HttpLuaModuleZh

= 名称 =

英文文档地址：http://wiki.nginx.org/HttpLuaModule

ngx_lua - Embed the power of Lua into Nginx

''此模块属第三方扩展，不在Nginx源码中. '' 安装介绍

= 状态 =

模块处于持续更新状态，并可用于生产环境.

= 版本 =

ngx_lua v0.3.1rc45 版本，最后更新 2012-01-05.

= 示例 = # 设置纯Lua扩展库PATH(';;' is the default path): lua_package_path '/foo/bar/?.lua;/blah/?.lua;;'; # 设置C扩展Lua模块，CPATH(can also use ';;'): lua_package_cpath '/bar/baz/?.so;/blah/blah/?.so;;'; server { location /inline_concat { # default_type指令设置默认MIME头: default_type 'text/plain'; set $a "hello"; set $b "world"; # 内嵌lua代码 set_by_lua $res "return ngx.arg[1]..ngx.arg[2]" $a $b; echo $res; }       location /rel_file_concat { set $a "foo"; set $b "bar"; # script path relative to nginx prefix # $ngx_prefix/conf/concat.lua contents: #           #    return ngx.arg[1]..ngx.arg[2] #           set_by_lua_file $res conf/concat.lua $a $b; echo $res; }       location /abs_file_concat { set $a "fee"; set $b "baz"; # absolute script path not modified set_by_lua_file $res /usr/nginx/conf/concat.lua $a $b; echo $res; }       location /lua_content { # MIME type determined by default_type: default_type 'text/plain'; content_by_lua "ngx.say('Hello,world!')" }        location /nginx_var { # MIME type determined by default_type: default_type 'text/plain'; # try access /nginx_var?a=hello,world content_by_lua "ngx.print(ngx.var['arg_a'], '\\n')"; }       location /request_body { # force reading request body (default off) lua_need_request_body on; client_max_body_size 50k; client_body_buffer_size 50k; content_by_lua 'ngx.print(ngx.var.request_body)'; }       # transparent non-blocking I/O in Lua via subrequests location /lua { # MIME type determined by default_type: default_type 'text/plain'; content_by_lua ' local res = ngx.location.capture("/some_other_location") if res.status == 200 then ngx.print(res.body) end'; }       # GET /recur?num=5 location /recur { # MIME type determined by default_type: default_type 'text/plain'; content_by_lua ' local num = tonumber(ngx.var.arg_num) or 0

if num > 50 then ngx.say("num too big") return end

ngx.say("num is: ", num) if num > 0 then res = ngx.location.capture("/recur?num=" .. tostring(num - 1)) ngx.print("status=", res.status, " ") ngx.print("body=", res.body) else ngx.say("end") end ';       }        location /foo { rewrite_by_lua ' res = ngx.location.capture("/memc",                   { args = { cmd = 'incr', key = ngx.var.uri } }                ) ';           proxy_pass http://blah.blah.com; }       location /blah { access_by_lua ' local res = ngx.location.capture("/auth") if res.status == ngx.HTTP_OK then return end if res.status == ngx.HTTP_FORBIDDEN then ngx.exit(res.status) end ngx.exit(ngx.HTTP_INTERNAL_SERVER_ERROR) ';           # proxy_pass/fastcgi_pass/postgres_pass/... }       location /mixed { rewrite_by_lua_file /path/to/rewrite.lua; access_by_lua_file /path/to/access.lua; content_by_lua_file /path/to/content.lua; }       # use nginx var in code path # WARN: contents in nginx var must be carefully filtered, # otherwise there'll be great security risk! location ~ ^/app/(.+) { content_by_lua_file /path/to/lua/app/root/$1.lua; }       location / { lua_need_request_body on; client_max_body_size 100k; client_body_buffer_size 100k; access_by_lua ' -- check the client IP addr is in our black list if ngx.var.remote_addr == "132.5.72.3" then ngx.exit(ngx.HTTP_FORBIDDEN) end -- check if the request body contains bad words if ngx.var.request_body and string.match(ngx.var.request_body, "fsck") then return ngx.redirect("/terms_of_use.html") end -- tests passed ';          # proxy_pass/fastcgi_pass/etc settings }   }

= Description =

This module embeds Lua, via the standard interpreter or LuaJIT, into Nginx and by leveraging Nginx subrequests, allows the integration of the powerful Lua threads ( Lua coroutines) into the Nginx event model.

Unlike Apache's mod_lua and Lighttpd's mod_magnet, Lua code executed using this module can be 100% non-blocking on network traffic as long as either of the ngx.location.capture or ngx.location.capture_multi interfaces are used to handle requests to upstream services such as mysql, postgresql, memcached, redis, or upstream http web services. (see HttpDrizzleModule, ngx_postgres, HttpMemcModule, HttpRedis2Module and HttpProxyModule modules for details).

The Lua interpreter or LuaJIT instance is shared across all the requests in a single nginx worker process but request contexts are segregated using lightweight Lua coroutines. Loaded Lua modules persist at the nginx worker process level resulting in a small memory footprint even when under heavy loads.

= Directives =

lua_code_cache
语法: lua_code_cache on | off

默认: lua_code_cache on

环境: main, server, location, location if

打开或关闭 Lua 代码缓存，影响以下指令： set_by_lua_file , content_by_lua_file, rewrite_by_lua_file, and access_by_lua_file, 以及强制加载或者reload Lua 模块等.

The Lua files referenced in set_by_lua_file, content_by_lua_file, access_by_lua_file, and rewrite_by_lua_file will not be cached at all, and Lua's  table will be cleared at the entry point of every request (such that Lua modules will not be cached either). With this in place, developers can follow the PHP way, i.e., edit-and-refresh.

Please note however, that Lua code inlined into nginx.conf such as those specified by set_by_lua, content_by_lua, access_by_lua, and rewrite_by_lua will always be cached because only the Nginx config file parser can correctly parse the file and the only ways to to reload the config file are to send a  signal or to restart Nginx.

The ngx_lua module currently does not support the "stat" mode like Apache's  does but this is planned for implementation in the future.

Disabling the Lua code cache is strongly discouraged for production use and should only be used during development as it has a significant impact on overall performance. In addition, race conditions when reloading Lua modules are common for concurrent requests when the code cache is disabled.

lua_regex_cache_max_entries
syntax: lua_regex_cache_max_entries 

default: lua_regex_cache_max_entries 1024

context: http

Specifies the maximum number of entries allowed in the worker process level compiled regex cache.

The regular expressions used in ngx.re.match, ngx.re.gmatch, ngx.re.sub, and ngx.re.gsub will be cached within this cache if the regex option  (i.e., compile-once flag) is specified.

The default number of entries allowed is 1024 and when this limit is reached, new regexes will not be cached (as if the  option was not specified) and there will be one, and only one, warning in the   file:

2011/08/27 23:18:26 [warn] 31997#0: *1 lua exceeding regex cache max entries (1024), ...

You should not activate the  option for regexes (and/or   string arguments for ngx.re.sub and ngx.re.gsub) that are generated on the fly and give rise to infinite variations to avoid hitting the specified limit.

lua_package_path
syntax: lua_package_path 

default: The content of LUA_PATH environ variable or Lua's compiled-in defaults.

context: main

Sets the Lua module search path used by scripts specified by set_by_lua, content_by_lua and others. The path string is in standard Lua path form, and can be used to stand for the original path.

lua_package_cpath
syntax: lua_package_cpath 

default: The content of LUA_CPATH environ variable or Lua's compiled-in defaults.

context: main

Sets the Lua C-module search path used by scripts specified by set_by_lua, content_by_lua and others. The cpath string is in standard Lua cpath form, and can be used to stand for the original cpath.

set_by_lua
syntax: set_by_lua $res  [$arg1 $arg2 ...]

context: main, server, location, server if, location if

phase: rewrite

Execute user code specified by  with input arguments , and set the script's return value to   in string form. The code in  can retrieve input arguments from the   table (index starts from   and increases sequentially) and the lua code may make API calls.

The set_by_lua directive is designed to execute short, fast running code blocks as the Nginx event loop is blocked during code execution. Time consuming code sequences should therefore be avoided.

Note that set_by_lua can only output a value to a single Nginx variable at a time but a workaround is possible by using the ngx.var.VARIABLE interface. For example,

location /foo { set $diff ''; # we have to predefine the $diff variable here set_by_lua $sum ' local a = 32 local b = 56 ngx.var.diff = a - b; -- write to $diff directly return a + b;         -- return the $sum value normally ';       echo "sum = $sum, diff = $diff"; }

This directive can be freely mixed with all the directives of HttpRewriteModule, HttpSetMiscModule, and HttpArrayVarModule. All of these directives will run in exactly the same order that they are written in the config file. For example,

set $foo 32; set_by_lua $bar 'tonumber(ngx.var.foo) + 1'; set $baz "bar: $bar"; # $baz == "bar: 33"

This directive requires the ngx_devel_kit module.

set_by_lua_file
syntax: set_by_lua_file $res  [$arg1 $arg2 ...]

context: main, server, location, server if, location if

phase: rewrite

Equivalent to set_by_lua, except that the file specified by  contains the lua code to be executed.

When the Lua code cache is on (default state), the user code is loaded once at the first request and cached and the Nginx config must be reloaded each time you modify the Lua source file. You can temporarily disable the Lua code cache during development by switching lua_code_cache  in your   to avoid reloading Nginx.

This directive requires the ngx_devel_kit module.

content_by_lua
syntax: content_by_lua 

context: location, location if

phase: content

Acts as a "content handler" and executes lua code string specified in  for every request. The lua code may make API calls and is executed as a new spawned coroutine in an independent global environment (i.e. a sandbox).

Do not use this directive and other content handler directives in the same location. For example, this directive and the proxy_pass directive should not be used in the same location.

content_by_lua_file
syntax: content_by_lua_file 

context: location, location if

phase: content

Equivalent to content_by_lua, except that the file specified by  contains the lua code to be executed.

Nginx variables can be used in the  string to provide flexibility. This however carries some risks and is not ordinarily recommended.

When the Lua code cache is on (default state), the user code is loaded once at the first request and cached and the Nginx config must be reloaded each time you modify the Lua source file. You can temporarily disable the Lua code cache during development by switching lua_code_cache  in your   to avoid reloading Nginx.

rewrite_by_lua
syntax: rewrite_by_lua 

context: http, server, location, location if

phase: post-rewrite

Acts as a rewrite phase handler and executes lua code string specified in  for every request. The Lua code may make API calls and is executed as a new spawned coroutine in an independent global environment (i.e. a sandbox).

Note that this handler always runs after the standard HttpRewriteModule. So the following will work as expected:

location /foo { set $a 12; # create and initialize $a set $b ""; # create and initialize $b rewrite_by_lua 'ngx.var.b = tonumber(ngx.var.a) + 1'; echo "res = $b"; }

because  and   run before rewrite_by_lua.

On the other hand, the following will not work as expected:

? location /foo { ?     set $a 12; # create and initialize $a ?     set $b ''; # create and initialize $b ?     rewrite_by_lua 'ngx.var.b = tonumber(ngx.var.a) + 1'; ?     if ($b = '13') { ?        rewrite ^ /bar redirect; ?        break; ?     }    ?    ?      echo "res = $b"; ? }

because  runs before rewrite_by_lua even if it is put after rewrite_by_lua in the config.

The right way of doing this is as follows:

location /foo { set $a 12; # create and initialize $a set $b ''; # create and initialize $b rewrite_by_lua ' ngx.var.b = tonumber(ngx.var.a) + 1 if tonumber(ngx.var.b) == 13 then return ngx.redirect("/bar"); end ';       echo "res = $b"; }

It is worth mentioning that, the  module can be approximately implemented by rewrite_by_lua. For example,

location / { eval $res { proxy_pass http://foo.com/check-spam; }       if ($res = 'spam') { rewrite ^ /terms-of-use.html redirect; }       fastcgi_pass ...; }

can be implemented in terms of  like this

location = /check-spam { internal; proxy_pass http://foo.com/check-spam; }   location / { rewrite_by_lua ' local res = ngx.location.capture("/check-spam") if res.body == "spam" then ngx.redirect("/terms-of-use.html") end ';       fastcgi_pass ...; }

Just as any other rewrite phase handlers, rewrite_by_lua also runs in subrequests.

Note that when calling  within a rewrite_by_lua handler, the nginx request processing control flow will still continue to the content handler. To terminate the current request from within a rewrite_by_lua handler, calling ngx.exit with status >= 200 and status < 300  for successful quits and   (or its friends) for failures.

If the HttpRewriteModule's rewrite directive is used to change the URI and initiate location re-lookups (internal redirections), then any rewrite_by_lua or rewrite_by_lua_file code sequences within the current location will not be executed. For example,

location /foo { rewrite ^ /bar; rewrite_by_lua 'ngx.exit(503)'; }   location /bar { ...   }

Here the Lua code  will never run. This will be the case if  is used as this will similarly initiate an internal redirection. If the  modifier is used instead, there will be no internal rediction and the rewrite_by_lua code will be executed.

rewrite_by_lua_file
syntax: rewrite_by_lua_file 

context: http, server, location, location if

phase: post-rewrite

Equivalent to rewrite_by_lua, except that the file specified by  contains the Lua code to be executed.

Nginx variables can be used in the  string to provide flexibility. This however carries some risks and is not ordinarily recommended.

When the Lua code cache is on (default state), the user code is loaded once at the first request and cached and the Nginx config must be reloaded each time you modify the Lua source file. You can temporarily disable the Lua code cache during development by switching lua_code_cache  in your   to avoid reloading Nginx.

access_by_lua
syntax: access_by_lua 

context: http, server, location, location if

phase: post-access

Acts as an access phase handler and executes lua code string specified in  for every request. The Lua code may make API calls and is executed as a new spawned coroutine in an independent global environment (i.e. a sandbox).

Note that this handler always runs after the standard HttpAccessModule. So the following will work as expected:

location / { deny   192.168.1.1; allow  192.168.1.0/24; allow  10.1.1.0/16; deny   all; access_by_lua ' local res = ngx.location.capture("/mysql", { ... }) ...       ';        # proxy_pass/fastcgi_pass/... }

That is, if a client IP address is in the blacklist, it will be denied before the MySQL query for more complex authentication is executed by access_by_lua.

It is worth mentioning that the  module can be approximated using access_by_lua:

location / { auth_request /auth; # proxy_pass/fastcgi_pass/postgres_pass/... }

can be implemented in terms of  like this

location / { access_by_lua ' local res = ngx.location.capture("/auth") if res.status == ngx.HTTP_OK then return end if res.status == ngx.HTTP_FORBIDDEN then ngx.exit(res.status) end ngx.exit(ngx.HTTP_INTERNAL_SERVER_ERROR) ';       # proxy_pass/fastcgi_pass/postgres_pass/... }

As with other access phase handlers, access_by_lua will not run in subrequests.

Note that when calling  within a access_by_lua handler, the nginx request processing control flow will still continue to the content handler. To terminate the current request from within a access_by_lua handler, calling ngx.exit with status >= 200 and status < 300  for successful quits and   (or its friends) for failures.

access_by_lua_file
syntax: access_by_lua_file 

context: http, server, location, location if

phase: post-access

Equivalent to access_by_lua, except that the file specified by  contains the Lua code to be executed.

Nginx variables can be used in the  string to provide flexibility. This however carries some risks and is not ordinarily recommended.

When the Lua code cache is on (default state), the user code is loaded once at the first request and cached and the Nginx config must be reloaded each time you modify the Lua source file. You can temporarily disable the Lua code cache during development by switching lua_code_cache  in your   to avoid reloading Nginx.

header_filter_by_lua
syntax: header_filter_by_lua 

context: http, server, location, location if

phase: output-header-filter

Uses Lua code specified in  to define an output header filter. Note that the following API functions are currently disabled within this context:


 * Output API functions (e.g., ngx.say and ngx.send_headers)
 * Control API functions (e.g., ngx.exit)
 * Subrequest API functions (e.g., ngx.location.capture and ngx.location.capture_multi)

Here is an example of overriding a response header (or adding one if absent) in our Lua header filter:

location / { proxy_pass http://mybackend; header_filter_by_lua 'ngx.header.Foo = "blah"'; }

This directive was first introduced in the  release.

header_filter_by_lua_file
syntax: header_filter_by_lua_file 

context: http, server, location, location if

phase: output-header-filter

Equivalent to header_filter_by_lua, except that the file specified by  contains the lua code to be executed.

This directive was first introduced in the  release.

lua_need_request_body
syntax: lua_need_request_body 

default: off

context: main | server | location

phase: depends on usage

Determines whether to force the request body data to be read before running rewrite/access/access_by_lua* or not. The Nginx core does not read the client request body by default and if request body data is required, then this directive should be turned  or the ngx.req.read_body function should be called within the Lua code.

To read the request body data within the $request_body variable, client_body_buffer_size must have the same value as client_max_body_size. Because when the content length exceeds client_body_buffer_size but less than client_max_body_size, Nginx will automatically buffer the data into a temporary file on the disk, which will lead to empty value in the $request_body variable.

If the current location includes rewrite_by_lua or rewrite_by_lua_file directives, then the request body will be read just before the rewrite_by_lua or rewrite_by_lua_file code is run (and also at the phase). Similarly, if only content_by_lua is specified, the request body will not be read until the content handler's Lua code is about to run (i.e., the request body will be read during the content phase).

It is recommended however, to use the ngx.req.read_body and ngx.req.discard_body functions for finer control over the request body reading process instead.

This also applies to access_by_lua and access_by_lua_file.

lua_shared_dict
syntax: lua_shared_dict 

default: no

context: main

phase: depends on usage

Declares a shared memory zone,, to serve as storage for the shm-based Lua dictionary.

The  argument accepts size units such as   and  :

http { lua_shared_dict dogs 10m; ...   }

See ngx.shared.DICT for details.

This directive was first introduced in the  release.

= Nginx API for Lua =

Introduction
The various  and   configuration directives serve as gateways to the Lua API within the   file. The Lua API described below can only be called within the user Lua code run in the context of these configuration directives.

The API is exposed to Lua in the form of two standard packages  and. These packages are in the default global scope within  and can also be introduced to external Lua modules by using the package.seeall option:

module("my_module", package.seeall)

function say(a) ngx.say(a) end

Alternatively, they can be imported to external Lua modules by using file scoped local Lua variables:

local ngx = ngx module("my_module")

function say(a) ngx.say(a) end

It is also possible to directly require the packages:

local ngx = require "ngx" local ndk = require "ndk"

The ability to require these packages was introduced in the  release.

Network I/O operations in user code should only be done through the Lua API calls as the Nginx event loop may be blocked and performance may drop off dramatically otherwise. Minor disk file operations may be done via Lua's standard  and   libraries but these should be avoided wherever possible as these also block the Nginx process. Delegating all network and disk I/O operations to Nginx subrequests (via the ngx.location.capture method and similar) is strongly recommended for maximum performance.

ngx.arg
语法: val = ngx.arg[index]

环境: set_by_lua*

当这个变量在set_by_lua或者set_by_lua_file内使用的时候是只读的，指的是在配置指令输入的参数:

value = ngx.arg[n]

下面是一个例子

location /foo { set $a 32; set $b 56; set_by_lua $res 'return tonumber(ngx.arg[1]) + tonumber(ngx.arg[2])' $a $b; echo $sum; }

与 的和是

When this table is used in the context of body_filter_by_lua or body_filter_by_lua_file, the first element holds the input data chunk to the output filter code and the second element holds the boolean flag for the "eof" flag indicating the end of the whole output data stream.

The data chunk and "eof" flag passed to the downstream Nginx output filters can also be overridden by assigning values directly to the corresponding table elements. When setting  or an empty Lua string value to , no data chunk will be passed to the downstream Nginx output filters at all.

ngx.var.VARIABLE
语法: ngx.var.VAR_NAME

环境: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

读写NGINX变量的值.

value = ngx.var.some_nginx_variable_name ngx.var.some_nginx_variable_name = value

需要注意的是你只能写NGINX已经定义的变量. 例如:

location /foo { set $my_var ''; # this line is required to create $my_var at config time content_by_lua ' ngx.var.my_var = 123; ...       ';    }

当然NGINX变量不能凭空来创造.

有一些NGINX变量比如： 和 能够被赋值,而有一些不能,比如：.

Nginx regex group capturing variables,  ,  , and etc, can be read by this interface as well, by writing,  ,  , and etc.

给变量 赋值为 ，将会取消NGINX的 变量.例如,

ngx.var.args = nil

注意 当读取nginx变量的时候,Ngnix会在每个请求的内存池中分配内存,当在请求结束的时候释放. 所以当你需要在Lua代码内重复的读取Nginx变量的时候,在Lua变量里缓存Ngnix的变量, 例如,

local val = ngx.var.some_var --- use the val repeatedly later

这样可以避免在当前请求的生命周期内内存的泄漏

Core constants
环境: init_by_lua*, set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*, body_filter_by_lua, *log_by_lua*, ngx.timer.*

ngx.OK (0) ngx.ERROR (-1) ngx.AGAIN (-2) ngx.DONE (-4) ngx.DECLINED (-5)

需要注意的是这三个常量只有在Nginx API for Lua使用 (例如, ngx.exit 可以接收,  , and   三个参数).

ngx.null

The  constant is a   light userdata usually used to represent nil values in Lua tables etc and is similar to the lua-cjson library's   constant. This constant was first introduced in the  release.

The  constant was first introduced in the   release.

HTTP method constants
环境: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

ngx.HTTP_GET ngx.HTTP_HEAD ngx.HTTP_PUT ngx.HTTP_POST ngx.HTTP_DELETE

这些常数经常在ngx.location.catpure和ngx.location.capture_multi方法中被调用.

HTTP status constants
context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

value = ngx.HTTP_OK (200) value = ngx.HTTP_CREATED (201) value = ngx.HTTP_SPECIAL_RESPONSE (300) value = ngx.HTTP_MOVED_PERMANENTLY (301) value = ngx.HTTP_MOVED_TEMPORARILY (302) value = ngx.HTTP_SEE_OTHER (303) value = ngx.HTTP_NOT_MODIFIED (304) value = ngx.HTTP_BAD_REQUEST (400) value = ngx.HTTP_UNAUTHORIZED (401) value = ngx.HTTP_FORBIDDEN (403) value = ngx.HTTP_NOT_FOUND (404) value = ngx.HTTP_NOT_ALLOWED (405) value = ngx.HTTP_GONE (410) value = ngx.HTTP_INTERNAL_SERVER_ERROR (500) value = ngx.HTTP_METHOD_NOT_IMPLEMENTED (501) value = ngx.HTTP_SERVICE_UNAVAILABLE (503) value = ngx.HTTP_GATEWAY_TIMEOUT (504) (first added in the v0.3.1rc38 release)

Nginx log level constants
context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

ngx.STDERR ngx.EMERG ngx.ALERT ngx.CRIT ngx.ERR ngx.WARN ngx.NOTICE ngx.INFO ngx.DEBUG

这些参数经常在ngx.log方法中被使用.

print
语法: print(...)

环境: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

于以 向 写日志.

它等价于一下内容：

ngx.log(ngx.NOTICE, 'lua print: ', a, b, ...)

LUA可以接收 参数，并且等价于字符串. 同样适用于ngx的布尔型的变量"true"和"false". ngx.null将会以字符串"null"输出.

NGINX有一个硬编码的错误，一条日志只能在2048个字节内. 如果多余这些个数，将会被截取. 如果想要修改这个限制，需要修改NGINX的源代码，位于src/core/ngx_log.h的NGX_MAX_ERROR_STR宏.

ngx.ctx
环境: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*, body_filter_by_lua, log_by_lua*, ngx.timer.*

这个table可以在每个请求用于保存Lua的上下文信息，并且在当前的请求生命周期内一直有效(类似于Ngnix的变量).

考虑下面这个例子,

location /test { rewrite_by_lua ' ngx.say("foo = ", ngx.ctx.foo) ngx.ctx.foo = 76 ';       access_by_lua ' ngx.ctx.foo = ngx.ctx.foo + 3 ';       content_by_lua ' ngx.say(ngx.ctx.foo) ';   }

执行 得出如下结果：

foo = nil 79

这说明 在rewrite, access, and content阶段一直有效.

每一个请求,包括子请求,都有自己的变量拷贝. 例如:

location /sub { content_by_lua ' ngx.say("sub pre: ", ngx.ctx.blah) ngx.ctx.blah = 32 ngx.say("sub post: ", ngx.ctx.blah) ';   }    location /main { content_by_lua ' ngx.ctx.blah = 73 ngx.say("main pre: ", ngx.ctx.blah) local res = ngx.location.capture("/sub") ngx.print(res.body) ngx.say("main post: ", ngx.ctx.blah) ';   }

执行 输出：

main pre: 73 sub pre: nil sub post: 32 main post: 73

这里需要注意的是 是子请求，但是不影响父请求的值. 这是因为他们有独立的.

Internal redirection will destroy the original request  data (if any) and the new request will have an empty   table. For instance, 内部重定向将会破坏主请求的 值，并且新的请求的 内容为nil，例如：

location /new { content_by_lua ' ngx.say(ngx.ctx.foo) ';   }    location /orig { content_by_lua ' ngx.ctx.foo = "hello" ngx.exec("/new") ';   }

执行 会得到

nil

而不是主请求的 这个值.

Arbitrary data values, including Lua closures and nested tables, can be inserted into this "magic" table. It also allows the registration of custom meta methods.

也支持lua的table,例如,

ngx.ctx = { foo = 32, bar = 54 }

ngx.location.capture
语法: res = ngx.location.capture(uri, options?)

环境: rewrite_by_lua*, access_by_lua*, content_by_lua*

Issue a synchronous but still non-blocking Nginx Subrequest using.

同步发出，又非阻塞的发起 的Nginx子请求

Nginx subrequests provide a powerful way to make non-blocking internal requests to other locations configured with disk file directory or any other nginx C modules like,  ,  , ,, and even   itself and etc etc etc.

需要注意的是子请求都是模仿HTTP协议，但是并没有额外的HTTP或者IPC流量开销，每一个都是在内部、高效的、C级别的

子请求完全不同于HTTP 301/302 重定向 (via ngx.redirect) 和内部重定向 (via ngx.exec).

这里有一个例子:

res = ngx.location.capture(uri)

返回的是一个Lua的table，包括三个成员 (, , and  ).

包括整个子请求返回的头部信息，它也是一个table. 对于多值相应头,这个值是lua的一个table，按照他们出现的先后顺序包括所有的值，如果子请求的相应头包含如下信息:

Set-Cookie: a=3 Set-Cookie: foo=bar Set-Cookie: baz=blah

那么  将会等价于 table.

URI query strings can be concatenated to URI itself, 例如,

res = ngx.location.capture('/foo/bar?a=3&b=4')

Named locations like  are not allowed due to a limitation in the nginx core. Use normal locations combined with the  directive to prepare internal-only locations.

An optional option table can be fed as the second argument, which support the options:

第二个参数是一个可选的table，它支持如下选项：


 * 指定子请求的请求方法, 目前只支持类似  这样的常数.
 * 指定子请求的请求方法, 目前只支持类似  这样的常数.


 * 指定子请求的body (字符串).
 * 指定子请求的body (字符串).


 * 指定子请求的URI参数 (both string value and Lua tables are accepted)
 * 指定子请求的URI参数 (both string value and Lua tables are accepted)


 * 给子请求指定的 Lua table ngx.ctx . It can be the current request's ngx.ctx table, which effectively make the parent and its subrequest to share exactly the same context table. This option was first introduced in the   release.
 * 给子请求指定的 Lua table ngx.ctx . It can be the current request's ngx.ctx table, which effectively make the parent and its subrequest to share exactly the same context table. This option was first introduced in the   release.


 * take a Lua table which holds the values to set the specified Nginx variables in the subrequest as this option's value. This option was first introduced in the  release.
 * take a Lua table which holds the values to set the specified Nginx variables in the subrequest as this option's value. This option was first introduced in the  release.


 * 指定是否拷贝当前请求的NGINX变量到子请求. 如果在子请求修改NGINX变量不会影响父请求的值. 在 版本后生效.
 * 指定是否拷贝当前请求的NGINX变量到子请求. 如果在子请求修改NGINX变量不会影响父请求的值. 在 版本后生效.


 * 指定是否共享NGINX变量到子请求. 如果在子请求修改NGINX变量，会影响父请求（当前请求）的值.
 * 指定是否共享NGINX变量到子请求. 如果在子请求修改NGINX变量，会影响父请求（当前请求）的值.

如果想要发出一个POST请求，可以参考：

res = ngx.location.capture(       '/foo/bar',        { method = ngx.HTTP_POST, body = 'hello, world' }    )

See HTTP method constants methods other than POST. 这个 默认是.

可以指定额外的URI参数, 例如,

ngx.location.capture('/foo?a=1',       { args = { b = 3, c = ':' } }    )

is equivalent to

ngx.location.capture('/foo?a=1&b=3&c=%3a')

that is, this method will automatically escape argument keys and values according to URI rules and concatenating them together into a complete query string. The format for the Lua table passed as the  argument is identical to the format used in the ngx.encode_args method.

这个选项也可以携带一些字符串的值:

ngx.location.capture('/foo?a=1',       { args = 'b=3&c=%3a' } }    )

这个功能上等同于上一个例子.

控制子请求是否共享父请求的变量.如果设置为, 那么当前的请求和相关的子请求将共享相同的Nginx的变量. 当然, 在子请求修改了变量，父请求也会跟着受到影响.

注意要谨慎使用这个功能，因为它可能引起很多意想不到的后果. ,, 或者   一般相对于 比较好.

这个选项默认是设置成  的.

location /other { set $dog "$dog world"; echo "$uri dog: $dog"; }                                        location /lua { set $dog 'hello'; content_by_lua ' res = ngx.location.capture("/other",               { share_all_vars = true }); ngx.print(res.body) ngx.say(ngx.var.uri, ": ", ngx.var.dog) ';     }

请求  会输出:

/other dog: hello world /lua: hello world

提供一份NGINX变量的副本给子请求. 子请求修改NGINX变量，不会影响父请求或者其他分享父请求变量的其他子请求的NGINX变量.

location /other { set $dog "$dog world"; echo "$uri dog: $dog"; }

location /lua { set $dog 'hello'; content_by_lua ' res = ngx.location.capture("/other",               { copy_all_vars = true });

ngx.print(res.body) ngx.say(ngx.var.uri, ": ", ngx.var.dog) ';   }

请求  会输出

/other dog: hello world /lua: hello

需要注意的是如果  和   都设置为 true, 那么   生效.

除了上述两个设置, 用 也可以向子请求赋值. These variables are set after the sharing or copying of variables has been evaluated, and provides a more efficient method of passing specific values to a subrequest over encoding them as URL arguments and unescaping them in the Nginx config file.

location /other { content_by_lua ' ngx.say("dog = ", ngx.var.dog) ngx.say("cat = ", ngx.var.cat) ';   }

location /lua { set $dog ''; set $cat ''; content_by_lua ' res = ngx.location.capture("/other",               { vars = { dog = "hello", cat = 32 }});

ngx.print(res.body) ';   }

执行  得出:

dog = hello cat = 32

能够用于提供一个LUA TABLE，类似 ngx.ctx 给子请求.

location /sub { content_by_lua ' ngx.ctx.foo = "bar"; ';   }       location /lua { content_by_lua ' local ctx = {} res = ngx.location.capture("/sub", { ctx = ctx }) ngx.say(ctx.foo); ngx.say(ngx.ctx.foo); ';                                           }

执行  这个请求的结果是

bar nil

这个参数  也能够用于分享相同的 ngx.ctx table 给子请求:

location /sub { content_by_lua ' ngx.ctx.foo = "bar"; ';   }    location /lua { content_by_lua ' res = ngx.location.capture("/sub", { ctx = ngx.ctx }) ngx.say(ngx.ctx.foo); ';   }

请求  输出:

bar

Note that subrequests issued by ngx.location.capture inherit all the request headers of the current request by default and that this may have unexpected side effects on the subrequest responses. For example, when using the standard  module to serve subrequests, an "Accept-Encoding: gzip" header in the main request may result in gzipped responses that cannot be handled properly in Lua code. Original request headers should be ignored by setting proxy_pass_request_headers to  in subrequest locations.

Please also refer to restrictions on capturing locations that include Echo Module directives.

ngx.location.capture_multi
语法: res1, res2, ... = ngx.location.capture_multi({ {uri, options?}, {uri, options?}, ... })

环境: rewrite_by_lua*, access_by_lua*, content_by_lua*

类似 ngx.location.capture, 但是支持并行多个子请求.

这个功能是在table表中平行的指定几个子请求，然后按照相同的顺序返回这些请求. 例如：

res1, res2, res3 = ngx.location.capture_multi{ { "/foo", { args = "a=3&b=4" } }, { "/bar" }, { "/baz", { method = ngx.HTTP_POST, body = "hello" } }, }   if res1.status == ngx.HTTP_OK then ...   end if res2.body == "BLAH" then ...   end

这个功能直到所有的子请求结束才返回数据​​. 总延时是子请求中单个的最大延时时间，而不是总共的延时总合.

Lua tables can be used for both requests and responses when the number of subrequests to be issued is not known in advance:

-- construct the requests table local reqs = {} table.insert(reqs, { "/mysql" }) table.insert(reqs, { "/postgres" }) table.insert(reqs, { "/redis" }) table.insert(reqs, { "/memcached" }) -- issue all the requests at once and wait until they all return local resps = { ngx.location.capture_multi(reqs) } -- loop over the responses table for i, resp in ipairs(resps) do       -- process the response table "resp" end

ngx.location.capture 仅仅是这个方法的一个特殊形式.逻辑上来说, ngx.location.capture 也可以这样实现

ngx.location.capture = function (uri, args) return ngx.location.capture_multi({ {uri, args} }) end

请同时参阅以上的 capturing locations that include Echo Module directives.

ngx.status
环境: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

读或者写当前请求的相应状态. 必须在输出相应头之前被调用.

ngx.status = ngx.HTTP_CREATED status = ngx.status 如果在相应之后调用的话，对输出没有影响不错会记录一条日志. attempt to set ngx.status after sending out response headers

ngx.header.HEADER
语法: ngx.header.HEADER = VALUE

语法: value = ngx.header.HEADER

环境: rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

添加、修改或者清空当前请求相应头的. 下划线 会被替换为横线 并且不区分大小写.

-- equivalent to ngx.header["Content-Type"] = 'text/plain' ngx.header.content_type = 'text/plain'; ngx.header["X-My-Header"] = 'blah blah';

当需要赋予多个值的时候，可以这样使用:

ngx.header['Set-Cookie'] = {'a=32; path=/', 'b=4; path=/'}

在相应头相会产出如下内容:

Set-Cookie: a=32; path=/ Set-Cookie: b=4; path=/

Only Lua tables are accepted (因为类似  只能接受一个值，所以只有在table中最后被赋值的内容才生效).

ngx.header.content_type = {'a', 'b'}

等价于

ngx.header.content_type = 'b'

给一个值设置为  时,相当于将这个值从header中移除:

ngx.header["X-My-Header"] = nil;

这相当于赋值一个空的table:

ngx.header["X-My-Header"] = {};

发送相应头之后在给  赋值 (either explicitly with ngx.send_headers or implicitly with ngx.print and similar) 将会抛出一个LUA的异常.

Reading  will return the value of the response header named.

下划线 也会被替换为横线  并且header的名字不区分大小写. 如果查看的相应的头的不存在, 那么将返回.

This is particularly useful in the context of filter_header_by_lua and filter_header_by_lua_file, for example,

location /test { set $footer '';

proxy_pass http://some-backend;

header_filter_by_lua ' if ngx.header["X-My-Header"] == "blah" then ngx.var.footer = "some value" end ';

echo_after_body $footer; }

如果headers里面有多个值, 所有的值将会按照顺序被收集并且以 Lua table 的形式返回. 例如, 相应的headers

Foo: bar Foo: baz

当读取  时返回的结果是

{"bar", "baz"}

需要注意的是  不是一个普通的 Lua table, 所以也不能用Lua的   方法遍历.

当需要读取 request 的headers, 用 ngx.req.get_headers 方法替换.

ngx.req.set_uri
语法: ngx.req.set_uri(uri, jump?)

环境: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

通过  这个参数，修改当前请求的URI. 这个参数必须是Lua的长度大于0的字符串，否则的话会抛出不可预料的异常.

布尔型的参数 能够像HttpRewriteModule的rewrite一样触发location跳转. 并且当 设置成为 (默认是 )的时候, 这个函数将不会返回, 并在在 阶段之后重新匹配并跳转到新的URI.当然你也发现，在 之后执行这个方法，页面不会跳转，只是修改了uri而已

默认的情况是只有当前的URI被修改，否则Location不会跳转. 当 参数是 或者不存在的时候,这个方法会返回但是没有返回值. 例如, 以下的NGINX配置文件片段

rewrite ^ /foo last;

可以用LUA这样实现:

ngx.req.set_uri("/foo", true)

当然，NGINX的这种配置

rewrite ^ /foo break;

也可以这样实现

ngx.req.set_uri("/foo", false)

或者等价于这样,

ngx.req.set_uri("/foo")

参数  只能够在rewrite_by_lua 和 rewrite_by_lua_file内设置为 , 否则会抛出不可预知的异常错误.

这里有一个关于正则表达式替换的更复杂的例子如下

location /test { rewrite_by_lua ' local uri = ngx.re.sub(ngx.var.uri, "^/test/(.*)", "$1", "o") ngx.req.set_uri(uri) ';       proxy_pass http://my_backend; }

等价于

location /test { rewrite ^/test/(.*) /$1 break; proxy_pass http://my_backend; }

注意的是你不能使用这个接口来修改URI参数，你只能使用ngx.req.set_uri_args完成这个功能，例如 rewrite ^ /foo?a=3? last;

can be coded as

ngx.req.set_uri_args("a=3") ngx.req.set_uri("/foo", true)

或者

ngx.req.set_uri_args({a = 3}) ngx.req.set_uri("/foo", true)

This interface was first introduced in the  release.

ngx.req.set_uri_args
语法: ngx.req.set_uri_args(args)

环境: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

根据 参数重新定义当前请求的URI参数. 参数 可以是Lua字符串, 比如

ngx.req.set_uri_args("a=3&b=hello%20world")

也可以是LUA TABLE

ngx.req.set_uri_args({ a = 3, b = "hello world" })

在后一种情况下, 这个方法会自动根据URI规则转义TABLE内的key和value.

This interface was first introduced in the  release.

也可参考 ngx.req.set_uri.

ngx.req.get_uri_args
语法: args = ngx.req.get_uri_args

环境: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

返回一个LUA TABLE，包含当前请求的全部的URL参数

这里有一个例子,

location = /test { content_by_lua ' local args = ngx.req.get_uri_args for key, val in pairs(args) do               if type(val) == "table" then ngx.say(key, ": ", table.concat(val, ", ")) else ngx.say(key, ": ", val) end end ';   }

那么 将会返回

foo: bar bar: baz, blah

Multiple occurrences of an argument key will result in a table value holding all of the values for that key in order.

参数和值将会根据URI的规则自动转义. 例如, 在上面的例子中,  将会输出

a b: 1a 2

当参数没有 的时候，将会按照布尔型参数处理. 例如, 将会输出

foo: true bar: true

可见，是LUA布尔型的. 然而, 他们不同于值为空的情况. 例如,  相会返回

foo: bar:

当key是空的时候，将不做处理, 例如,  什么都不会输出.

通过修改NGINX的变量 (或者在LUA内修改变量 ) 也会在返回值内生效:

ngx.var.args = "a=3&b=42" local args = ngx.req.get_uri_args

不管实际请求的字符串是什么，这里的 将会生效

{a = 3, b = 42}

ngx.req.get_post_args
语法: ngx.req.get_post_args

环境: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

返回一个LUA TABLE，包括所有当前请求的POST参数(of the MIME type ). 这个方法依赖于通过先调用ngx.req.read_body方法先读取请求的body 或者打开lua_need_request_body指令(设置lua_need_request_body为on), 否则将会抛出异常错误.

这里有一个例子,

location = /test { content_by_lua ' ngx.req.read_body local args = ngx.req.get_post_args for key, val in pairs(args) do               if type(val) == "table" then ngx.say(key, ": ", table.concat(val, ", ")) else ngx.say(key, ": ", val) end end ';   }

那么

# Post request with the body 'foo=bar&bar=baz&bar=blah' $ curl --data 'foo=bar&bar=baz&bar=blah' localhost/test

将会得到

foo: bar bar: baz, blah

当POST包含多个参数的时候，返回的TABLE将会按照KEY给他们排序. 当然内容会根据URI规则重新转义，例如在上面的情况下

# POST request with body 'a%20b=1%61+2' $ curl -d 'a%20b=1%61+2' localhost/test

将会输出

a b: 1a 2

当请求的URI内的参数后面没有 或者为空的情况，规则同ngx.req.get_uri_args

ngx.req.get_headers
语法: headers = ngx.req.get_headers

环境: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

返回一个包含多有当前请求头信息的lua table.

这里有个例子,

local h = ngx.req.get_headers for k, v in pairs(h) do       ... end

如果读取个别的请求头部信息:

ngx.say("Host: ", ngx.req.get_headers["Host"])

对于请求头信息包含多个事例,如:

Foo: foo Foo: bar Foo: baz

那么 则表示一个Lua (array) table,如下:

{"foo", "bar", "baz"}

另外一种读取单个请求头部信息的方法是使用, 它是nginx的标准变量$http_HEADER.

ngx.req.set_header
语法: ngx.req.set_header(header_name, header_value)

环境: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

给当前请求的头部信息名称是 的内容附值为, 可以覆盖任何现有的信息.当前请求的子请求不会受到影响.

这里是一个给头部信息 附值的例子:

ngx.req.set_header("Content-Type", "text/css")

也可以给 设置一个数组的内容, 例如,

ngx.req.set_header("Foo", {"a", "abc"})

将会产生两个值:

Foo: a   Foo: abc

如果给 附值是 的话,那这个头部信息会被移出. 所以

ngx.req.set_header("X-Foo", nil)

等价于

ngx.req.clear_header("X-Foo")

ngx.req.read_body
语法: ngx.req.read_body

环境: rewrite_by_lua*, access_by_lua*, content_by_lua*

在不阻塞ngnix其他事件的情况下同步读取客户端的body信息

如果因为开启了lua_need_request_body而已经读取的请求的body,那么这个方法不会做任何操作,立刻返回信息 如果因为开启了ngx.req.discard_body那么这个方法也不会做任何操作,立刻返回

如果因为连接失败或者读取内容失败, 这个方法会立刻抛出lua异常,500错误.

你也可以过一会儿通过ngx.req.get_body_data获取请求的body信息,获取获取通过ngx.req.get_body_file获取缓存在磁盘上的临时文件名称,这依赖于
 * 1) 当前请求的body大小是否已经超过了你设置的client_body_buffer_size大小,
 * 2) 和你是否开启了 client_body_in_file_only.

In case that you do not want to read the request body and the current request may have a request body, then it is crucial to use the ngx.req.discard_body function to explicitly discard the request body, or you'll break HTTP 1.1 keepalive and HTTP 1.1 pipelining. 如果你不想要去读请求的body并且当前请求的有body的情况下,那么开启ngx.req.discard_body明确放弃读取请求主体非常重要,否则你将破坏HTTP 1.1 长连接和 HTTP 1.1 流水线.

这里是个小例子:

ngx.req.read_body local args = ngx.req.get_post_args

This function was first introduced in the  release.

ngx.req.discard_body
syntax: ngx.req.discard_body

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Explicitly discard the request body, i.e., read the data on the connection and throw it away immediately. Please note that, simply ignoring request body is not the right way to discard it, you need to call this function, or you'll break things under HTTP 1.1 keepalive or HTTP 1.1 pipelining.

This function is an asynchronous call and returns immediately.

If the request body has already been read, this function does nothing and returns immediately.

This function was first introduced in the  release.

See also ngx.req.read_body.

ngx.req.get_body_data
syntax: data = ngx.req.get_body_data

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Retrieves the in-memory request body data. It returns a Lua string rather than a Lua table holding all the parsed query arguments. If you want the latter, use ngx.req.get_post_args instead.

This function returns  if
 * 1) the request body has not been read,
 * 2) the request body has been read into disk temporary files,
 * 3) or the request body has zero size.

If the request body has not been read yet, call ngx.req.read_body first (or turned on lua_need_request_body to force this module to read the request body automatically, but this is not recommended).

If the request body has been read into disk files, try calling the ngx.req.get_body_file function instead.

In case that you want to enforce in-memory request bodies, try setting client_body_buffer_size to the same size value in client_max_body_size.

Note that calling this function instead of using  or   is more efficient because it can save one dynamic memory allocation and one data copy.

This function was first introduced in the  release.

See also ngx.req.get_body_file.

ngx.req.get_body_file
syntax: file_name = ngx.req.get_body_file

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Retrieves the file name for the in-file request body data. Returns  if the request body has not been read or has been read into memory.

The returned file is read only and is usually cleaned up automatically by Nginx's memory pool. It should not be modified, renamed, or removed by your own Lua code.

If the request body has not been read yet, call ngx.req.read_body first (or turned on lua_need_request_body to force this module to read the request body automatically, but this is not recommended).

If the request body has been read into memory, try calling the ngx.req.get_body_data function instead.

In case that you want to enforce in-file request bodies, try turning on client_body_in_file_only.

This function was first introduced in the  release.

See also ngx.req.get_body_data.

ngx.req.set_body_data
syntax: ngx.req.set_body_data(data)

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Set the current request's request body using the in-memory data specified by the  argument.

If the current request's request body has not been read, then it will be properly discarded. When the current request's request body has been read into memory or buffered into a disk file, then the old request body's memory will be freed or the disk file will be cleaned up immediately, respectively.

This function requires patching the Nginx core to function properly because the Nginx core does not allow modifying request bodies by the current design. Here is a patch for Nginx 1.0.9: nginx-1.0.9-allow_request_body_updating.patch, and this patch should be applied cleanly to other releases of Nginx as well.

This patch has already been applied to ngx_openresty 1.0.8.17 and above.

This function was first introduced in the  release.

See also ngx.req.set_body_file.

ngx.req.set_body_file
syntax: ngx.req.set_body_file(file_name, auto_clean?)

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Set the current request's request body using the in-file data specified by the  argument.

If the optional  argument is given a   value, then this file will be automatically removed at request completion or the next time this function or ngx.req.set_body_data are called in the same request. The  is default to.

You must ensure that the file specified by the  argument exists and is readable by an Nginx worker process by setting its permission properly. Otherwise a Lua exception will be thrown.

If the current request's request body has not been read, then it will be properly discarded. When the current request's request body has been read into memory or buffered into a disk file, then the old request body's memory will be freed or the disk file will be cleaned up immediately, respectively.

This function requires patching the Nginx core to function properly because the Nginx core does not allow modifying request bodies by the current design. Here is a patch for Nginx 1.0.9: nginx-1.0.9-allow_request_body_updating.patch, and this patch should be applied cleanly to other releases of Nginx as well.

If you're using ngx_openresty 1.0.8.17+, then you've already had this patch applied.

This function was first introduced in the  release.

See also ngx.req.set_body_data.

ngx.req.clear_header
syntax: ngx.req.clear_header(header_name)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Clear the current request's request header named. None of the current request's subrequests will be affected.

ngx.exec
syntax: ngx.exec(uri, args?)

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Does an internal redirect to  with.

ngx.exec('/some-location'); ngx.exec('/some-location', 'a=3&b=5&c=6'); ngx.exec('/some-location?a=3&b=5', 'c=6');

Named locations are also supported, but query strings are ignored. For example,

location /foo { content_by_lua ' ngx.exec("@bar"); ';   }    location @bar { ...   }

The optional second  can be used to specify extra URI query arguments, for example:

ngx.exec("/foo", "a=3&b=hello%20world")

Alternatively, you can pass a Lua table for the  argument and let ngx_lua do URI escaping and string concatenation automatically for you, for instance,

ngx.exec("/foo", { a = 3, b = "hello world" })

The result is exactly the same as the previous example. The format for the Lua table passed as the  argument is identical to the format used in the ngx.encode_args method.

Note that this is very different from ngx.redirect in that it is just an internal redirect and no new HTTP traffic is involved.

This method never returns.

This method must be called before ngx.send_headers or explicit response body outputs by either ngx.print or ngx.say.

This method is similar to the echo_exec directive of the HttpEchoModule.

ngx.redirect
syntax: ngx.redirect(uri, status?)

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Issue an  redirection to.

The optional  parameter specifies whether or  to be used. It is   by default.

Here is an example assuming the current server name is  and that it is listening on Port 1984:

return ngx.redirect("/foo")

which is equivalent to

return ngx.redirect("http://localhost:1984/foo", ngx.HTTP_MOVED_TEMPORARILY)

We can also use the numberical code directly as the second  argument:

return ngx.redirect("/foo", 301)

This method must be called before ngx.send_headers or explicit response body outputs by either ngx.print or ngx.say.

This method never returns.

This method is very much like the rewrite directive with the  modifier in the standard HttpRewriteModule, for example, this  snippet

rewrite ^ /foo? redirect; # nginx config

is equivalent to the following Lua code

return ngx.redirect('/foo'); -- lua code

while

rewrite ^ /foo? permanent; # nginx config

is equivalent to

return ngx.redirect('/foo', ngx.HTTP_MOVED_PERMANENTLY) -- Lua code

URI arguments can be specified as well, for example:

return ngx.redirect('/foo?a=3&b=4')

ngx.send_headers
syntax: ngx.send_headers

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Explicitly send out the response headers.

Usually you do not have to send headers yourself. will automatically send out headers right before you output contents via ngx.say or ngx.print.

Headers will also be sent automatically when content_by_lua exits normally.

ngx.headers_sent
syntax: value = ngx.headers_sent

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Returns  if the response headers have been sent (by ngx_lua), and   otherwise.

This API was first introduced in ngx_lua v0.3.1rc6.

ngx.print
syntax: ngx.print(...)

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Emit arguments concatenated to the HTTP client (as response body). If response headers have not been sent, this function will send headers out first and then output body data.

Lua  values will output string   and Lua boolean values will output literal strings   and   respectively.

Nested arrays of strings are permitted and the elements in the arrays will be sent one by one:

local table = { "hello, ", {"world: ", true, " or ", false, {": ", nil}} }   ngx.print(table)

will yield the output

hello, world: true or false: nil

Non-array table arguments will cause a Lua exception to be thrown.

This is an asynchronous call and will return immediately without waiting for all the data to be written into the system send buffer. To run in synchronous mode, call  after calling. This can be particularly useful for streaming output. See ngx.flush for more details.

ngx.say
syntax: ngx.say(...)

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Just as ngx.print but also emit a trailing newline.

ngx.log
syntax: ngx.log(log_level, ...)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Log arguments concatenated to error.log with the given logging level.

Lua  arguments are accepted and result in literal , and Lua booleans result in literal   or   outputs.

The  argument can take constants like   and. Check out Nginx log level constants for details.

ngx.flush
syntax: ngx.flush(wait?)

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Flushes response output to the client.

accepts an optional boolean  argument (Default:  ) first introduced in the   release. When called with the default argument, it issues an asynchronous call (Returns immediately without waiting for output data to be written into the system send buffer). Calling the function with the  argument set to   switches to synchronous mode.

In synchronous mode, the function will not return until all output data has been written into the system send buffer or until the send_timeout setting has expired. Note that using the Lua coroutine mechanism means that this function does not block the Nginx event loop even in the synchronous mode.

When  is called immediately after   or , it causes the latter functions to run in synchronous mode. This can be particularly useful for streaming output.

Note that  is non functional when in the HTTP 1.0 output buffering mode. See HTTP 1.0 support.

ngx.exit
syntax: ngx.exit(status)

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

When  (i.e.,   and above), it will interrupt the execution of the current request and return status code to nginx.

When  (i.e.,  ), it will only quit the current phase handler (or the content handler if the content_by_lua directive is used) and continue to run later phases (if any) for the current request.

The  argument can be ,  ,  , , or other HTTP status constants.

To return an error page with custom contents, use code snippets like this:

ngx.status = ngx.HTTP_GONE ngx.say("This is our own content") -- to cause quit the whole request rather than the current phase handler ngx.exit(ngx.HTTP_OK)

The effect in action:

$ curl -i http://localhost/test HTTP/1.1 410 Gone Server: nginx/1.0.6 Date: Thu, 15 Sep 2011 00:51:48 GMT Content-Type: text/plain Transfer-Encoding: chunked Connection: keep-alive

This is our own content

Number literals can be used directly as the argument, for instance,

ngx.exit(501)

ngx.eof
syntax: ngx.eof

context: rewrite_by_lua*, access_by_lua*, content_by_lua*

Explicitly specify the end of the response output stream.

ngx.escape_uri
syntax: newstr = ngx.escape_uri(str)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Escape  as a URI component.

ngx.unescape_uri
syntax: newstr = ngx.unescape_uri(str)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Unescape  as an escaped URI component.

For example,

ngx.say(ngx.unescape_uri("b%20r56+7"))

gives the output

b r56 7

ngx.encode_args
syntax: str = ngx.encode_args(table)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Encode the Lua table to a query args string according to the URI encoded rules.

For example,

ngx.encode_args({foo = 3, ["b r"] = "hello world"})

yields

foo=3&b%20r=hello%20world

The table keys must be Lua strings.

Multi-value query args are also supported. Just use a Lua table for the arg's value, for example:

ngx.encode_args({baz = {32, "hello"}})

gives

baz=32&baz=hello

If the value table is empty and the effect is equivalent to the  value.

Boolean argument values are also supported, for instance,

ngx.encode_args({a = true, b = 1})

yields

a&b=1

If the argument value is, then the effect is equivalent to the   value.

This method was first introduced in the  release.

ngx.encode_base64
syntax: newstr = ngx.encode_base64(str)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Encode  to a base64 digest.

ngx.decode_base64
syntax: newstr = ngx.decode_base64(str)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Decodes the  argument as a base64 digest to the raw form. Returns  if   is not well formed.

ngx.crc32_short
syntax: intval = ngx.crc32_short(str)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Calculates the CRC-32 (Cyclic Redundancy Code) digest for the  argument.

This method performs better on relatively short  inputs (i.e., less than 30 ~ 60 bytes), as compared to ngx.crc32_long. The result is exactly the same as ngx.crc32_long.

Behind the scene, it is just a thin wrapper around the  function defined in the Nginx core.

This API was first introduced in the  release.

ngx.crc32_long
syntax: intval = ngx.crc32_long(str)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Calculates the CRC-32 (Cyclic Redundancy Code) digest for the  argument.

This method performs better on relatively long  inputs (i.e., longer than 30 ~ 60 bytes), as compared to ngx.crc32_short. The result is exactly the same as ngx.crc32_short.

Behind the scene, it is just a thin wrapper around the  function defined in the Nginx core.

This API was first introduced in the  release.

ngx.hmac_sha1
syntax: digest = ngx.hmac_sha1(secret_key, str)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Computes the HMAC-SHA1 digest of the argument  and turns the result using the secret key.

The raw binary form of the  digest will be generated, use ngx.encode_base64, for example, to encode the result to a textual representation if desired.

For example,

local key = "thisisverysecretstuff" local src = "some string we want to sign" local digest = ngx.hmac_sha1(key, src) ngx.say(ngx.encode_base64(digest))

yields the output

R/pvxzHC4NLtj7S+kXFg/NePTmk=

This API requires the OpenSSL library enabled in your Nignx build (usually by passing the  option to the   script).

This function was first introduced in the  release.

ngx.md5
syntax: digest = ngx.md5(str)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Returns the hexadecimal representation of the MD5 digest of the  argument.

For example,

location = /md5 { content_by_lua 'ngx.say(ngx.md5("hello"))'; }

yields the output

5d41402abc4b2a76b9719d911017c592

See also ngx.md5_bin if you want the raw binary MD5 digest.

ngx.md5_bin
syntax: digest = ngx.md5_bin(str)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Returns the binary form of the MD5 digest of the  argument.

See also ngx.md5 if you want the hexadecimal form of the MD5 digest.

ngx.today
syntax: str = ngx.today

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Returns current date (in the format ) from the nginx cached time (no syscall involved unlike Lua's date library).

This is the local time.

ngx.time
syntax: secs = ngx.time

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Returns the elapsed seconds from the epoch for the current timestamp from the nginx cached time (no syscall involved unlike Lua's date library).

You can enforce updating the Nginx time cache by calling ngx.update_time first.

ngx.now
syntax: secs = ngx.now

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Returns a floating-point number for the elapsed time in seconds (including microseconds as the decimal part) from the epoch for the current timestamp from the nginx cached time (no syscall involved unlike Lua's date library).

You can use the standard Nginx directive timer_resolution to adjust the accuracy or forcibly updating the Nginx time cache by calling ngx.update_time first.

This API was first introduced in.

ngx.update_time
syntax: ngx.update_time

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Forcibly updating the Nginx current time cache. This call involves a syscall and thus has some overhead, so do not abuse it.

This API was first introduced in.

ngx.localtime
syntax: str = ngx.localtime

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Returns the current timestamp (in the format ) of the nginx cached time (no syscall involved unlike Lua's os.date function).

This is the local time.

ngx.utctime
syntax: str = ngx.utctime

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Returns the current timestamp (in the format ) of the nginx cached time (no syscall involved unlike Lua's os.date function).

This is the UTC time.

ngx.cookie_time
syntax: str = ngx.cookie_time(sec)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Returns a formated string can be used as the cookie expiration time. The parameter  is the timestamp in seconds (like those returned from ngx.time).

ngx.say(ngx.cookie_time(1290079655)) -- yields "Thu, 18-Nov-10 11:27:35 GMT"

ngx.http_time
syntax: str = ngx.http_time(sec)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Returns a formated string can be used as the http header time (for example, being used in  header). The parameter  is the timestamp in seconds (like those returned from ngx.time).

ngx.say(ngx.http_time(1290079655)) -- yields "Thu, 18 Nov 10 11:27:35 GMT"

ngx.parse_http_time
syntax: sec = ngx.parse_http_time(str)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Parse the http time string (as returned by ngx.http_time) into seconds. Returns the seconds or  if the input string is in bad forms.

local time = ngx.parse_http_time("Thu, 18 Nov 10 11:27:35 GMT") if time == nil then ...   end

ngx.is_subrequest
syntax: value = ngx.is_subrequest

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Returns  if the current request is an nginx subrequest, or   otherwise.

ngx.re.match
syntax: captures = ngx.re.match(subject, regex, options?, ctx?)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Matches the  string using the Perl-compatible regular expression   with the optional.

Only the first occurrence of the match is returned, or  if no match is found. In case of fatal errors, like seeing bad  sequences in   mode, a Lua exception will be raised.

When a match is found, a Lua table  is returned, where   holds the whole substring being matched, and   holds the first parenthesized subpattern's capturing,   the second, and so on.

local m = ngx.re.match("hello, 1234", "[0-9]+") -- m[0] == "1234"

local m = ngx.re.match("hello, 1234", "([0-9])[0-9]+") -- m[0] == "1234" -- m[1] == "1"

Unmatched subpatterns will have  values in their   table fields.

local m = ngx.re.match("hello, world", "(world)|(hello)") -- m[0] == "hello" -- m[1] == nil -- m[2] == "hello"

You can also specify  to control how the match will be performed. The following option characters are supported:

a            anchored mode (only match from the beginning)

d            enable the DFA mode (or the longest token match semantics). this requires PCRE 6.0+ or else a Lua exception will be thrown. first introduced in ngx_lua v0.3.1rc30.

i            caseless mode (similar to Perl's /i modifier)

j            enable PCRE JIT compilation, this requires PCRE 8.20+ and PCRE must be built with the --enable-jit option, or it is                 a no-op. this should always be used along with the option to gain the full performance benefit. first introduced in ngx_lua v0.3.1rc30.

m            multi-line mode (similar to Perl's /m modifier)

o            compile-once mode (similar to Perl's /o modifer), to enable the worker-process-level compiled-regex cache

s            single-line mode (similar to Perl's /s modifier)

u            UTF-8 mode. this requires PCRE to be built with the --enable-utf8 option or else a Lua exception will be thrown.

x            extended mode (similar to Perl's /x modifier)

These options can be combined:

local m = ngx.re.match("hello, world", "HEL LO", "ix") -- m[0] == "hello"

local m = ngx.re.match("hello, 美好生活", "HELLO, (.{2})", "iu") -- m[0] == "hello, 美好" -- m[1] == "美好"

The  option is useful for performance tuning, because the regex pattern in question will only be compiled once, cached in the worker-process level, and shared among all requests in the current Nginx worker process. You can tune the upper limit of the regex cache via the lua_regex_cache_max_entries directive.

The optional fourth argument,, can be a Lua table holding an optional   field. When the  field in the   table argument is specified,   will start matching from that offset. Regardless of the presence of the  field in the   table,   will always set this   field to the position after the substring matched by the whole pattern in case of a successful match. When match fails, the  table will be left intact.

local ctx = {} local m = ngx.re.match("1234, hello", "[0-9]+", "", ctx) -- m[0] = "1234" -- ctx.pos == 4

local ctx = { pos = 2 } local m = ngx.re.match("1234, hello", "[0-9]+", "", ctx) -- m[0] = "34" -- ctx.pos == 4

The  table argument combined with the   regex modifier can be used to construct a lexer atop.

Note that, the  argument is not optional when the   argument is specified; use the empty Lua string  as the placeholder for   if you do not want to specify any regex options.

This method requires the PCRE library enabled in your Nginx build. (Known Issue With Special PCRE Sequences).

This feature was introduced in the  release.

ngx.re.gmatch
syntax: iterator = ngx.re.gmatch(subject, regex, options?)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Similar to ngx.re.match, but returns a Lua iterator instead, so as to let the user programmer iterate all the matches over the  string argument with the PCRE.

Here is a small exmple to demonstrate its basic usage:

local iterator = ngx.re.gmatch("hello, world!", "([a-z]+)", "i") local m   m = iterator    -- m[0] == m[1] == "hello" m = iterator   -- m[0] == m[1] == "world" m = iterator   -- m == nil

More often we just put it into a Lua  loop:

for m in ngx.re.gmatch("hello, world!", "([a-z]+)", "i") ngx.say(m[0]) ngx.say(m[1]) end

The optional  argument takes exactly the same semantics as the ngx.re.match method.

The current implementation requires that the iterator returned should only be used in a single request. That is, one should not assign it to a variable belonging to persistent namespace like a Lua package.

This method requires the PCRE library enabled in your Nginx build. (Known Issue With Special PCRE Sequences).

This feature was first introduced in the  release.

ngx.re.sub
syntax: newstr, n = ngx.re.sub(subject, regex, replace, options?)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Substitutes the first match of the Perl-compatible regular expression  on the   argument string with the string or function argument. The optional  argument has exactly the same meaning as in ngx.re.match.

This method returns the resulting new string as well as the number of successful substitutions, or throw out a Lua exception when an error occurred (syntax errors in the  string argument, for example).

When the  is a string, then it is treated as a special template for string replacement. For example,

local newstr, n = ngx.re.sub("hello, 1234", "([0-9])[0-9]", "[$0][$1]") -- newstr == "hello, [12][1]34" -- n == 1

where  referring to the whole substring matched by the pattern and   referring to the first parenthesized capturing substring.

You can also use curly braces to disambiguate variable names from the background string literals:

local newstr, n = ngx.re.sub("hello, 1234", "[0-9]", "${0}00") -- newstr == "hello, 10034" -- n == 1

Literal dollar sign characters in the   string argument can be escaped by another dollar sign, for instance,

local newstr, n = ngx.re.sub("hello, 1234", "[0-9]", "$$") -- newstr == "hello, $234" -- n == 1

Do not use backlashes to escape dollar signs; it will not work as expected.

When the  argument is of type "function", then it will be invoked with the "match table" as the argument to generate the replace string literal for substitution. The "match table" fed into the  function is exactly the same as the return value of ngx.re.match. Here is an example:

local func = function (m) return "[" .. m[0] .. "][" .. m[1] .. "]"   end local newstr, n = ngx.re.sub("hello, 1234", "( [0-9] ) [0-9]", func, "x") -- newstr == "hello, [12][1]34" -- n == 1

The dollar sign characters in the return value of the  function argument are not special at all.

This method requires the PCRE library enabled in your Nginx build. (Known Issue With Special PCRE Sequences).

This feature was first introduced in the  release.

ngx.re.gsub
syntax: newstr, n = ngx.re.gsub(subject, regex, replace, options?)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Just like ngx.re.sub, but does global substitution.

Here is some examples:

local newstr, n = ngx.re.gsub("hello, world", "([a-z])[a-z]+", "[$0,$1]", "i") -- newstr == "[hello,h], [world,w]" -- n == 2

local func = function (m) return "[" .. m[0] .. "," .. m[1] .. "]"   end local newstr, n = ngx.re.gsub("hello, world", "([a-z])[a-z]+", func, "i") -- newstr == "[hello,h], [world,w]" -- n == 2

This method requires the PCRE library enabled in your Nginx build. (Known Issue With Special PCRE Sequences).

This feature was first introduced in the  release.

ngx.shared.DICT
syntax: dict = ngx.shared.DICT

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Fetching the shm-based Lua dictionary object for the shared memory zone named  defined by the lua_shared_dict directive.

The resulting object  has the following methods:


 * get
 * set
 * add
 * replace
 * incr
 * delete

Here is an example:

http { lua_shared_dict dogs 10m; server { location /set { content_by_lua ' local dogs = ngx.shared.dogs dogs:set("Jim", 8) ngx.say("STORED") ';           }            location /get { content_by_lua ' local dogs = ngx.shared.dogs ngx.say(dogs:get("Jim")) ';           }        }    }

Let us test it:

$ curl localhost/set STORED

$ curl localhost/get 8

$ curl localhost/get 8

You will consistently get the output  when accessing   regardless how many Nginx workers there are because the   dictionary resides in the shared memory and visible to all of the worker processes.

The shared dictionary will retain its contents through a server config reload (either by sending the  signal to the Nginx process or by using the   command-line option).

The contents in the dictionary storage will be lost, however, when the Nginx server quits.

This feature was first introduced in the  release.

ngx.shared.DICT.get
syntax: value = ngx.shared.DICT:get(key)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Retrieving the value in the dictionary ngx.shared.DICT for the key. If the key does not exist or has been expired, then  will be returned.

The value returned will have the original data type when they were inserted into the dictionary, for example, Lua booleans, numbers, or strings.

The first argument to this method must be the dictionary object itself, for example,

local cats = ngx.shared.cats local value = cats.get(cats, "Marry")

or use Lua's syntactic sugar for method calls:

local cats = ngx.shared.cats local value = cats:get("Marry")

These two forms are fundamentally equivalent.

This feature was first introduced in the  release.

See also ngx.shared.DICT.

ngx.shared.DICT.set
syntax: success, err, forcible = ngx.shared.DICT:set(key, value, exptime?)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Unconditionally sets a key-value pair into the shm-based dictionary ngx.shared.DICT. Returns three values:


 * : boolean value to indicate whether the key-value pair is stored or not.
 * : textual error message, can be.
 * : a boolean value to indicate whether other valid items have been removed forcibly when out of storage in the shared memory zone.

The  argument inserted can be Lua booleans, numbers, strings, or. Their value type will also be stored into the dictionary, thus you can get exactly the same data type when later retrieving the value out of the dictionary via the get method.

The optional  argument specifies expiration time (in seconds) for the inserted key-value pair. The time resolution is  seconds. If the  takes the value   (which is the default), then the item will never be expired.

When it fails to allocate memory for the current key-value item, then  will try removing existing items in the storage according to the Least-Recently Used (LRU) algorithm. Note that, LRU takes priority over expiration time here. If up to tens of existing items have been removed and the storage left is still insufficient (either due to the total capacity limit specified by lua_shared_dict or memory segmentation), then the  return value will be   and   will be.

If this method succeeds in storing the current item by forcibly removing other not-yet-expired items in the dictionary via LRU, the  return value will be. If it stores the item without forcibly removing other valid items, then the return value  will be.

The first argument to this method must be the dictionary object itself, for example,

local cats = ngx.shared.cats local succ, err, forcible = cats.set(cats, "Marry", "it is a nice cat!")

or use Lua's syntactic sugar for method calls:

local cats = ngx.shared.cats local succ, err, forcible = cats:set("Marry", "it is a nice cat!")

These two forms are fundamentally equivalent.

This feature was first introduced in the  release.

See also ngx.shared.DICT.

ngx.shared.DICT.add
syntax: success, err, forcible = ngx.shared.DICT:add(key, value, exptime?)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Just like the set method, but only stores the key-value pair into the dictionary ngx.shared.DICT if the key does not exist.

If the  argument already exists in the dictionary (and not expired for sure), the   return value will be   and the   return value will be.

This feature was first introduced in the  release.

See also ngx.shared.DICT.

ngx.shared.DICT.replace
syntax: success, err, forcible = ngx.shared.DICT:replace(key, value, exptime?)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Just like the set method, but only stores the key-value pair into the dictionary ngx.shared.DICT if the key does exist.

If the  argument does not exist in the dictionary (or expired already), the   return value will be   and the   return value will be.

This feature was first introduced in the  release.

See also ngx.shared.DICT.

ngx.shared.DICT.delete
syntax: ngx.shared.DICT:delete(key)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Unconditionally removes the key-value pair from the shm-based dictionary ngx.shared.DICT.

It is equivalent to.

This feature was first introduced in the  release.

See also ngx.shared.DICT.

ngx.shared.DICT.incr
syntax: newval, err = ngx.shared.DICT:incr(key, value)

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

Increments the (numerical) value for  in the shm-based dictionary ngx.shared.DICT by the step value. Returns the new resulting number if the operation is successfully completed or  and an error message otherwise.

The key must already exist in the dictionary, otherwise it will return  and.

If the original value is not a valid Lua number in the dictionary, it will return  and.

The  argument can be any valid Lua numbers, like negative numbers or floating-point numbers.

This feature was first introduced in the  release.

See also ngx.shared.DICT.

ndk.set_var.DIRECTIVE
syntax: res = ndk.set_var.DIRECTIVE_NAME

context: set_by_lua*, rewrite_by_lua*, access_by_lua*, content_by_lua*, header_filter_by_lua*

This mechanism allows calling other nginx C modules' directives that are implemented by Nginx Devel Kit (NDK)'s set_var submodule's.

For example, the following HttpSetMiscModule directives can be invoked this way:


 * set_quote_sql_str
 * set_quote_pgsql_str
 * set_quote_json_str
 * set_unescape_uri
 * set_escape_uri
 * set_encode_base32
 * set_decode_base32
 * set_encode_base64
 * set_decode_base64
 * set_encode_hex
 * set_decode_hex
 * set_sha1
 * set_md5

For instance,

local res = ndk.set_var.set_escape_uri('a/b'); -- now res == 'a%2fb'

Similarly, the following directives provided by HttpEncryptedSessionModule can be invoked from within Lua too:


 * set_encrypt_session
 * set_decrypt_session

This feature requires the ngx_devel_kit module.

= HTTP 1.0 support =

The HTTP 1.0 protocol does not support chunked outputs and always requires an explicit  header when the response body is non-empty. So when an HTTP 1.0 request is present, This module will automatically buffer all the outputs of user calls of ngx.say and ngx.print and postpone sending response headers until it sees all the outputs in the response body, and at that time ngx_lua can calculate the total length of the body and construct a proper  header for the HTTP 1.0 client.

Note that, common HTTP benchmark tools like  and   always issue HTTP 1.0 requests by default. To force  to send HTTP 1.0 requests, use the  option.

= Data Sharing within an Nginx Worker =

'''NOTE: This mechanism behaves differently when code cache is turned off, and should be considered as a DIRTY TRICK. Backward compatibility is NOT guaranteed. Use at your own risk! We're going to design a whole new data-sharing mechanism.'''

If you want to globally share user data among all the requests handled by the same nginx worker process, you can encapsulate your shared data into a Lua module, require the module in your code, and manipulate shared data through it. It works because required Lua modules are loaded only once, and all coroutines will share the same copy of the module. Note however that Lua global variables WILL NOT persist between requests because of the one-coroutine-per-request isolation design.

Here is a complete small example:

-- mydata.lua module("mydata", package.seeall) local data = { dog = 3, cat = 4, pig = 5, }   function get_age(name) return data[name] end

and then accessing it from your nginx.conf:

location /lua { content_lua_by_lua ' local mydata = require("mydata") ngx.say(mydata.get_age("dog")) ';   }

Your  module in this example will only be loaded and run on the first request to the location , and all those subsequent requests to the same nginx worker process will use the reloaded instance of the module as well as the same copy of the data in it, until you send a  signal to the nginx master process to enforce a reload.

This data sharing technique is essential for high-performance Lua apps built atop this module. It is common to cache reusable data globally.

It is worth noting that this is per-worker sharing, not per-server sharing. That is, when you have multiple nginx worker processes under an nginx master, this data sharing cannot pass process boundary. If you indeed need server-wide data sharing, you can


 * 1) Use only a single nginx worker and a single server. This is not recommended when you have a multi-core CPU or multiple CPUs in a single machine.
 * 2) Use some true backend storage like ,  , or an RDBMS like.

= Known Issues =

Lua Coroutine Yielding/Resuming

 * As the module's predefined Nginx I/O API uses the coroutine yielding/resuming mechanism, user code should not call any Lua modules that use the Lua coroutine mechanism in order to prevent conflicts with the module's predefined Nginx API methods such as ngx.location.capture (Actually, coroutine modules have been masked off in content_by_lua directives and others). This limitation is significant and work is ongoing on an alternative coroutine implementation that can fit into the Nginx event model to address this. When this is done, it will be possible to use the Lua coroutine mechanism freely as it is in standard Lua implementations.
 * Lua's  builtin is implemented as a C function in both Lua 5.1 and LuaJIT 2.0 and when you call ngx.location.capture, ngx.exec, ngx.exit or ngx.req.read_body or similar in the file to be loaded by , a coroutine yield across the C function boundary will be initiated. This however is not allowed within ngx_lua and will usually result in error messages like  . To avoid this, define a real Lua module in your   file and use Lua's   builtin instead.
 * Because the standard Lua 5.1 interpreter's VM is not fully resumable, the methods ngx.location.capture, ngx.location.capture_multi, ngx.redirect, ngx.exec, and ngx.exit cannot be used within the context of a Lua pcall or xpcall when the standard Lua 5.1 interpreter is used; you'll get the error . To fix this, please use LuaJIT 2.0 instead, because LuaJIT 2.0 supports a fully resume-able VM.

Lua Variable Scope
Care should be taken when importing modules and this form should be used:

local xxx = require('xxx')


 * instead of the old deprecated form:

require('xxx')


 * If you have to use the old form, force reload the module for every request by using the  command:

package.loaded.xxx = nil require('xxx')

It is recommended to always place the following piece of code at the end of Lua modules that use the ngx.location.capture or ngx.location.capture_multi directives to prevent casual use of module-level global variables that are shared among all requests:

getmetatable(foo.bar).__newindex = function (table, key, val) error('Attempt to write to undeclared variable "' .. key .. '": '               .. debug.traceback) end

Assuming your current Lua module is named, this will guarantee that local variables in module   functions have been declared as "local". It prevents undesirable race conditions while accessing such variables. See Data Sharing within an Nginx Worker for the reasons behind this.

Locations With HttpEchoModule Directives
The ngx.location.capture and ngx.location.capture_multi directives cannot capture locations that include the echo_location, echo_location_async, echo_subrequest, or echo_subrequest_async directives.

location /foo { content_by_lua ' res = ngx.location.capature("/bar") ';   }    location /bar { echo_location /blah; }   location /blah { echo "Success!"; }

$ curl -i http://example.com/foo

will not work as expected.

Special PCRE Sequences
PCRE sequences such as,  , or  , require special attention because in string literals, the backslash character,  , is stripped out by both the Lua language parser and by the Nginx config file parser before processing. So the following snippet will not work as expected:

# nginx.conf ? location /test { ?    content_by_lua ' ?        local regex = "\d+"  -- THIS IS WRONG!! ?        local m = ngx.re.match("hello, 1234", regex) ?        if m then ngx.say(m[0]) else ngx.say("not matched!") end ?    ';    ? }    # evaluates to "not matched!"

To avoid this, you need to double escape the backslash:

# nginx.conf location /test { content_by_lua ' local regex = "\\\\d+" local m = ngx.re.match("hello, 1234", regex) if m then ngx.say(m[0]) else ngx.say("not matched!") end ';   }    # evaluates to "1234"

Here,  is stripped down to   by the Nginx config file parser and this is further stripped down to   by the Lua language parser before running.

Alternatively, you can present the regex pattern as a long-bracketed lua string literal by encasing it in "long brackets",, in which case backslashes have to only be escaped once for the Nginx config file parser.

# nginx.conf location /test { content_by_lua ' local regex = \\d+ local m = ngx.re.match("hello, 1234", regex) if m then ngx.say(m[0]) else ngx.say("not matched!") end ';   }    # evaluates to "1234"

Here,  is stripped down to   by the Nginx config file parser and this is processed correctly.

Note that you may need to use a longer from of the long bracket,, if the regex pattern contains   sequences. You can also, if you wish, use the  form as your default form and it may help with readability if you put a space between your long brackets and your regex patterns.

# nginx.conf location /test { content_by_lua ' local regex = [=[ [0-9]+ ]=] local m = ngx.re.match("hello, 1234", regex) if m then ngx.say(m[0]) else ngx.say("not matched!") end ';   }    # evaluates to "1234"

An alternative approach to escaping PCRE sequences is to ensure that Lua code is placed in external script files and executed using the various  directives. With this approach, the backslashes are only stripped by the Lua language parser and therefore only need to be escaped once each.

-- test.lua local regex = "\\d+" local m = ngx.re.match("hello, 1234", regex) if m then ngx.say(m[0]) else ngx.say("not matched!") end -- evaluates to "1234"

Within external script files, PCRE sequences presented as long-bracketed lua string literals do not require modification. -- test.lua local regex = \d+ local m = ngx.re.match("hello, 1234", regex) if m then ngx.say(m[0]) else ngx.say("not matched!") end -- evaluates to "1234"

= Typical Uses =

Just to name a few:


 * Mashup'ing and processing outputs of various nginx upstream outputs (proxy, drizzle, postgres, redis, memcached, and etc) in Lua,
 * doing arbitrarily complex access control and security checks in Lua before requests actually reach the upstream backends,
 * manipulating response headers in an arbitrary way (by Lua)
 * fetching backend information from external storage backends (like redis, memcached, mysql, postgresql) and use that information to choose which upstream backend to access on-the-fly,
 * coding up arbitrarily complex web applications in a content handler using synchronous but still non-blocking access to the database backends and other storage,
 * doing very complex URL dispatch in Lua at rewrite phase,
 * using Lua to implement advanced caching mechanism for nginx subrequests and arbitrary locations.

The possibilities are unlimited as the module allows bringing together various elements within Nginx as well as exposing the power of the Lua language to the user. The module provides the full flexibility of scripting while offering performance levels comparable with native C language programs both in terms of CPU time as well as memory footprint. This is particularly the case when LuaJIT 2.0 is enabled.

Other scripting language implementations typically struggle to match this performance level.

The Lua state (Lua VM instance) is shared across all the requests handled by a single nginx worker process to miminize memory use.

On a ThinkPad T400 2.80 GHz laptop, the HelloWorld example readily achieves 28k req/sec using. By contrast, Nginx + php-fpm 5.2.8 + Unix Domain Socket yields 6k req/sec and Node.js v0.6.1 yields 10.2k req/sec for their HelloWorld equivalents.

This module performs best when built with LuaJIT 2.0.

= Nginx Compatibility = The module is compatible with the following versions of Nginx:


 * 1.1.x (last tested: 1.1.5)
 * 1.0.x (last tested: 1.0.10)
 * 0.9.x (last tested: 0.9.4)
 * 0.8.x >= 0.8.54 (last tested: 0.8.54)

= Code Repository =

The code repository of this project is hosted on github at chaoslawful/lua-nginx-module.

= 安装 =

You may wish to consider using the ngx_openresty bundle to install Nginx, ngx_lua, either one of the Standard Lua interpreter or LuaJIT, as well as a package of powerful companion Nginx modules:

http://openresty.org

最简单的安装方式是.

当然，你也可以选择单独编译 ngx_lua 模块到 Nginx 中：


 * 1) 安装 Lua 5.1 或者 LuaJIT 2.0 （推荐）. Lua 可以从 www.lua.org 免费获取. 你的发行版中可能已经包括 Lua.
 * 2) 下载最新的 ngx_devel_kit (NDK) 模块 HERE.
 * 3) 下载最新的 HERE.
 * 4) 下载最新的 Nginx HERE (See Nginx Compatibility)

从源码编译模块：

wget 'http://nginx.org/download/nginx-1.0.10.tar.gz' tar -xzvf nginx-1.0.10.tar.gz   cd nginx-1.0.10/ # 设置 Lua 环境变量 export LUA_LIB=/path/to/lua/lib export LUA_INC=/path/to/lua/include # 或者设置 LuaJIT 环境变量 # export LUAJIT_LIB=/path/to/luajit/lib # export LUAJIT_INC=/path/to/luajit/include/luajit-2.0 # 这里假设 Nginx 安装路径是 /opt/nginx/ ./configure --prefix=/opt/nginx \ --add-module=/path/to/ngx_devel_kit \ --add-module=/path/to/lua-nginx-module make -j2 make install

= Bugs and Patches =

Please report bugs or submit patches by:


 * 1) Creating a ticket on the GitHub issue tracker (Recommended)
 * 2) Posting to the Nginx mailing list.

= TODO =

Short Term

 * add,  , and   options to ngx.location.capture and ngx.location.capture_multi` methods, to allow micro performance tuning on the user side.
 * add directives to run lua codes when nginx stops/reloads.
 * deal with TCP 3-second delay problem under great connection harness.
 * add options to ngx.location.capture and ngx.location.capture_multi in order to share and copy a particular set of nginx variables with subrequests, specified by the user.
 * add an option to ngx.location.capture and ngx.location.capture_multi so as to specify the ngx.ctx table for subrequests.
 * add support for multi-value arguments to if its   argument is a Lua table.
 * add APIs to access cookies as key/value pairs.

Longer Term

 * add the  directive to load module into main thread's globals.
 * add the "cosocket" mechamism that will emulate a common set of Lua socket API that will give you totally transparently non-blocking capability out of the box by means of a completely new upstream layer atop the nginx event model and no nginx subrequest overheads.
 * add Lua code automatic time slicing support by yielding and resuming the Lua VM actively via Lua's debug hooks.
 * make set_by_lua using the same mechanism as content_by_lua.
 * add coroutine API back to the Lua land.

= Changes =

v0.3.0
New features


 * added the header_filter_by_lua and header_filter_by_lua_file directives. thanks Liseen Wan (万珣新).
 * implemented the PCRE regex API for Lua: ngx.re.match, ngx.re.gmatch, ngx.re.sub, and ngx.re.gsub.
 * now we add the  and   table into   such that the user can write   and  . thanks @Lance.
 * added new directive lua_regex_cache_max_entries to control the upper limit of the worker-process-level compiled-regex cache enabled by the  regex option.
 * implemented the special ngx.ctx Lua table for user programmers to store per-request Lua context data for their applications. thanks 欧远宁 for suggesting this feature.
 * now ngx.print and ngx.say allow (nested) array-like table arguments. the array elements in them will be sent piece by piece. this will avoid string concatenation for templating engines like ltp.
 * implemented the ngx.req.get_post_args method for fetching url-encoded POST query arguments from within Lua.
 * implemented the ngx.req.get_uri_args method to fetch parsed URL query arguments from within Lua. thanks Bertrand Mansion (golgote).
 * added new function ngx.parse_http_time, thanks James Hurst.
 * now we allow Lua boolean and  values in arguments to ngx.say, ngx.print, ngx.log and print.
 * added support for user C macros  and  . for now we can only define them in  's   file because nginx  's   option hates values with double-quotes in them. sigh. ngx_openresty is already using this feature to bundle 3rd-party Lua libraries.

Bug fixes


 * worked-around the "stack overflow" issue while using  and disabling lua_code_cache, as described as github issue #27. thanks Patrick Crosby.
 * fixed the  alert while combining lua_need_request_body on + access_by_lua/rewrite_by_lua + proxy_pass/fastcgi_pass. thanks Liseen Wan (万珣新).
 * fixed issues with HTTP 1.0 HEAD requests.
 * made setting  after sending out response headers throw out a Lua exception to help debugging issues like github issue #49. thanks Bill Donahue (ikhoyo).
 * fixed an issue regarding defining global variables in C header files: we should have defined the global  in a single compilation unit. thanks @姜大炮.

= 测试用例 =

运行测试用例，有以下依赖：


 * Nginx 版本 >= 0.8.54


 * Perl 模块:
 * test-nginx: http://github.com/agentzh/test-nginx


 * Nginx 模块:
 * echo-nginx-module: http://github.com/agentzh/echo-nginx-module
 * drizzle-nginx-module: http://github.com/chaoslawful/drizzle-nginx-module
 * rds-json-nginx-module: http://github.com/agentzh/rds-json-nginx-module
 * set-misc-nginx-module: http://github.com/agentzh/set-misc-nginx-module
 * headers-more-nginx-module: http://github.com/agentzh/headers-more-nginx-module
 * memc-nginx-module: http://github.com/agentzh/memc-nginx-module
 * srcache-nginx-module: http://github.com/agentzh/srcache-nginx-module
 * ngx_auth_request: http://mdounin.ru/hg/ngx_http_auth_request_module/


 * C 库:
 * yajl: https://github.com/lloyd/yajl


 * Lua 模块:
 * lua-yajl: https://github.com/brimworks/lua-yajl
 * Note: the compiled module has to be placed in '/usr/local/lib/lua/5.1/'


 * Applications:
 * mysql: create database 'ngx_test', grant all privileges to user 'ngx_test', password is 'ngx_test'
 * memcached

下面的模块在配置时候顺序很重要，正确的顺序如下：


 * 1) ngx_devel_kit
 * 2) set-misc-nginx-module
 * 3) ngx_http_auth_request_module
 * 4) echo-nginx-module
 * 5) memc-nginx-module
 * 6) lua-nginx-module (i.e. this module)
 * 7) headers-more-nginx-module
 * 8) srcache-nginx-module
 * 9) drizzle-nginx-module
 * 10) rds-json-nginx-module

= Copyright and License =

This module is licenced under the BSD license.

Copyright (C) 2009-2011, by Xiaozhe Wang (chaoslawful) .

Copyright (C) 2009-2011, by Zhang "agentzh" Yichun (章亦春) .

All rights reserved.

Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:


 * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.


 * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

= 其他参考 =


 * 根据 URL 参数动态构建 MySQL 查询
 * 使用 Redis 和 Lua 创建动态路由规则
 * 使用 LuaRocks 安装 ngx_lua
 * ngx_lua 介绍
 * ngx_devel_kit
 * HttpEchoModule
 * HttpDrizzleModule
 * postgres-nginx-module
 * HttpMemcModule
 * The ngx_openresty bundle