具有相同请求标头的 BASH cURL 和 PHP cURL 的不同响应

Question

This is a kind of very interesting problem.这是一个非常有趣的问题。

I'm writing an API for a micro web radio player.我正在为微型网络广播播放器编写 API。 To get the current track I'm parsing the cast servers frontend.为了获得当前曲目，我正在解析演员服务器前端。

In the example of http://deepmix.ru , which is a SHOUTcast 1.x server, I'd able to parse the URI http://85.21.79.31:7128/played.html .在http://deepmix.ru的示例中，这是一个 SHOUTcast 1.x 服务器，我能够解析 URI http://85.21.79.31:7128/played.html 。

When I request the URI in FireFox I get the webpage displaying the list of played tracks.当我在 FireFox 中请求 URI 时，我得到显示播放曲目列表的网页。 If I request that URI with cURL from BASH of my server hosting my API I get a 404 of that server.如果我从托管 API 的服务器的 BASH 请求带有 cURL 的 URI，我会得到该服务器的 404。

$ curl -v -G http://85.21.79.31:7128/played.html
*   Trying 85.21.79.31...
* Connected to 85.21.79.31 (85.21.79.31) port 7128 (#0)
> GET /played.html HTTP/1.1
> Host: 85.21.79.31:7128
> User-Agent: curl/7.47.0
> Accept: */*
> 
ICY 404 Resource Not Found
icy-notice1:<BR>SHOUTcast Distributed Network Audio Server/Linux v1.9.8<BR>
icy-notice2:The resource requested was not found<BR>

* Connection #0 to host 85.21.79.31 left intact

I assumed a proper user agent may help and add Mozilla to receive the webpage.我认为适当的用户代理可能会有所帮助并添加Mozilla以接收网页。 So that worked.所以这奏效了。

$ curl -v -A "Mozilla" -G http://85.21.79.31:7128/played.html
*   Trying 85.21.79.31...
* Connected to 85.21.79.31 (85.21.79.31) port 7128 (#0)
> GET /played.html HTTP/1.1
> Host: 85.21.79.31:7128
> User-Agent: Mozilla
> Accept: */*
> 
* HTTP 1.0, assume close after body
< HTTP/1.0 200 OK
< content-type:text/html
< 
<HTML>[...]<title>SHOUTcast Administrator</title>[...]</HEAD><BODY topmargin=0 leftmargin=0 marginheight=0 marginwidth=0 bgcolor=#000000 text=#EEEEEE link=#001155 vlink=#001155 alink=#FF0000><font class=default><table width=100% border=0 cellpadding=0 cellspacing=0><tr><td height=50><font class=logoText>&nbsp;SHOUTcast Song History</font></td></tr><tr><td height=14 align=right><font class=ltv><a id=ltv href="http://www.shoutcast.com/">SHOUTcast Server Version 1.9.8/Linux</a>[...]</body></html>

According to my discoveries I transferred the request into my PHP cURL implementation.根据我的发现，我将请求转移到我的 PHP cURL 实现中。

$curlHandler = curl_init();
curl_setopt_array(
    $curlHandler,
    [
        CURLINFO_HEADER_OUT     => true,
        CURLOPT_URL             => 'http://85.21.79.31:7128/played.html',
        CURLOPT_CUSTOMREQUEST   => 'GET',
        CURLOPT_USERAGENT       => 'Mozilla',
        CURLOPT_RETURNTRANSFER  => true,
        CURLOPT_HEADER          => false,
        CURLOPT_NOBODY          => true
    ]
);
$response = curl_exec( $curlHandler );
$info     = curl_getinfo( $curlHandler );
curl_close( $curlHandler );
var_dump( $info );
var_dump( $response );

But I got the 404 response.但是我得到了 404 响应。

array(31) {
  ["request_header"]=>
  string(87) "GET /played.html HTTP/1.1
Host: 85.21.79.31:7128
User-Agent: Mozilla
Accept: */*

"
}

Warning: file_get_contents(http://85.21.79.31:7128/played.html): failed to open stream: HTTP request failed! ICY 404 Resource Not Found
 in /vagrant/src/Readers/CurrentTrackReader.php on line 50

While comparing the request headers there's no differences.在比较请求标头时，没有区别。 So I assume there's differences between BASH's and PHP's cURL implementations I cannot see.所以我假设我看不到 BASH 和 PHP 的 cURL 实现之间存在差异。

So what happens here?那么这里会发生什么呢？

Update (2019-07-31)更新 (2019-07-31)

Additonal runtime environment information其他运行时环境信息

OS操作系统

Linux hostname 4.19.0-0.bpo.5-amd64 #1 SMP Debian 4.19.37-4~bpo9+1 (2019-06-19) x86_64 GNU/Linux

PHP PHP

PHP Version 7.3.7-2+0~20190725.42+debian9~1.gbp848ca5

Update (2019-07-31)更新 (2019-07-31)

Additional information about cURL有关 cURL 的其他信息

BASH巴什

$ curl --version
curl 7.47.0 (x86_64-pc-linux-gnu) libcurl/7.47.0 GnuTLS/3.4.10 zlib/1.2.8 libidn/1.32 librtmp/2.3
Protocols: dict file ftp ftps gopher http https imap imaps ldap ldaps pop3 pop3s rtmp rtsp smb smbs smtp smtps telnet tftp 
Features: AsynchDNS IDN IPv6 Largefile GSS-API Kerberos SPNEGO NTLM NTLM_WB SSL libz TLS-SRP UnixSockets

phpinfo() phpinfo()

cURL support     | enabled
cURL Information | 7.52.1
Age              | 3

Features

AsynchDNS        | Yes
CharConv         | No
Debug            | No
GSS-Negotiate    | No
IDN              | Yes
IPv6             | Yes
krb4             | No
Largefile        | Yes
libz             | Yes
NTLM             | Yes
NTLMWB           | Yes
SPNEGO           | Yes
SSL              | Yes
SSPI             | No
TLS-SRP          | Yes
HTTP2            | Yes
GSSAPI           | Yes
KERBEROS5        | Yes
UNIX_SOCKETS     | Yes
PSL              | Yes
HTTPS_PROXY      | Yes
Protocols        | dict, file, ftp, ftps, gopher, http, https, imap, imaps, ldap, ldaps, pop3, pop3s, rtmp, rtsp, scp, sftp, smb, smbs, smtp, smtps, telnet, tftp
Host             | x86_64-pc-linux-gnu
SSL Version      | OpenSSL/1.0.2s
ZLib Version     | 1.2.8
libSSH Version   | libssh2/1.7.0


Directive   | Local Value | Master Value
curl.cainfo | no value    | no value

Answer 1

It's a very big fault of myself.这是我自己的一个很大的错误。

First I left that CURLOPT_NOBODY => true in the code (see the question) which leads to an empty response.首先，我在代码中留下了CURLOPT_NOBODY => true （请参阅问题），这会导致空响应。

Second I left file_get_contents( $uri ) in my code which leads to the warning posted in the question.其次，我在我的代码中留下了file_get_contents( $uri ) ，这导致了问题中发布的警告。

In conclusion while removing these both cURL works as expected.总之，在删除这些时，两个 cURL 都按预期工作。

具有相同请求标头的 BASH cURL 和 PHP cURL 的不同响应

问题描述

1 个解决方案

解决方案1
2 已采纳 2019-08-01 08:13:10

具有相同请求标头的 BASH cURL 和 PHP cURL 的不同响应

问题描述

1 个解决方案

解决方案1 2 已采纳 2019-08-01 08:13:10

解决方案1
2 已采纳 2019-08-01 08:13:10