Method ParseURL

Sources/Dependencies/ixwebsocket/IXUrlParser.cpp:139–378

Practical URI parser for ws/wss/http/https URLs. Standards references: - RFC 3986: generic URI syntax used for scheme, authority, path, query, fragment. - RFC 3986 Section 3.2.2: IP-literal host syntax (bracketed IPv6 in authority). - RFC 6455 Section 3: ws/wss URI schemes (default ports handled in getProtocolPort). Note: this is not a full RFC validator; it is a fast parser with targeted syntax

	//
	// Note: this is not a full RFC validator; it is a fast parser with targeted syntax checks.
	clParseURL clParseURL::ParseURL(const std::string& URL)
	{
		clParseURL Result;
		const std::size_t npos = std::string::npos;
		std::size_t current = 0;

		// Step 1: read scheme (<scheme>:) and normalize to lowercase.
		std::size_t schemeEnd = URL.find(':', current);
		if (schemeEnd == npos)
		{
			return clParseURL(LUrlParserError_NoUrlCharacter);
		}

		Result.m_Scheme = URL.substr(current, schemeEnd - current);
		if (!IsSchemeValid(Result.m_Scheme))
		{
			return clParseURL(LUrlParserError_InvalidSchemeName);
		}

		std::transform(Result.m_Scheme.begin(),
					   Result.m_Scheme.end(),
					   Result.m_Scheme.begin(),
					   ::tolower);

		current = schemeEnd + 1;

		// Step 2: require authority prefix "//".
		if (current + 1 >= URL.size() || URL[current] != '/' || URL[current + 1] != '/')
		{
			return clParseURL(LUrlParserError_NoDoubleSlash);
		}
		current += 2;

		// Step 3: locate the authority boundary (<authority> ends at '/' or '?').
		std::size_t authorityEnd = URL.find_first_of("/?", current);
		if (authorityEnd == npos)
		{
			authorityEnd = URL.size();
		}

		// Step 4: parse optional user info (<user>[:<password>]@).
		bool hasUserInfo = false;
		std::size_t atPos = URL.find('@', current);
		if (atPos != npos && atPos < authorityEnd)
		{
			hasUserInfo = true;
		}

		if (hasUserInfo)
		{
			// User info can contain ':' but must contain exactly one '@' separator
			// within the authority segment.
			if (URL.find('@', atPos + 1) != npos && URL.find('@', atPos + 1) < authorityEnd)
			{
				return clParseURL(LUrlParserError_MultipleAtSignsInAuthority);
			}

			std::size_t colonPos = URL.find(':', current);