7

I want to build a regular expression in c++{MFC} which validates the URL.

The regular expression must satisfy following conditions.

Valid URL:- http://cu-241.dell-tech.co.in/MyWebSite/ISAPIWEBSITE/Denypage.aspx/ http://www.google.com http://www.google.co.in

Invalid URL:-

  1. http://cu-241.dell-tech.co.in/\MyWebSite/\ISAPIWEBSITE/\Denypage.aspx/ = Regx must check & invalid URL as '\' character between "/\MyWebSite/\ISAPIWEBSITE/\Denypage.aspx/"

  2. http://cu-241.dell-tech.co.in//////MyWebSite/ISAPIWEBSITE/Denypage.aspx/ = Regx must check & invalidate URL due to multiple entries of "///////" in url.

  3. http://news.google.co.in/%5Cnwshp?hl=en&tab=wn = Regex must check & invalidate URL for additional insertion of %5C & %2F character.

How can we develop a generic Regular Expression satisfying above condition. Please, Help us by providing a regular expression that will handle above scenario's in CPP{MFC}

Spudley
  • 161,975
  • 39
  • 229
  • 303
user701960
  • 71
  • 1
  • 1
  • 2

2 Answers2

12

Have you tried using the RFC 3986 suggestion? If you're capable of using GCC-4.9 then you can go directly with <regex>.

It states that with ^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\?([^#]*))?(#(.*))? you can get as submatches:

  scheme    = $2
  authority = $4
  path      = $5
  query     = $7
  fragment  = $9

For example:

int main(int argc, char *argv[])
{
  std::string url (argv[1]);
  unsigned counter = 0;

  std::regex url_regex (
    R"(^(([^:\/?#]+):)?(//([^\/?#]*))?([^?#]*)(\?([^#]*))?(#(.*))?)",
    std::regex::extended
  );
  std::smatch url_match_result;

  std::cout << "Checking: " << url << std::endl;

  if (std::regex_match(url, url_match_result, url_regex)) {
    for (const auto& res : url_match_result) {
      std::cout << counter++ << ": " << res << std::endl;
    }
  } else {
    std::cerr << "Malformed url." << std::endl;
  }

  return EXIT_SUCCESS;
}

Then:

./url-matcher http://localhost.com/path\?hue\=br\#cool

Checking: http://localhost.com/path?hue=br#cool
0: http://localhost.com/path?hue=br#cool
1: http:
2: http
3: //localhost.com
4: localhost.com
5: /path
6: ?hue=br
7: hue=br
8: #cool
9: cool
Community
  • 1
  • 1
Ciro Costa
  • 2,265
  • 21
  • 25
  • This is working really great. Could you show me how to use this to extract all matching urls in a string using the regex? I tried to use it with sregex_iterator but I don't get any matches. Thank you very much! – Julius Feb 28 '16 at 10:31
  • 6
    Unforutnately this is not for validating but for splitting a correct URI into its parts. It will not even detect the most simple cases, like unencoded spaces. – Lothar Jul 27 '16 at 03:27
  • Thanks for such a useful and well-explained answer. This is the best all-around URL parsing script I've found for accuracy, ease of use, and quickness of implementation. And you don't need to download any special libraries! It would make a great answer to this question: https://stackoverflow.com/q/2616011/1043704 – Lorien Brune Sep 30 '17 at 03:19
  • Note that if a port is present, it will be included in elements 3 and 4. I.e., `http://localhost.com:8888/path?hue=br#cool` results in `3: //localhost.com:8888` and `4: localhost.com:8888`. – Lorien Brune Sep 30 '17 at 03:49
  • not working, "aaa" will be accepted by this expression – Andrey Nekrasov Dec 02 '20 at 12:52
0

look at http://gskinner.com/RegExr/, there is a community tab on the right where you find contributed regex's. There is a URI category, not sure you'll find exactly what you need but this is a good start

davka
  • 13,226
  • 11
  • 58
  • 82