基于libcurl实现的HTTP客户端
文章目录
- libcurl安装
- HttpClient
libcurl安装
libcurl的编译安装请参考博客https://blog.csdn.net/gg_simida/article/details/80536860
HttpClient
我们的目标是封装一个HttpClient类，支持GET、POST或者自定义方法，支持发送和接受文本、json、xml、form-data、x-www-form-urlencoded数据，支持自定义头部Headers等
// HttpRequest.h
#ifndef HTTP_REQUEST_H_
#define HTTP_REQUEST_H_
#include
#include
#include HttpRequest.h头文件中定义了ContentType,FormData,HttpRequest,HttpResponse等数据结构；
// HttpClient.h
#ifndef HTTP_CLIENT_H_
#define HTTP_CLIENT_H_
/***************************************************************
HttpClient based libcurl
***************************************************************/
#include
#include
#include "HttpRequest.h"
class HttpClient {
public:
HttpClient();
~HttpClient();
int Send(const HttpRequest& req, HttpResponse* res);
static const char* strerror(int errcode);
void SetTimeout(int sec) {m_timeout = sec;}
void AddHeader(string key, string value) {
m_headers[key] = value;
}
void DelHeader(string key) {
auto iter = m_headers.find(key);
if (iter != m_headers.end()) {
m_headers.erase(iter);
}
}
void ClearHeader() {
m_headers.clear();
}
protected:
int curl(const HttpRequest& req, HttpResponse* res);
private:
static std::atomic_flag s_bInit;
int m_timeout; // unit:s default:10s
KeyValue m_headers;
};
#endif // HTTP_CLIENT_H_
HttpClient.h头文件声明了基于libcurl的HTTP客户端类；
// HttpClient.cpp
#include "HttpClient.h"
#include
#include
using std::string;
#define SPACE_CHARS " \t\r\n"
static string trim(const string& str) {
string::size_type pos1 = str.find_first_not_of(SPACE_CHARS);
if (pos1 == string::npos) return "";
string::size_type pos2 = str.find_last_not_of(SPACE_CHARS);
return str.substr(pos1, pos2-pos1+1);
}
std::atomic_flag HttpClient::s_bInit;
#define DEFAULT_TIMEOUT 10
HttpClient::HttpClient() {
if (!s_bInit.test_and_set()) {
curl_global_init(CURL_GLOBAL_ALL);
}
m_timeout = DEFAULT_TIMEOUT;
}
HttpClient::~HttpClient() {
}
int HttpClient::Send(const HttpRequest& req, HttpResponse* res) {
return curl(req, res);
}
static size_t s_formget_cb(void *arg, const char *buf, size_t len) {
return len;
}
static size_t s_header_cb(char* buf, size_t size, size_t cnt, void* userdata) {
if (buf == NULL || userdata == NULL) return 0;
HttpResponse* res = (HttpResponse*)userdata;
string str(buf);
string::size_type pos = str.find_first_of(':');
if (pos == string::npos) {
if (res->version.empty() && strncmp(buf, "HTTP", 4) == 0) {
// status line
// HTTP/1.1 200 OK\r\n
char* space1 = strchr(buf, ' ');
char* space2 = strchr(space1+1, ' ');
if (space1 && space2) {
*space1 = '\0';
*space2 = '\0';
res->version = buf;
res->status_code = atoi(space1+1);
res->status_message = trim(space2+1);
}
}
}
else {
// headers
string key = trim(str.substr(0, pos));
string value = trim(str.substr(pos+1));
res->headers[key] = value;
}
return size*cnt;
}
static size_t s_body_cb(char *buf, size_t size, size_t cnt, void *userdata) {
if (buf == NULL || userdata == NULL) return 0;
HttpResponse* res = (HttpResponse*)userdata;
res->body.append(buf, size*cnt);
return size*cnt;
}
static string escape(const string& param) {
string res;
const char* p = param.c_str();
int len = param.size();
char escape[4] = {0};
for (int i = 0; i < len; ++i) {
if (p[i] == ' ' ||
p[i] == '?' ||
p[i] == '=' ||
p[i] == '&' ||
p[i] == '%') {
sprintf(escape, "%%%02X", p[i]);
res += escape;
}
else {
res += p[i];
}
}
return res;
}
inline bool is_hex(char c) {
return (c >= '0' && c <= '9') ||
(c >= 'A' && c <= 'F') ||
(c >= 'a' && c <= 'f');
}
inline int c2i(char c) {
return c-'0';
}
static string unescape(const string& escape_param) {
string res;
const char* p = escape_param.c_str();
int len = escape_param.size();
int i = 0;
while (i < len) {
if (p[i] == '%' && i < len-2 && is_hex(p[i+1]) && is_hex(p[i+2])) {
res += c2i(p[i+1])*16 + c2i(p[i+2]);
i += 3;
}
else {
res += p[i];
++i;
}
}
return res;
}
int HttpClient::curl(const HttpRequest& req, HttpResponse* res) {
CURL* handle = curl_easy_init();
// SSL
curl_easy_setopt(handle, CURLOPT_SSL_VERIFYPEER, 0);
curl_easy_setopt(handle, CURLOPT_SSL_VERIFYHOST, 0);
// method
curl_easy_setopt(handle, CURLOPT_CUSTOMREQUEST, req.method.c_str());
// url
curl_easy_setopt(handle, CURLOPT_URL, req.url.c_str());
// header
struct curl_slist *headers = NULL;
if (m_headers.size() != 0) {
for (auto& pair : m_headers) {
string header = pair.first;
header += ": ";
header += pair.second;
headers = curl_slist_append(headers, header.c_str());
}
}
const char* psz = "text/plain";
switch (req.content_type) {
#define CASE_CONTENT_TYPE(id, str) \
case id: psz = str; break;
FOREACH_CONTENT_TYPE(CASE_CONTENT_TYPE)
#undef CASE_CONTENT_TYPE
}
string strContentType("Content-type: ");
strContentType += psz;
headers = curl_slist_append(headers, strContentType.c_str());
curl_easy_setopt(handle, CURLOPT_HTTPHEADER, headers);
//hlogd("%s %s", req.method.c_str(), req.url.c_str());
//hlogd("%s", strContentType.c_str());
// body or params
struct curl_httppost* httppost = NULL;
struct curl_httppost* lastpost = NULL;
switch (req.content_type) {
case FORM_DATA:
{
for (auto& pair : req.form) {
CURLformoption opt = pair.second.type == FormData::FILENAME ? CURLFORM_FILE : CURLFORM_COPYCONTENTS;
curl_formadd(&httppost, &lastpost,
CURLFORM_COPYNAME, pair.first.c_str(),
opt, pair.second.data.c_str(),
CURLFORM_END);
}
if (httppost) {
curl_easy_setopt(handle, CURLOPT_HTTPPOST, httppost);
curl_formget(httppost, NULL, s_formget_cb);
}
}
break;
case QUERY_STRING:
case X_WWW_FORM_URLENCODED:
{
string params;
auto iter = req.kvs.begin();
while (iter != req.kvs.end()) {
if (iter != req.kvs.begin()) {
params += '&';
}
params += escape(iter->first);
params += '=';
params += escape(iter->second);
iter++;
}
if (req.content_type == QUERY_STRING) {
string url_with_params(req.url);
url_with_params += '?';
url_with_params += params;
curl_easy_setopt(handle, CURLOPT_URL, url_with_params.c_str());
//hlogd("%s", url_with_params.c_str());
}
else {
curl_easy_setopt(handle, CURLOPT_POSTFIELDS, params.c_str());
//hlogd("%s", params.c_str());
}
}
break;
default:
{
if (req.text.size() != 0) {
curl_easy_setopt(handle, CURLOPT_POSTFIELDS, req.text.c_str());
//hlogd("%s", req.text.c_str());
}
}
break;
}
if (m_timeout != 0) {
curl_easy_setopt(handle, CURLOPT_TIMEOUT, m_timeout);
}
curl_easy_setopt(handle, CURLOPT_WRITEFUNCTION, s_body_cb);
curl_easy_setopt(handle, CURLOPT_WRITEDATA, res);
curl_easy_setopt(handle, CURLOPT_HEADER, 0);
curl_easy_setopt(handle, CURLOPT_HEADERFUNCTION, s_header_cb);
curl_easy_setopt(handle, CURLOPT_HEADERDATA, res);
int ret = curl_easy_perform(handle);
if (ret != 0) {
//hloge("%d: %s", ret, curl_easy_strerror((CURLcode)ret));
}
//if (res->body.length() != 0) {
//hlogd("Response:%s", res->body.c_str());
//}
//double total_time, name_time, conn_time, pre_time;
//curl_easy_getinfo(handle, CURLINFO_TOTAL_TIME, &total_time);
//curl_easy_getinfo(handle, CURLINFO_NAMELOOKUP_TIME, &name_time);
//curl_easy_getinfo(handle, CURLINFO_CONNECT_TIME, &conn_time);
//curl_easy_getinfo(handle, CURLINFO_PRETRANSFER_TIME, &pre_time);
//hlogd("TIME_INFO: %lf,%lf,%lf,%lf", total_time, name_time, conn_time, pre_time);
if (headers) {
curl_slist_free_all(headers);
}
if (httppost) {
curl_formfree(httppost);
}
curl_easy_cleanup(handle);
return ret;
}
const char* HttpClient::strerror(int errcode) {
return curl_easy_strerror((CURLcode)errcode);
}
HttpClient.cpp源文件根据HttpRequest利用libcurl完成请求，在回调函数中解析填充HttpResponse
// test.cpp
#include "HttpClient.h"
#include
int main(int argc, char* argv[]) {
HttpClient session;
HttpRequest req;
req.method = "GET";
req.url = "www.baidu.com";
HttpResponse res;
int ret = session.Send(req, &res);
if (ret != 0) {
printf("%s %s failed => %d:%s\n", req.method.c_str(), req.url.c_str(), ret, HttpClient::strerror(ret));
}
else {
printf("%s %d %s\r\n", res.version.c_str(), res.status_code, res.status_message.c_str());
for (auto& header : res.headers) {
printf("%s: %s\r\n", header.first.c_str(), header.second.c_str());
}
printf("\r\n");
printf("%s", res.body.c_str());
printf("\n");
}
return ret;
}
封装过后使用起来相当方便吧，一点不输python的requests