C++正则表达式匹配URL失效，无法提取端口问题求助

It’s frustrating when a regex works perfectly online but fails in C++—let’s break down why this is happening and how to fix it.

Key Issues in Your Code #

1. Wrong Regex Syntax Flag: You’re using std::regex_constants::extended, which enables POSIX Extended Regular Expression syntax. Most online regex testers use ECMAScript syntax (the default for std::regex), so switching flags will align your code with the behavior you saw online.
2. Invalid Test String: Your test string "http" doesn’t match your regex because the pattern requires https?:// (the :// part is missing from "http"). You need to test with valid URLs like "http://example.com" or "https://example.com:8080/path".
3. Potential Misuse of Matching Functions: Ensure you’re using std::regex_search if you want to find the pattern anywhere in the string, or std::regex_match if you want the entire string to match the regex (your current regex ends with .*, so regex_match should work for full URLs).

Working Code Example #

Here’s a corrected version that matches URLs as expected and extracts ports:

#include <regex>
#include <string>
#include <iostream>

int main() {
    // Use default ECMAScript syntax (no extended flag)
    std::regex r(R"(https?://[0-9a-zA-Z.-]+(:([0-9]+))?.*)");
    
    // Test with valid URLs (including edge cases like subdomains and ports)
    std::string urls[] = {
        "http://example.com",
        "https://example.com:8080/path/to/page",
        "http://127.0.0.1:3000",
        "http://sub-domain.co.uk",
        "https://test.com"
    };
    
    for (const auto& url : urls) {
        std::smatch match;
        if (std::regex_match(url, match, r)) {
            std::cout << "Matched: " << url << "\n";
            
            // Extract port if it exists
            if (!match[2].str().empty()) {
                std::cout << "→ Port found: " << match[2].str() << "\n";
            }
        } else {
            std::cout << "Not matched: " << url << "\n";
        }
    }
    
    return 0;
}

Explanation #

• Removed extended Flag: The default ECMAScript syntax matches what online tools use, so your regex will behave consistently.
• Raw String Literal: Using R"(...)" avoids messy escape sequences (though slashes don’t need escaping here, it’s a good habit for regex in C++).
• Updated Character Class: Added - to the domain character class to support hyphenated subdomains (a common edge case).
• Port Extraction: Uses std::smatch to retrieve the captured port value from group 2 of the regex.

Additional Notes #

For production-grade URL parsing, consider using a dedicated library (like Boost.Regex or a purpose-built URL parser) instead of regex—URLs have complex edge cases (like international domains or query parameters) that regexes struggle to handle perfectly.

内容的提问来源于stack exchange，提问作者Jtvd78