将 cURL 内容结果保存到 C++ 中的字符串中

Save cURL content result into a string in C++

提问by Grego

int main(void)
{
  CURL *curl;
  CURLcode res;

  curl = curl_easy_init();
  if(curl) {
    curl_easy_setopt(curl, CURLOPT_URL, "http://www.google.com");
    curl_easy_setopt(curl, CURLOPT_SSL_VERIFYPEER, 0L);
    curl_easy_setopt(curl, CURLOPT_SSL_VERIFYHOST, 0L);
    res = curl_easy_perform(curl);
    curl_easy_cleanup(curl);
  }
  _getch();
  return 0;
}

string contents = "";

I would like to save the result of the curl html content in a string, how do I do this? It's a silly question but unfortunately, I couldn't find anywhere in the cURL examples for C++ thanks!

我想将 curl html 内容的结果保存在一个字符串中，我该怎么做？这是一个愚蠢的问题，但不幸的是，我在 C++ 的 cURL 示例中找不到任何地方，谢谢！

回答by Joachim Isaksson

You will have to use CURLOPT_WRITEFUNCTIONto set a callback for writing. I can't test to compile this right now, but the function should look something close to;

您将不得不使用CURLOPT_WRITEFUNCTION设置回调以进行写入。我现在无法测试编译它，但该函数应该看起来很接近；

static std::string readBuffer;

static size_t WriteCallback(void *contents, size_t size, size_t nmemb, void *userp)
{ 
    size_t realsize = size * nmemb;
    readBuffer.append(contents, realsize);
    return realsize;
}

Then call it by doing;

然后通过执行调用它；

readBuffer.clear();
curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, WriteCallback);
// ...other curl options
res = curl_easy_perform(curl);

After the call, readBuffershould have your contents.

通话后，readBuffer应该有你的内容。

Edit: You can use CURLOPT_WRITEDATAto pass the buffer string instead of making it static. In this case I just made it static for simplicity. A good page to look (besides the linked example above) is herefor an explanation of the options.

编辑：您可以使用CURLOPT_WRITEDATA传递缓冲区字符串而不是使其静态。在这种情况下，为了简单起见，我只是将其设为静态。一个很好看的页面（除了上面的链接示例）是这里对选项的解释。

Edit2: As requested, here's a complete working example without the static string buffer;

Edit2：根据要求，这是一个没有静态字符串缓冲区的完整工作示例；

#include <iostream>
#include <string>
#include <curl/curl.h>


static size_t WriteCallback(void *contents, size_t size, size_t nmemb, void *userp)
{
    ((std::string*)userp)->append((char*)contents, size * nmemb);
    return size * nmemb;
}

int main(void)
{
  CURL *curl;
  CURLcode res;
  std::string readBuffer;

  curl = curl_easy_init();
  if(curl) {
    curl_easy_setopt(curl, CURLOPT_URL, "http://www.google.com");
    curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, WriteCallback);
    curl_easy_setopt(curl, CURLOPT_WRITEDATA, &readBuffer);
    res = curl_easy_perform(curl);
    curl_easy_cleanup(curl);

    std::cout << readBuffer << std::endl;
  }
  return 0;
}

回答by Mark Laagland

Using the 'new' C++11 lambda functionality, this can be done in a few lines of code.

使用“新的”C++11 lambda 功能，只需几行代码即可完成。

#ifndef WIN32 #define __stdcall "" #endif //For compatibility with both Linux and Windows
std::string resultBody { };
curl_easy_setopt(curl, CURLOPT_WRITEDATA, &resultBody);
curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, static_cast<size_t (__stdcall *)(char*, size_t, size_t, void*)>(
    [](char* ptr, size_t size, size_t nmemb, void* resultBody){
        *(static_cast<std::string*>(resultBody)) += std::string {ptr, size * nmemb};
        return size * nmemb;
    }
));

CURLcode curlResult = curl_easy_perform(curl);
std::cout << "RESULT BODY:\n" << resultBody << std::endl;
// Cleanup etc

Note the __stdcall cast is needed to comply to the C calling convention (cURL is a C library)

请注意，需要 __stdcall 强制转换以符合 C 调用约定（cURL 是 C 库）

回答by perreal

This might not work right away but should give you an idea:

这可能不会立即起作用，但应该给你一个想法：

#include <string>
#include <curl.h>
#include <stdio.h>
size_t write_data(void *ptr, size_t size, size_t nmemb, FILE *stream) {
    size_t written;
    written = fwrite(ptr, size, nmemb, stream);
    return written;
}

int main() {
    std::string tempname = "temp";
    CURL *curl;
    CURLcode res;
    curl = curl_easy_init();
    if(curl) {
      FILE *fp = fopen(tempname.c_str(),"wb");
      curl_easy_setopt(curl, CURLOPT_URL, "http://www.google.com");
      curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, write_data); 
      curl_easy_setopt(curl, CURLOPT_WRITEDATA, fp);
      res = curl_easy_perform(curl);
      curl_easy_cleanup(curl);
      fclose(fp);
      fp = fopen(tempname.c_str(),"rb");
      fseek (fp , 0 , SEEK_END);
      long lSize = ftell (fp);
      rewind(fp);
      char *buffer = new char[lSize+1];
      fread (buffer, 1, lSize, fp);
      buffer[lSize] = 0;
      fclose(fp);
      std::string content(buffer);
      delete [] buffer;
    }
}

回答by Uli K?hler

On my blog I have publisheda simple wrapper class to perform this task.

在我的博客上，我发布了一个简单的包装类来执行此任务。

Usage example:

用法示例：

#include "HTTPDownloader.hpp"

int main(int argc, char** argv) {
    HTTPDownloader downloader;
    std::string content = downloader.download("https://stackoverflow.com");
    std::cout << content << std::endl;
}

Here's the header file:

这是头文件：

/**
 * HTTPDownloader.hpp
 *
 * A simple C++ wrapper for the libcurl easy API.
 *
 * Written by Uli K?hler (techoverflow.net)
 * Published under CC0 1.0 Universal (public domain)
 */
#ifndef HTTPDOWNLOADER_HPP
#define HTTPDOWNLOADER_HPP

#include <string>

/**
 * A non-threadsafe simple libcURL-easy based HTTP downloader
 */
class HTTPDownloader {
public:
    HTTPDownloader();
    ~HTTPDownloader();
    /**
     * Download a file using HTTP GET and store in in a std::string
     * @param url The URL to download
     * @return The download result
     */
    std::string download(const std::string& url);
private:
    void* curl;
};

#endif  /* HTTPDOWNLOADER_HPP */

Here's the source code:

这是源代码：

/**
 * HTTPDownloader.cpp
 *
 * A simple C++ wrapper for the libcurl easy API.
 *
 * Written by Uli K?hler (techoverflow.net)
 * Published under CC0 1.0 Universal (public domain)
 */
#include "HTTPDownloader.hpp"
#include <curl/curl.h>
#include <curl/easy.h>
#include <curl/curlbuild.h>
#include <sstream>
#include <iostream>
using namespace std;

size_t write_data(void *ptr, size_t size, size_t nmemb, void *stream) {
    string data((const char*) ptr, (size_t) size * nmemb);
    *((stringstream*) stream) << data;
    return size * nmemb;
}

HTTPDownloader::HTTPDownloader() {
    curl = curl_easy_init();
}

HTTPDownloader::~HTTPDownloader() {
    curl_easy_cleanup(curl);
}

string HTTPDownloader::download(const std::string& url) {
    curl_easy_setopt(curl, CURLOPT_URL, url.c_str());
    /* example.com is redirected, so we tell libcurl to follow redirection */
    curl_easy_setopt(curl, CURLOPT_FOLLOWLOCATION, 1L);
    curl_easy_setopt(curl, CURLOPT_NOSIGNAL, 1); //Prevent "longjmp causes uninitialized stack frame" bug
    curl_easy_setopt(curl, CURLOPT_ACCEPT_ENCODING, "deflate");
    std::stringstream out;
    curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, write_data);
    curl_easy_setopt(curl, CURLOPT_WRITEDATA, &out);
    /* Perform the request, res will get the return code */
    CURLcode res = curl_easy_perform(curl);
    /* Check for errors */
    if (res != CURLE_OK) {
        fprintf(stderr, "curl_easy_perform() failed: %s\n",
                curl_easy_strerror(res));
    }
    return out.str();
}

回答by Jakub Koszuliński

Came out with useful, yet simple solution, which overloads std::ostream::operator<<

提出了有用但简单的解决方案，它重载了 std::ostream::operator<<

#include <ostream>

#include <curl/curl.h>

size_t curlCbToStream (
    char * buffer,
    size_t nitems,
    size_t size,
    std::ostream * sout
)
{
    *sout << buffer;

    return nitems * size;
}

std::ostream & operator<< (
    std::ostream & sout,
    CURL * request
)
{
    ::curl_easy_setopt(request, CURLOPT_WRITEDATA, & sout);
    ::curl_easy_setopt(request, CURLOPT_WRITEFUNCTION, curlCbToStream);
    ::curl_easy_perform(request);

    return sout;
}

Possible drawback of taken approach could be:

采取的方法的可能缺点可能是：

typedef void CURL;

That means it covers all known pointer types.

这意味着它涵盖了所有已知的指针类型。

回答by Stefan Rogin

Based on @JoachimIsaksson answer, here is a more verbose output that handles out-of-memory and has a limit for the maximum output from curl (as CURLOPT_MAXFILESIZElimits only based on header information and not on the actual size transferred ).

基于@JoachimIsaksson 的回答，这里有一个更详细的输出，它处理内存不足，并且对 curl 的最大输出有限制（因为CURLOPT_MAXFILESIZE限制仅基于头信息而不是实际传输的大小）。

#DEFINE MAX_FILE_SIZE = 10485760 //10 MiB

size_t curl_to_string(void *ptr, size_t size, size_t count, void *stream)
{
    if(((string*)stream)->size() + (size * count) > MAX_FILE_SIZE)
    {
        cerr<<endl<<"Could not allocate curl to string, output size (current_size:"<<((string*)stream)->size()<<"bytes + buffer:"<<(size * count) << "bytes) would exceed the MAX_FILE_SIZE ("<<MAX_FILE_SIZE<<"bytes)";
        return 0;
    }
    int retry=0;
    while(true)
    {
        try{
            ((string*)stream)->append((char*)ptr, 0, size*count);
            break;// successful
        }catch (const std::bad_alloc&) {
            retry++;
            if(retry>100)
            {
                cerr<<endl<<"Could not allocate curl to string, probably not enough memory, aborting after : "<<retry<<" tries at 10s apart";
                return 0;
            }
            cerr<<endl<<"Could not allocate curl to string, probably not enough memory, sleeping 10s, try:"<<retry;
            sleep(10);
        }
    }
  return size*count;
}

回答by Simog

I use Joachim Isaksson's answer with a modern C++ adaptation of CURLOPT_WRITEFUNCTION.

我将 Joachim Isaksson 的答案与CURLOPT_WRITEFUNCTION的现代 C++ 改编版结合使用。

No nagging by the compiler for C-style casts.

编译器不会对 C 风格的转换进行唠叨。

static auto WriteCallback(char* ptr, size_t size, size_t nmemb, void* userdata) -> size_t {
  static_cast<string*>(userdata)->append(ptr, size * nmemb);
  return size * nmemb;
}