dpull / lua-webclient Goto Github PK

我用nginx挂了一个test.php。里面就一句。 浏览器访问能正常显示“hello”字符串。但是skynet.call("webclient", "lua", "request", "http://localhost/test.php")的时候，会在webclient.lua的第37行(xpcall那一句)报错。

遇到了这个issues中同样的问题，加入链接-lcurl后仍然报错

#8
undefined symbol: curl_easy_escape

已解决，在这里补充一下，链接的时候要把-lcurl加在后面，像这样：
cc -g -O2 -Wall -Iskynet/3rd/lua -fPIC --shared -Iserver/lualib-src/webclient server/lualib-src/webclient.c -o server/luaclib/webclient.so -lcurl
如果是这样把链接库选项加在中间则会报错：
cc -g -O2 -Wall -Iskynet/3rd/lua -fPIC --shared -Iserver/lualib-src/webclient -lcurl server/lualib-src/webclient.c -o server/luaclib/webclient.so

附一个参考makefile：

include platform.mk

LUA_CLIB_PATH ?= server/luaclib

LUA_CLIB = webclient

CFLAGS = -g -O2 -Wall -I$(LUA_INC) $(MYCFLAGS)

# lua
LUA_STATICLIB := skynet/3rd/lua/liblua.a
LUA_LIB ?= $(LUA_STATICLIB)
LUA_INC ?= skynet/3rd/lua

all : \
    $(foreach v, $(LUA_CLIB), $(LUA_CLIB_PATH)/$(v).so)

$(LUA_CLIB_PATH)/webclient.so : server/lualib-src/webclient.c | $(LUA_CLIB_PATH)
	$(CC) $(CFLAGS) $(SHARED) -Iserver/lualib-src/webclient $^ -o $@ -lcurl

可关闭该issues或保留作为参考

skynet中使用, 报以下错误:
:00000018(1486429408.35)[2017-02-07 09:03:28]: lua loader error : error loading module 'webclient' from file './luaclib/webclient.so':
./luaclib/webclient.so: undefined symbol: curl_easy_escape
stack traceback:
[C]: in ?
[C]: in function 'require'
./mainserver/webclient.lua:7: in local 'main'
./skynet/lualib/loader.lua:48: in main chunk
:00000018(1486429408.35)[2017-02-07 09:03:28]: KILL self
:00000008(1486429408.35)[2017-02-07 09:03:28]: init service failed: ./skynet/lualib/skynet.lua:535: ./skynet/service/service_mgr.lua:25: ./skynet/lualib/skynet.lua:381: call failed
stack traceback:
[C]: in function 'assert'
./skynet/lualib/skynet.lua:535: in function 'skynet.uniqueservice'
./mainserver/main.lua:15: in local 'start'
./skynet/lualib/skynet.lua:619: in function <./skynet/lualib/skynet.lua:616>
[C]: in function 'xpcall'
./skynet/lualib/skynet.lua:623: in function 'skynet.pcall'
./skynet/lualib/skynet.lua:627: in function 'skynet.init_service'
./skynet/lualib/skynet.lua:640: in upvalue 'f'
./skynet/lualib/skynet.lua:104: in function <./skynet/lualib/skynet.lua:103>

具体是webclient_destory函数里的curl_multi_cleanup。
用skynet.abort关闭，有时候会报错。

➜ CardGameHall git:(develop) ✗ curl --version
curl 7.35.0 (x86_64-pc-linux-gnu) libcurl/7.35.0 OpenSSL/1.0.1f zlib/1.2.8 libidn/1.28 librtmp/2.3
Protocols: dict file ftp ftps gopher http https imap imaps ldap ldaps pop3 pop3s rtmp rtsp smtp smtps telnet tftp
Features: AsynchDNS GSS-Negotiate IDN IPv6 Largefile NTLM NTLM_WB SSL libz TLS-SRP

CURLOPT_POSTFIELDS 在7.17.1版本以后不会拷贝post数据

另外新增了一个选项:CURLOPT_COPYPOSTFIELDS

提此issue，希望后面使用这个库的同学不要掉坑了...

我将test.lua中urls改成https://api.mch.weixin.qq.com/pay/unifiedorder，运行出现Resolving timed out
https://api.mch.weixin.qq.com/pay/unifiedorder 0 Resolving timed out after 5000 milliseconds 0 -1.0 0

但是同一台机器用curl能返回结果：
curl -f https://api.mch.weixin.qq.com/pay/unifiedorder
<return_code></return_code><return_msg></return_msg>%

from: lgc

发起一个https请求，服务器没有返回content-length回导致你的库拿不到正确的结果
比如说这个网站：https://www.ddpkcc.com/

curl -I https://www.ddpkcc.com
HTTP/1.1 200 OK
Server: nginx/1.11.13
Date: Wed, 30 Aug 2017 09:32:58 GMT
Content-Type: text/html; charset=UTF-8
Connection: keep-alive
X-Powered-By: PHP/7.0.20

偶尔出现的情况

skynet中使用webclient
生产环境：
Linux VM-16-11-ubuntu 4.15.0-142-generic
curl 7.47.0 (x86_64-pc-linux-gnu) libcurl/7.47.0
OpenSSL 1.0.2g 1 Mar 2016

比较奇怪的地方：
libcurl 第三方库为啥会调用到 skynet-src/malloc_hook.c 的free方法

以下是报错堆栈：
Program terminated with signal SIGSEGV, Segmentation fault.
#0 je_san_check_stashed_ptrs (ptrs=0xfffffffffffff9c0, nstashed=nstashed@entry=200, usize=32)
at src/san.c:182
182 src/san.c: No such file or directory.
[Current thread is 1 (Thread 0x7fc0b9bef700 (LWP 28184))]
(gdb) where
#0 je_san_check_stashed_ptrs (ptrs=0xfffffffffffff9c0, nstashed=nstashed@entry=200, usize=32)
at src/san.c:182
#1 0x00000000004be120 in je_tcache_bin_flush_stashed (is_small=true, binind=2, cache_bin=0x7fc0b9bef048,
tcache=0x7fc0b9bef010, tsd=0x7fc0b9beecb8) at src/tcache.c:570
#2 tcache_bin_flush_bottom (small=true, rem=100, binind=2, cache_bin=0x7fc0b9bef048, tcache=0x7fc0b9bef010,
tsd=0x7fc0b9beecb8) at src/tcache.c:508
#3 je_tcache_bin_flush_small (tsd=tsd@entry=0x7fc0b9beecb8, tcache=tcache@entry=0x7fc0b9bef010,
cache_bin=0x7fc0b9bef048, binind=, rem=100) at src/tcache.c:529
#4 0x000000000044b87c in tcache_dalloc_small (slow_path=false, binind=, ptr=0x7fc265345f80,
tcache=, tsd=) at include/jemalloc/internal/tcache_inlines.h:157
#5 arena_dalloc (slow_path=false, caller_alloc_ctx=, tcache=,
ptr=0x7fc265345f80, tsdn=) at include/jemalloc/internal/arena_inlines_b.h:331
#6 idalloctm (slow_path=false, is_internal=false, alloc_ctx=, tcache=,
ptr=0x7fc265345f80, tsdn=) at include/jemalloc/internal/jemalloc_internal_inlines_c.h:120
#7 ifree (slow_path=false, tcache=, ptr=0x7fc265345f80, tsd=0x7fc0b9beecb8)
at src/jemalloc.c:2887
#8 je_free_default (ptr=0x7fc265345f80) at src/jemalloc.c:3014
#9 0x000000000041fded in free (ptr=0x7fc265345f80) at skynet-src/malloc_hook.c:215
#10 0x00007fc2d176e956 in ?? () from /usr/lib/x86_64-linux-gnu/libcurl.so.4
#11 0x00007fc2e0add6ba in start_thread (arg=0x7fc0b9bef700) at pthread_create.c:333
#12 0x00007fc2e00fe51d in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:109

麻烦问下这个库是否有微信昵称中文乱码的解决方案？

作者你好，目前的接口curl连接还是使用阻塞方式，这个能改成非阻塞模式吗？

在拉取依赖的lua版本号时，目标库的版本不存在，请更新依赖的lua版本号，谢谢。

在接入微信官方充值时，调用https请求报错， 如下： NSS: client certificate not found (nickname not specified)。

1.环境

系统：CentOS Linux release 7.2.1511 (Core) 
curl: curl 7.29.0 (x86_64-redhat-linux-gnu)
请求url: https://api.mch.weixin.qq.com/pay/unifiedorder

2.问题

然而在测试的时候（自己写curl例子测试）， 发现这个报错是可以忽略的（在NSS: client certificate not found 报错时，仍然会返回）。
经过定位，发现是 write_callback 时，只要有报错，就return 了：

static size_t write_callback(char* buffer, size_t block_size, size_t count, void* arg)
{
    ...
    //这里return 导致后面没有接受数据
    if (webrequest->error[0] != '\0')
        return length;
    ...
}

把这个write_callback和webclient_getrespond的 webrequest->errorr 判断修改后，成功获取到返回（这里我打开了CURLOPT_VERBOSE的调试输出）：

* About to connect() to api.mch.weixin.qq.com port 443 (#0)
*   Trying 140.207.69.102...
* Connected to api.mch.weixin.qq.com (140.207.69.102) port 443 (#0)
* Initializing NSS with certpath: sql:/etc/pki/nssdb
* skipping SSL peer certificate verification
* NSS: client certificate not found (nickname not specified)
* SSL connection using TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384
* Server certificate:
* 	subject: CN=payapp.weixin.qq.com,OU=R&D,O=Tencent Technology (Shenzhen) Company Limited,L=shenzhen,ST=guangdong,C=CN
* 	start date: Mar 22 00:00:00 2017 GMT
* 	expire date: Jun 21 23:59:59 2018 GMT
* 	common name: payapp.weixin.qq.com
* 	issuer: CN=GeoTrust SSL CA - G3,O=GeoTrust Inc.,C=US
> POST /pay/unifiedorder HTTP/1.1
Host: api.mch.weixin.qq.com
Accept: */*
Content-Length: 548
Content-Type: application/x-www-form-urlencoded

* upload completely sent off: 548 out of 548 bytes
< HTTP/1.1 200 OK
< Server: nginx
< Date: Sat, 01 Apr 2017 06:01:44 GMT
< Content-Type: text/plain
< Content-Length: 454
< Connection: keep-alive
< Keep-Alive: timeout=8
< 
* Connection #0 to host api.mch.weixin.qq.com left intact

返回如下
true	<xml><return_code><![CDATA[SUCCESS]]></return_code>
<return_msg><![CDATA[OK]]></return_msg>
...
<trade_type><![CDATA[APP]]></trade_type>
</xml>

如题，使用skynet，如果post的表格中有参数带有布尔值就直接把skynet干掉了。

服务器使用的是skynet框架，最近有时出现崩溃问题，不好重现
收集几个coredump，gdb里看到的都是在webclient_destory的时候触发崩溃
以下是gdb的栈：
(gdb) bt
#0 0x00007fdfdbc1988a in ?? ()
#1 0x00007fdfe942098f in ?? () from /lib/x86_64-linux-gnu/libcrypto.so.1.0.0
#2 0x00007fdfe941ed56 in ?? () from /lib/x86_64-linux-gnu/libcrypto.so.1.0.0
#3 0x00007fdfe942e440 in sk_pop_free () from /lib/x86_64-linux-gnu/libcrypto.so.1.0.0
#4 0x00007fdfe941f0dc in ENGINE_cleanup () from /lib/x86_64-linux-gnu/libcrypto.so.1.0.0
#5 0x00007fdfe01e352e in ?? () from /usr/lib/x86_64-linux-gnu/libcurl.so.4
#6 0x00007fdfe01e3ec9 in ?? () from /usr/lib/x86_64-linux-gnu/libcurl.so.4
#7 0x00007fdfe01bf695 in curl_global_cleanup () from /usr/lib/x86_64-linux-gnu/libcurl.so.4
#8 0x00007fdfe03fe47c in webclient_destory (l=0x7fdfe1b5ed08) at lualib-src/lua-webclient.c:87
#9 0x0000000000414739 in luaD_precall (L=L@entry=0x7fdfe1b5ed08, func=0x7fdfe1a79930, nresults=0) at ldo.c:434
#10 0x0000000000414a03 in luaD_call (L=L@entry=0x7fdfe1b5ed08, func=, nResults=) at ldo.c:498
#11 0x0000000000414a61 in luaD_callnoyield (L=0x7fdfe1b5ed08, func=, nResults=) at ldo.c:509
#12 0x0000000000413e6f in luaD_rawrunprotected (L=L@entry=0x7fdfe1b5ed08, f=0x416010 , ud=0x0) at ldo.c:142
#13 0x0000000000414d4d in luaD_pcall (L=0x7fdfe1b5ed08, func=, u=, old_top=48, ef=) at ldo.c:729
#14 0x0000000000415f1c in GCTM (L=L@entry=0x7fdfe1b5ed08, propagateerrors=propagateerrors@entry=0) at lgc.c:833
#15 0x00000000004175aa in callallpendingfinalizers (L=0x7fdfe1b5ed08) at lgc.c:872
#16 luaC_freeallobjects (L=0x7fdfe1b5ed08) at lgc.c:981
#17 0x000000000041bbce in close_state (L=0x7fdfe1b5ed08) at lstate.c:245
#18 0x000000000041c0c0 in lua_close (L=) at lstate.c:344
#19 0x00007fdfe7dfc7ac in snlua_release (l=0x7fdfe8b79470) at service-src/service_snlua.c:192
#20 0x000000000040abfa in delete_context (ctx=0x7fdfe0991400) at skynet-src/skynet_server.c:213
#21 skynet_context_release (ctx=ctx@entry=0x7fdfe0991400) at skynet-src/skynet_server.c:223
#22 0x000000000040b1fd in skynet_context_message_dispatch (sm=sm@entry=0x7fdfe8a168a0, q=0x7fdfe1b9c500, q@entry=0x0, weight=weight@entry=0) at skynet-src/skynet_server.c:348
#23 0x000000000040b87d in thread_worker (p=) at skynet-src/skynet_start.c:162
#24 0x00007fdfe9e3e184 in start_thread (arg=0x7fdfe2df0700) at pthread_create.c:312
#25 0x00007fdfe907d03d in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:111

请问这可能是什么问题？从哪个方向查？谢谢！

设置cURL允许执行的最长毫秒数无效，超过设定时间后没有返回

看了源码
if (top > 2 && lua_istable(l, 3)) {
http_post = webclient_tohttppost(l, 3);
if (!http_post)
return luaL_argerror(l, 3, "parameter post_form invalid");
}
else if (top > 2 && lua_isstring(l, 3)) {
postdata = lua_tolstring(l,3, &postdatalen);
}

if (top > 3 && lua_isnumber(l, 4)) {
    connect_timeout_ms = (long)lua_tointeger(l, 4);
    if (connect_timeout_ms < 0)
        return luaL_argerror(l, 4, "parameter connect_timeout_ms invalid");
}

源码中if (top > 2 && lua_istable(l, 3)) { 栈上第三个参数如果是table就设置headers，然后下面postdata也是栈上第三个参数，这里是否有误，我理解是url，headers，postdata，timeout。。。这样的排序，现在headers和data是不能同时设置的。

还有就是headers的table格式是什么样的，我这边设置是
local headers = {
["content_type"] = "application/json"
}
实际测试会出现Segmentation fault (core dumped)