tweaked your code slightly to take into a/c all words including the 'X' from my OP example

If you put [] around a hyphen it just turns it into a character class.
\w is really too permissive when you mean X (you could include x but frankly I would reject that as well).

Additionally, the () in the regex is largely for grouping. In my last one, match groups 1 through 3 would be all the numbers, and 5 is just the check digit.