add http_url class for parsing url

lib_acl_cpp/include/acl_cpp/http/http_utils.hpp

@@ -1,5 +1,6 @@
 #pragma once
 #include "../acl_cpp_define.hpp"
+#include "../stdlib/string.hpp"
 
 namespace acl {
 
@@ -28,6 +29,62 @@ public:
 	 */
 	static bool get_addr(const char* url, char* domain, size_t size,
 		unsigned short* port);
+
+	/**
+	 * 解析输入的完整或部分 URL
+	 * @param url {const char*} 非空完整或部分 URL 字符串
+	 * @param domain {string&} 用来存放域名地址信息
+	 * @param size {size_t} domain 缓冲区大小
+	 * @param port {unsigned short*} 用来存放 url 中的端口号
+	 * @param url_path {string&} 用来存放 url 中不含域名和参数的部分，即针对类似
+	 *  于 URL：http://test.com.cn/cgi-bin/test?name=value 则只提取其中的字
+	 *  符串：/cgi-bin/test
+	 * @param url_params {string&} 用来存放 url 中的参数部分字符串
+	 * @return {bool} 解析 url 是否成功
+	 */
+	static bool parse_url(const char* url, string& domain,
+		unsigned short* port, string& url_path, string& url_params);
+};
+
+class ACL_CPP_API http_url {
+public:
+	http_url(void);
+	~http_url(void) {}
+
+	bool parse(const char* url);
+
+public:
+	const char* get_proto(void) const {
+		return proto_;
+	}
+
+	const char* get_domain(void) const {
+		return domain_.c_str();
+	}
+
+	unsigned short get_port(void) const {
+		return port_;
+	}
+
+	const char* get_url_path(void) const {
+		return url_path_.c_str();
+	}
+
+	const char* get_url_params(void) const {
+		return url_params_.c_str();
+	}
+
+	void reset(void);
+
+private:
+	char proto_[16];
+	string domain_;
+	unsigned short port_;
+	string url_path_;
+	string url_params_;
+
+	bool parse_url_part(const char* url);
+	const char* parse_domain(const char* url);
 };
 
 } // namespace acl

lib_acl_cpp/samples/url_coder/url_coder.cpp

@@ -4,6 +4,7 @@
 #include "stdafx.h"
 #include "acl_cpp/stdlib/string.hpp"
 #include "acl_cpp/stdlib/url_coder.hpp"
+#include "acl_cpp/http/http_utils.hpp"
 
 using namespace acl;
 
@@ -75,11 +76,45 @@ int main(void)
 		v0, found0 ? "found it" : "not found", coder5.get("n1"),
 		v3, found3 ? "found it" : "not found");
 
-#ifdef WIN32
-	printf("enter any key to exit ...\r\n");
+	printf("enter any key to continue ...\r\n");
+	getchar();
+
+	printf("\r\n");
+
+	acl::http_url hu;
+	const char* urls[] = {
+		"http://www.google.com/",
+		"https://www.google.com/",
+		"https://www.google.com/test",
+		"http://www.google.com/test?name=value&name2=value2",
+		"/test",
+		"/",
+		"/test?",
+		"/test?name1=value1&name2=value2",
+		"/path/test",
+		"/path/test?name=value",
+		NULL,
+	};
+
+	for (size_t i = 0; urls[i] != NULL; i++) {
+		if (!hu.parse(urls[i])) {
+			printf("parse url=%s failed\r\n", urls[i]);
+			break;
+		}
+
+		printf("url:%s\r\n", urls[i]);
+		printf("proto=%s, port=%d, domain=%s, path=%s, params=%s\r\n",
+			hu.get_proto(), hu.get_port(), hu.get_domain(),
+			hu.get_url_path(), hu.get_url_params());
+		printf("\r\n");
+
+		hu.reset();
+	}
+
+#if defined(_WIN32) || defined(_WIN64)
+	printf("Enter any key to exit ...\r\n");
 	getchar();
 #endif
-
 	return 0;
 }
 

lib_acl_cpp/src/http/http_utils.cpp

@@ -19,13 +19,12 @@ bool http_utils::get_addr(const char* url, char* addr, size_t size)
 	safe_snprintf(addr, size, "%s:%d", buf, port);
 	return true;
 }
+#define	HTTP_PREFIX	"http://"
+#define	HTTPS_PREFIX	"https://"
 
 bool http_utils::get_addr(const char* url, char* domain, size_t size,
 	unsigned short* pport)
 {
-#define	HTTP_PREFIX	"http://"
-#define	HTTPS_PREFIX	"https://"
-
 	const char* ptr;
 	unsigned short default_port;
 
@@ -73,4 +72,111 @@ bool http_utils::get_addr(const char* url, char* domain, size_t size,
 	return true;
 }
 
+//////////////////////////////////////////////////////////////////////////////
+
+http_url::http_url(void) {
+	ACL_SAFE_STRNCPY(proto_, "http", sizeof(proto_));
+	port_ = 80;
+}
+
+void http_url::reset(void) {
+	ACL_SAFE_STRNCPY(proto_, "http", sizeof(proto_));
+	port_ = 80;
+	domain_.clear();
+	url_path_.clear();
+	url_params_.clear();
+}
+
+bool http_url::parse(const char *url) {
+	const char* ptr;
+
+	if (!strncasecmp(url, HTTP_PREFIX, sizeof(HTTP_PREFIX) - 1)) {
+		ptr = url + sizeof(HTTP_PREFIX) - 1;
+	} else if (!strncasecmp(url, HTTPS_PREFIX, sizeof(HTTPS_PREFIX) - 1)) {
+		ptr = url + sizeof(HTTPS_PREFIX) - 1;
+		port_ = 443;
+		ACL_SAFE_STRNCPY(proto_, "https", sizeof(proto_));
+	} else if (*url == '/'){
+		ptr = url;
+	} else {
+		logger_error("invalid url: %s", url);
+		return false;
+	}
+
+	if (*ptr == 0) {
+		logger_error("invalid url: %s", url);
+		return false;
+	}
+
+	if (ptr == url) {
+		return parse_url_part(url);
+	} else {
+		ptr = parse_domain(ptr);
+		if (ptr == NULL) {
+			url_path_ = "/";
+			return true;
+		}
+		return parse_url_part(ptr);
+	}
+}
+
+#define SKIP_WHILE(cond, ptr) { while(*ptr && (cond)) ptr++; }
+
+bool http_url::parse_url_part(const char *url) {
+	if (*url != '/') {
+		logger_error("invalid url: %s", url);
+		return false;
+	}
+
+	const char* ptr = url;
+	SKIP_WHILE(*ptr == '/', ptr);
+	if (*ptr == 0) {
+		url_path_ = "/";
+		return true;
+	}
+
+	const char* qm = strchr(ptr, '?');
+	if (qm == NULL) {
+		url_path_ = url;
+		return true;
+	}
+	url_path_.copy(url, qm - url);
+	++qm;
+	if (*qm != 0) {
+		url_params_ = qm;
+	}
+	return true;
+}
+
+const char* http_url::parse_domain(const char *url) {
+	if (*url == '/') {
+		logger_error("invalid url: %s", url);
+		return NULL;
+	}
+	const char* ptr = strchr(url, '/');
+	if (ptr == NULL) {
+		domain_ = url;
+		return NULL;
+	}
+
+	char buf[256];
+	size_t size = ptr - url + 1;
+	if (size > sizeof(buf)) { // xxx: sanity check
+		size = sizeof(buf);
+	}
+	ACL_SAFE_STRNCPY(buf, url, size);
+
+	// fixme: Is it error if buf contains IPV6 Addr ---zsx
+	char* col = strchr(buf, ':');
+	if (col != NULL) {
+		*col++ = 0;
+		port_ = (unsigned short) atoi(col);
+		if (port_ == 0 || port_ == 65535) {
+			port_ = 80;
+		}
+	}
+	domain_ = buf;
+	return ptr;
+}
+
 } // namespace acl