HTTP标头错误报告内容长度时该怎么办 [英] What to do when http header wrongly reports content-length
问题描述
我正在尝试通过以下方式通过https下载网页:首先下载带有HEAD请求的标头，然后解析以获得Content-Length，然后使用Content-Length加上标头的一些空间来分配用于存储缓冲区的内存GET请求的结果.似乎stackoverflow.com提供的Content-Length太小，因此我的代码存在段错误.
I am trying to download web pages over https by first downloading the headers with a HEAD request, then parsing to obtain the Content-Length and then using the Content-Length plus some space for headers to allocate memory for a buffer to store results from a GET request. It seems that stackoverflow.com gives a Content-Length that is too small and thus my code segfaults.
我尝试遍历堆栈溢出的问题，以了解如何动态分配内存来处理那些错误报告其Content-Length但无法找到任何合适答案的页面.
I've tried looking through stack overflow past questions to see how to go about dynamically allocating memory to handle pages which misreport their Content-Length but haven't been able to find any suitable answers.
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <openssl/bio.h>
#include <openssl/ssl.h>
#include <openssl/err.h>
#define MAX_HEADER_SIZE 8192
/**
* Main SSL demonstration code entry point
*/
int main() {
char* host_and_port = "stackoverflow.com:443";
char* head_request = "HEAD / HTTP/1.1\r\nHost: stackoverflow.com\r\n\r\n";
char* get_request = "GET / HTTP/1.1\r\nHost: stackoverflow.com\r\n\r\n";
char* store_path = "mycert.pem";
char *header_token, *line_token, content_length_line[1024];
char *cmp = "\r\n";
char *html;
char *get;
int content_length;
size_t i = 0;
char buffer[MAX_HEADER_SIZE];
buffer[0] = 0;
BIO* bio;
SSL_CTX* ctx = NULL;
SSL* ssl = NULL;
/* initilise the OpenSSL library */
SSL_load_error_strings();
SSL_library_init();
ERR_load_BIO_strings();
OpenSSL_add_all_algorithms();
bio = NULL;
int r = 0;
/* Set up the SSL pointers */
ctx = SSL_CTX_new(TLS_client_method());
ssl = NULL;
r = SSL_CTX_load_verify_locations(ctx, store_path, NULL);
if (r == 0) {
fprintf(stdout,"Unable to load the trust store from %s.\n", store_path);
fprintf(stdout, "Error: %s\n", ERR_reason_error_string(ERR_get_error()));
fprintf(stdout, "%s\n", ERR_error_string(ERR_get_error(), NULL));
ERR_print_errors_fp(stdout);
}
/* Setting up the BIO SSL object */
bio = BIO_new_ssl_connect(ctx);
BIO_get_ssl(bio, &ssl);
if (!(ssl)) {
printf("Unable to allocate SSL pointer.\n");
fprintf(stdout, "Error: %s\n", ERR_reason_error_string(ERR_get_error()));
fprintf(stdout, "%s\n", ERR_error_string(ERR_get_error(), NULL));
ERR_print_errors_fp(stdout);
bio = NULL;
}
SSL_set_mode(ssl, SSL_MODE_AUTO_RETRY);
/* Attempt to connect */
BIO_set_conn_hostname(bio, host_and_port);
/* Verify the connection opened and perform the handshake */
if (BIO_do_connect(bio) < 1) {
fprintf(stdout, "Unable to connect BIO.%s\n", host_and_port);
fprintf(stdout, "Error: %s\n", ERR_reason_error_string(ERR_get_error()));
fprintf(stdout, "%s\n", ERR_error_string(ERR_get_error(), NULL));
ERR_print_errors_fp(stdout);
bio = NULL;
}
if (SSL_get_verify_result(ssl) != X509_V_OK) {
printf("Unable to verify connection result.\n");
fprintf(stdout, "Error: %s\n", ERR_reason_error_string(ERR_get_error()));
fprintf(stdout, "%s\n", ERR_error_string(ERR_get_error(), NULL));
ERR_print_errors_fp(stdout);
}
if (bio == NULL)
return (EXIT_FAILURE);
r = -1;
while (r < 0) {
r = BIO_write(bio, head_request, strlen(head_request));
if (r <= 0) {
if (!BIO_should_retry(bio)) {
printf("BIO_read should retry test failed.\n");
fprintf(stdout, "Error: %s\n", ERR_reason_error_string(ERR_get_error()));
fprintf(stdout, "%s\n", ERR_error_string(ERR_get_error(), NULL));
ERR_print_errors_fp(stdout);
continue;
}
/* It would be prudent to check the reason for the retry and handle
* it appropriately here */
}
}
r = -1;
while (r < 0) {
r = BIO_read(bio, buffer, MAX_HEADER_SIZE);
if (r == 0) {
printf("Reached the end of the data stream.\n");
fprintf(stdout, "Error: %s\n", ERR_reason_error_string(ERR_get_error()));
fprintf(stdout, "%s\n", ERR_error_string(ERR_get_error(), NULL));
ERR_print_errors_fp(stdout);
continue;
} else if (r < 0) {
if (!BIO_should_retry(bio)) {
printf("BIO_read should retry test failed.\n");
fprintf(stdout, "Error: %s\n", ERR_reason_error_string(ERR_get_error()));
fprintf(stdout, "%s\n", ERR_error_string(ERR_get_error(), NULL));
ERR_print_errors_fp(stdout);
continue;
}
/* It would be prudent to check the reason for the retry and handle
* it appropriately here */
}
};
printf("%s\r\n", buffer);
header_token = strtok(buffer, cmp);
while (header_token != NULL)
{
//printf ("header_token: %s\n\n", header_token);
if (strncmp(header_token, "Content-Length:", strlen("Content-Length:")) == 0
|| strncmp(header_token, "content-length:", strlen("content-length:")) == 0)
{
//printf ("header_token %s is equal to Content-Length:\n", header_token);
strcpy(content_length_line, header_token);
}
header_token = strtok(NULL, cmp);
}
if (strlen(content_length_line) > 0)
{
line_token = strtok(content_length_line, " ");
line_token = strtok(NULL, " ");
content_length = atoi(line_token);
printf ("Content-Length = %d\n", content_length);
}
//char get[content_length + MAX_HEADER_SIZE];
get = malloc((content_length + MAX_HEADER_SIZE)*sizeof(char));
if (get == NULL) {
fprintf(stdout, "Out of memory\n");
return (EXIT_FAILURE);
}
r = -1;
while (r < 0) {
r = BIO_write(bio, get_request, strlen(get_request));
if (r <= 0) {
if (!BIO_should_retry(bio)) {
printf("BIO_read should retry test failed.\n");
fprintf(stdout, "Error: %s\n", ERR_reason_error_string(ERR_get_error()));
fprintf(stdout, "%s\n", ERR_error_string(ERR_get_error(), NULL));
ERR_print_errors_fp(stdout);
continue;
}
/* It would be prudent to check the reason for the retry and handle
* it appropriately here */
}
}
r = -1;
while (r) {
while (r < 0) {
r = BIO_read(bio, buffer, 4096);
if (r == 0) {
printf("Reached the end of the data stream.\n");
fprintf(stdout, "Error: %s\n", ERR_reason_error_string(ERR_get_error()));
fprintf(stdout, "%s\n", ERR_error_string(ERR_get_error(), NULL));
ERR_print_errors_fp(stdout);
continue;
} else if (r < 0) {
if (!BIO_should_retry(bio)) {
printf("BIO_read should retry test failed.\n");
fprintf(stdout, "Error: %s\n", ERR_reason_error_string(ERR_get_error()));
fprintf(stdout, "%s\n", ERR_error_string(ERR_get_error(), NULL));
ERR_print_errors_fp(stdout);
continue;
}
/* It would be prudent to check the reason for the retry and handle
* it appropriately here */
}
};
printf("Received %d bytes\n",r);
printf("Received total of %ld bytes of %d\n", i+r, content_length);
memcpy(get+i, buffer, r);
i += r;
}
printf("%s\r\n", buffer);
/* clean up the SSL context resources for the encrypted link */
SSL_CTX_free(ctx);
free(get);
return (EXIT_SUCCESS);
}
我通常希望能够打印出完整的网页，但是由于错误的Content-Length，我得到了以下输出和段错误.
I would usually expect to be able to print out the full web page but because of the erroneous Content-Length I get the following output and segfault.
Received 1752 bytes
Received total of 248784 bytes of 105585
Program received signal SIGSEGV, Segmentation fault.
__memmove_sse2_unaligned_erms () at ../sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S:404
404 ../sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S: No such file or directory.
我应该如何处理内容长度不正确的页面?
How should I handle pages that give incorrect Content-Length?
推荐答案
对HEAD请求的响应中的Content-length不相关.仅包含实际正文的响应中的Content-length是相关的(即对GET，POST ...的响应).并且此Content-length应该用于读取HTTP正文，即，首先读取HTTP标头，确定长度，然后按指定的方式读取正文.即使可以读取更多数据，它们也不属于响应主体.
The Content-length in the response to a HEAD request is of no relevance. Only the Content-length in the response containing the actual body is relevant (i.e. response to GET, POST...). And this Content-length should be used to read the HTTP body, i.e. first read the HTTP header, determine the length and then read the body as specified. Even if more data could be read they don't belong to the response body.
除此之外，您正在执行HTTP/1.1请求.这意味着服务器可能使用Transfer-Encoding: chunked，在这种情况下，Content-length的值也无关紧要.相反，分块编码具有优先权，您需要根据每个给定块的长度读取主体的所有块.
Apart from that you are doing a HTTP/1.1 request. This means that the server might use Transfer-Encoding: chunked in which case the value of Content-length is irrelevant too. Instead chunked encoding takes preference and you need to read all the chunks of the body based on the length of each given chunk.
这篇关于HTTP标头错误报告内容长度时该怎么办的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
