This module embeds the Lua interpreter or LuaJIT into the nginx core and integrates the powerful Lua threads (aka Lua coroutines) into the nginx event model
Unlike [Apache's mod_lua](http://httpd.apache.org/docs/2.3/mod/mod_lua.html) and [Lighttpd's mod_magnet](http://redmine.lighttpd.net/wiki/1/Docs:ModMagnet), Lua code written atop this module can be *100% non-blocking* on network traffic
[HttpDrizzleModule](http://wiki.nginx.org/HttpDrizzleModule), [ngx_postgres](http://github.com/FRiCKLE/ngx_postgres/), [HttpMemcModule](http://wiki.nginx.org/HttpMemcModule), [HttpRedis2Module](http://wiki.nginx.org/HttpRedis2Module) and [HttpProxyModule](http://wiki.nginx.org/HttpProxyModule) modules for details).
Enable or disable the Lua code cache for [set_by_lua_file](http://wiki.nginx.org/HttpLuaModule#set_by_lua_file),
[content_by_lua_file](http://wiki.nginx.org/HttpLuaModule#content_by_lua_file), [rewrite_by_lua_file](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua_file), and
[access_by_lua_file](http://wiki.nginx.org/HttpLuaModule#access_by_lua_file), and also force Lua module reloading on a per-request basis.
like those specified by [set_by_lua](http://wiki.nginx.org/HttpLuaModule#set_by_lua), [content_by_lua](http://wiki.nginx.org/HttpLuaModule#content_by_lua),
[access_by_lua](http://wiki.nginx.org/HttpLuaModule#access_by_lua), and [rewrite_by_lua](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua) will *always* be
Specifies the maximal entries allowed in the worker-process-level compiled-regex cache.
The regular expressions used in [ngx.re.match](http://wiki.nginx.org/HttpLuaModule#ngx.re.match), [ngx.re.gmatch](http://wiki.nginx.org/HttpLuaModule#ngx.re.gmatch), [ngx.re.sub](http://wiki.nginx.org/HttpLuaModule#ngx.re.sub), and [ngx.re.gsub](http://wiki.nginx.org/HttpLuaModule#ngx.re.gsub) will be cached in this cache if the regex option `o` (i.e., compile-once flag) is specified.
The default entries allowed is 1024.
When the user Lua programs are exceeding this limit, those new regexes will not be cached at all (as if no `o` option is ever specified), and there will be one (and only one) warning in nginx's `error.log` file, like this
You shouldn't specify the `o` regex option for regexes (and/or `replace` string arguments for [ngx.re.sub](http://wiki.nginx.org/HttpLuaModule#ngx.re.sub) and [ngx.re.gsub](http://wiki.nginx.org/HttpLuaModule#ngx.re.gsub)) that are generated *on the fly* and give rise to infinite variations, or you'll quickly reach the limit specified here.
Execute user code specified by `<lua-script-str>` with input arguments `$arg1 $arg2 ...`, and set the script's return value to `$res` in string form. In
This directive can be freely mixed with all the directives of [HttpRewriteModule](http://wiki.nginx.org/HttpRewriteModule), [HttpSetMiscModule](http://wiki.nginx.org/HttpSetMiscModule), and [HttpArrayVarModule](http://wiki.nginx.org/HttpArrayVarModule). All of these directives will run in exactly the same order that they are written in the config file. For example,
The use code is executed in a new spawned coroutine with independent global environment (i.e. a sandbox).
Do not use this directive and other content handler directives in a same location. For example, it's bad to use this directive with a [proxy_pass](http://wiki.nginx.org/HttpProxyModule#proxy_pass) directive in the same location.
Basically the same as [content_by_lua](http://wiki.nginx.org/HttpLuaModule#content_by_lua), except the code to be executed is in
the file specified by `<path-lua-script>`.
Nginx variables can be used in `<path-to-lua-script>` string, in order to provide
greater flexibility in practice. But this feature must be used carefully, so is
not recommend for beginners.
When the Lua code cache is on (this is the default), the user code is loaded once at the first request and cached. Nginx config must be reloaded if you modified the file and expected to see updated behavior. You can disable the Lua code cache by setting [lua_code_cache](http://wiki.nginx.org/HttpLuaModule#lua_code_cache) `off` in your `nginx.conf` file.
Note that this handler always runs *after* the standard [HttpRewriteModule](http://wiki.nginx.org/HttpRewriteModule). So the following will work as expected:
because `if` runs *before* [rewrite_by_lua](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua) even if it's put after [rewrite_by_lua](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua) in the config.
It's worth mentioning that, the `ngx_eval` module can be approximately implemented by [rewrite_by_lua](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua). For example,
Note that calling `ngx.exit(ngx.OK)` just returning from the current [rewrite_by_lua](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua) handler, and the nginx request processing control flow will still continue to the content handler. To terminate the current request from within the current [rewrite_by_lua](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua) handler, calling [ngx.exit](http://wiki.nginx.org/HttpLuaModule#ngx.exit) with status >= 200 (`ngx.HTTP_OK`) and status <300(`ngx.HTTP_SPECIAL_RESPONSE`)forsuccessfulquitsand`ngx.exit(ngx.HTTP_INTERNAL_SERVER_ERROR)`(oritsfriends)forfailures.
If one uses [HttpRewriteModule](http://wiki.nginx.org/HttpRewriteModule)'s [rewrite](http://wiki.nginx.org/HttpRewriteModule#rewrite) directive to change the URI and initiate location re-lookups (kinda like internal redirections), then [rewrite_by_lua](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua) and [rewrite_by_lua_file](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua_file) will be skipped altogether in the current location. For example,
location /foo {
rewrite ^ /bar;
rewrite_by_lua 'ngx.exit(503)';
}
location /bar {
...
}
Here the Lua code `ngx.exit(503)` will never run while all the Lua code (except access phase handlers) in the `/bar` location will not be affected anyway. Similarly, `rewrite ^ /bar last` will also initiate a location re-lookup. If you use the `break` modifier for the [rewrite](http://wiki.nginx.org/HttpRewriteModule#rewrite) directive, however, no location re-lookup will be triggered, and therefore, the rewrite-phase Lua code will still be run as normal.
Act as an access phase handler and execute user code specified by `<lua-script-str>` for every request. The user code may call predefined APIs to generate response content.
This hook uses exactly the same mechanism as [content_by_lua](http://wiki.nginx.org/HttpLuaModule#content_by_lua) so all the nginx APIs defined there are also available here.
Note that this handler always runs *after* the standard [HttpAccessModule](http://wiki.nginx.org/HttpAccessModule). So the following will work as expected:
That is, if a client address appears in the blacklist, then we don't have to bother sending a MySQL query to do more advanced authentication in [access_by_lua](http://wiki.nginx.org/HttpLuaModule#access_by_lua).
It's worth mentioning that, the `ngx_auth_request` module can be approximately implemented by [access_by_lua](http://wiki.nginx.org/HttpLuaModule#access_by_lua). For example,
Note that calling `ngx.exit(ngx.OK)` just returning from the current [access_by_lua](http://wiki.nginx.org/HttpLuaModule#access_by_lua) handler, and the nginx request processing control flow will still continue to the content handler. To terminate the current request from within the current [access_by_lua](http://wiki.nginx.org/HttpLuaModule#access_by_lua) handler, calling `ngx.exit(status)` where status >= 200 (`ngx.HTTP_OK`) and status <300(`ngx.HTTP_SPECIAL_RESPONSE`)forsuccessfulquitsand`ngx.exit(ngx.HTTP_INTERNAL_SERVER_ERROR)`oritsfriendsforfailures.
* Output API (e.g., [ngx.say](http://wiki.nginx.org/HttpLuaModule#ngx.say) and [ngx.send_headers](http://wiki.nginx.org/HttpLuaModule#ngx.send_headers))
* Control APIs (e.g., [ngx.exit](http://wiki.nginx.org/HttpLuaModule#ngx.exit))
* Subrequest APIs (e.g., [ngx.location.capture](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture) and [ngx.location.capture_multi](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture_multi))
This is very much like [header_filter_by_lua](http://wiki.nginx.org/HttpLuaModule#header_filter_by_lua) except that it loads Lua code from an external Lua source file.
Force reading request body data or not. The client request body won't be read, so you have to explicitly force reading the body if you need its content.
If you want to read the request body data from the [$request_body](http://wiki.nginx.org/HttpCoreModule#.24request_body) variable, make sure that
your have configured [client_body_buffer_size](http://wiki.nginx.org/HttpCoreModule#client_body_buffer_size) to have exactly the same value as [client_max_body_size](http://wiki.nginx.org/HttpCoreModule#client_max_body_size).
If the current location defines [rewrite_by_lua](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua) or [rewrite_by_lua_file](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua_file),
then the request body will be read just before the [rewrite_by_lua](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua) or [rewrite_by_lua_file](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua_file) code is run (and also at the
`rewrite` phase). Similarly, if only [content_by_lua](http://wiki.nginx.org/HttpLuaModule#content_by_lua) is specified,
You're recommended to use the [ngx.req.read_body](http://wiki.nginx.org/HttpLuaModule#ngx.req.read_body) function and [ngx.req.discard_body](http://wiki.nginx.org/HttpLuaModule#ngx.req.discard_body) for finer control over the request body reading process though.
The same applies to [access_by_lua](http://wiki.nginx.org/HttpLuaModule#access_by_lua) and [access_by_lua_file](http://wiki.nginx.org/HttpLuaModule#access_by_lua_file).
The Nginx API exposed to the Lua land is provided in the form of two standard packages `ngx` and `ndk`. These packages are in the default global scope.
When you're writing your own external Lua modules, however, you can introduce these packages by using the [package.seeall](http://www.lua.org/manual/5.1/manual.html#pdf-package.seeall) option:
module("my_module", package.seeall)
function say(a) ngx.say(a) end
Alternatively, import them to your Lua modules by using file-scoped local Lua variables, like this:
local ngx = ngx
module("my_module")
function say(a) ngx.say(a) end
You can directly require the standard packages `ngx` and `ndk` introduced by this Nginx module, like this:
local ngx = require "ngx"
local ndk = require "ndk"
The ability to require these packages was introduced in the `v0.2.1rc19` release.
Network I/O operations in user code should only be done through our Nginx APIs defined below, otherwise Nginx event loop may be blocked and performance may drop off dramatically. Small disk file operations can be done via Lua's standard `io` and `file` libraries but should be eliminated wherever possible because these also block the Nginx process. Delegating all network and disk I/O operations to Nginx subrequests (via the [ngx.location.capture](http://wiki.nginx.org/HttpLuaModule#ngx.location.catpure) method and its friends) are strongly recommended.
Index the input arguments to the [set_by_lua](http://wiki.nginx.org/HttpLuaModule#set_by_lua) and [set_by_lua_file](http://wiki.nginx.org/HttpLuaModule#set_by_lua_file) directives:
These constants are usually used in [ngx.location.catpure](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture) and [ngx.location.capture_multi](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture_multi) method calls.
We can see that modification of the `ngx.ctx.blah` entry in the subrequest does not affect the one in its parent request. They do have two separate versions of `ngx.ctx.blah` per se.
Internal redirection will destroy the original request's `ngx.ctx` data (if any) and the new request will have an emptied `ngx.ctx` table. For instance,
Arbitrary data values can be inserted into this "matic" table, including Lua closures and nested tables. You can also register your own meta methods with it.
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 `ngx_proxy`, `ngx_fastcgi`, `ngx_memc`,
`ngx_postgres`, `ngx_drizzle`, and even `ngx_lua` itself and etc etc etc.
Also note that subrequests just mimic the HTTP interface but there's *no* extra HTTP/TCP traffic *nor* IPC involved. Everything works internally, efficiently, on the C level.
Subrequests are completely different from HTTP 301/302 redirection (via [ngx.redirect](http://wiki.nginx.org/HttpLuaModule#ngx.redirect)) and internal redirection (via [ngx.exec](http://wiki.nginx.org/HttpLuaModule#ngx.exec)).
[proxy_pass_request_headers](http://wiki.nginx.org/HttpProxyModule#proxy_pass_request_headers) `off` in your subrequest location to ignore the original request headers.
The [ngx.location.capture](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture) function is just a special form
of this function. Logically speaking, the [ngx.location.capture](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture) can be implemented like this
When assigning to `ngx.header.HEADER` will set, add, or clear the current request's response header named `HEADER`. Underscores (`_`) in the header names will be replaced by dashes (`-`) and the header names will be matched case-insensitively.
Setting `ngx.header.HEADER` after sending out response headers (either explicitly with [ngx.send_headers](http://wiki.nginx.org/HttpLuaModule#ngx.send_headers) or implicitly with [ngx.print](http://wiki.nginx.org/HttpLuaModule#ngx.print) and its friends) will throw out a Lua exception.
Reading `ngx.header.HEADER` will return the value of the response header named `HEADER`. Underscores (`_`) in the header names will also be replaced by dashes (`-`) and the header names will be matched case-insensitively. If the response header is not present at all, `nil` will be returned.
This is particularly useful in the context of [filter_header_by_lua](http://wiki.nginx.org/HttpLuaModule#filter_header_by_lua) and [filter_header_by_lua_file](http://wiki.nginx.org/HttpLuaModule#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;
}
For multi-value headers, all of the values of header will be collected in order and returned as a Lua table. For example, response headers
Foo: bar
Foo: baz
will result in
{"bar", "baz"}
to be returned when reading `ngx.header.Foo`.
Note that `ngx.header` is not a normal Lua table so you cannot iterate through it using Lua's `ipairs` function.
For reading *request* headers, use the [ngx.req.get_headers](http://wiki.nginx.org/HttpLuaModule#ngx.req.get_headers) function instead.
Rewrite the current request's (parsed) URI by the `uri` argument. The `uri` argument must be a Lua string and cannot be of zero length, or a Lua exception will be thrown.
The optional boolean `jump` argument can trigger location rematch (or location jump) as [HttpRewriteModule](http://wiki.nginx.org/HttpRewriteModule)'s [rewrite](http://wiki.nginx.org/HttpRewriteModule#rewrite) directive, that is, when `jump` is `true` (default to `false`), this function will never return and it will tell Nginx to try re-searching locations with the new URI value at the later `post-rewrite` phase and jumping to the new location. Location jump will not be triggered otherwise, and only the current request's URI will be modified, which is also the default behavior. This function will return but with no returned values when the `jump` argument is `false` or absent altogether.
The `jump` can only be set to `true` in [rewrite_by_lua](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua) and [rewrite_by_lua_file](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua_file). Use of jump in other contexts is prohibited and will throw out a Lua exception.
Note that you cannot use this interface to rewrite URI arguments, and you need to use [ngx.req.set_uri_args](http://wiki.nginx.org/HttpLuaModule#ngx.req.set_uri_args) for that. For instance, Nginx config
Multiple occurrences of an argument key will result in a table value holding all of the values for that key in order.
Keys and values will be automatically unescaped according to URI escaping rules. For example, in the above settings, `GET /test?a%20b=1%61+2` will yield the output
That is, they will take Lua boolean values `true`. However, they're different from arguments taking empty string values. For example, `GET /test?foo=&bar=` will give something like
Returns a Lua table holds all of the current request's POST query arguments. It's required to read the request body first by calling [ngx.req.read_body](http://wiki.nginx.org/HttpLuaModule#ngx.req.read_body) or to turn on the [lua_need_request_body](http://wiki.nginx.org/HttpLuaModule#lua_need_request_body) directive, or a Lua exception will be thrown.
That is, they will take Lua boolean values `true`. However, they're different from arguments taking empty string values. For example, `POST /test` with request body `foo=&bar=` will give something like
Another way to read individual request headers is to use `ngx.var.http_HEADER`, that is, nginx's standard [$http_HEADER](http://wiki.nginx.org/HttpCoreModule#.24http_HEADER) variables.
If the request body is already read previously by turning on [lua_need_request_body](http://wiki.nginx.org/HttpLuaModule#lua_need_request_body) or by using other modules, then this function is a no-op and returns immediately.
If the request body has already been explicitly discarded, either by this module's [ngx.req.discard_body](http://wiki.nginx.org/HttpLuaModule#ngx.req.discard_body) or other modules, this function is a no-op and returns immediately.
In case of errors, like connection errors while reading the data, this method will throw out a Lua exception *or* terminate the current request with the 500 status code immediately.
You can later either retrieve the request body data via [ngx.req.get_body_data](http://wiki.nginx.org/HttpLuaModule#ngx.req.get_body_data) or retrieve the temporary file name for the body data cached to disk via [ngx.req.get_body_file](http://wiki.nginx.org/HttpLuaModule#ngx.req.get_body_file), depending on
1. and whether you have turned on [client_body_in_file_only](http://wiki.nginx.org/HttpCoreModule#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's crucial to use the [ngx.req.discard_body](http://wiki.nginx.org/HttpLuaModule#ngx.req.discard_body) function to explicitly discard the request body, or you'll break HTTP 1.1 keepalive and HTTP 1.1 pipelining.
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.
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](http://wiki.nginx.org/HttpLuaModule#ngx.req.get_post_args) instead.
1. the request body has been read into disk temporary files,
1. or the request body has zero size.
If the request body has not been read yet, call [ngx.req.read_body](http://wiki.nginx.org/HttpLuaModule#ngx.req.read_body) first (or turned on [lua_need_request_body](http://wiki.nginx.org/HttpLuaModule#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](http://wiki.nginx.org/HttpLuaModule#ngx.req.get_body_file) function instead.
In case that you want to enforce in-memory request bodies, try setting [client_body_buffer_size](http://wiki.nginx.org/HttpCoreModule#client_body_buffer_size) to the same size value in [client_max_body_size](http://wiki.nginx.org/HttpCoreModule#client_max_body_size).
Note that calling this function instead of using `ngx.var.request_body` or `ngx.var.echo_request-body` is more efficient because it can save one dynamic memory allocation and one data copy.
This function was first introduced in the `v0.3.1rc17` release.
See also [ngx.req.get_body_file](http://wiki.nginx.org/HttpLuaModule#ngx.req.get_body_file).
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](http://wiki.nginx.org/HttpLuaModule#ngx.req.read_body) first (or turned on [lua_need_request_body](http://wiki.nginx.org/HttpLuaModule#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](http://wiki.nginx.org/HttpLuaModule#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](http://wiki.nginx.org/HttpCoreModule#client_body_in_file_only).
This function was first introduced in the `v0.3.1rc17` release.
See also [ngx.req.get_body_data](http://wiki.nginx.org/HttpLuaModule#ngx.req.get_body_data).
Set the current request's request body using the in-memory data specified by the `data` 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 1.0.8: [nginx-1.0.8-allow_request_body_updating.patch](https://github.com/agentzh/ngx_openresty/blob/master/patches/nginx-1.0.8-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](http://openresty.org/) 1.0.8.17+, then you've already had this patch applied.
This function was first introduced in the `v0.3.1rc18` release.
See also [ngx.req.set_body_file](http://wiki.nginx.org/HttpLuaModule#ngx.req.set_body_file).
Set the current request's request body using the in-file data specified by the `file_name` argument.
If the optional `auto_clean` argument is given a `true` value, then this file will be automatically removed at request completion or the next time this function or [ngx.req.set_body_data](http://wiki.nginx.org/HttpLuaModule#ngx.req.set_body_data) are called in the same request. The `auto_clean` is default to `false`.
You must ensure that the file specified by the `file_name` 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 1.0.8: [nginx-1.0.8-allow_request_body_updating.patch](https://github.com/agentzh/ngx_openresty/blob/master/patches/nginx-1.0.8-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](http://openresty.org/) 1.0.8.17+, then you've already had this patch applied.
This function was first introduced in the `v0.3.1rc18` release.
See also [ngx.req.set_body_data](http://wiki.nginx.org/HttpLuaModule#ngx.req.set_body_data).
The optional second `args` 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 `args` 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.
This method is very much like the [echo_exec](http://wiki.nginx.org/HttpEchoModule#echo_exec) directive in [HttpEchoModule](http://wiki.nginx.org/HttpEchoModule).
This method *must* be called before [ngx.send_headers](http://wiki.nginx.org/HttpLuaModule#ngx.send_headers) or explicit response body outputs by either [ngx.print](http://wiki.nginx.org/HttpLuaModule#ngx.print) or [ngx.say](http://wiki.nginx.org/HttpLuaModule#ngx.say).
Emit arguments concatenated to the HTTP client (as response body). If response headers have not been sent yet, this function will first send the headers out, and then output the body data.
The `log_level` argument can take constants like `ngx.ERR` and `ngx.WARN`. Check out [Nginx log level constants](http://wiki.nginx.org/HttpLuaModule#Nginx_log_level_constants) for details.
Force flushing the response outputs. This operation has no effect in HTTP 1.0 buffering output mode. See [HTTP 1.0 support](http://wiki.nginx.org/HttpLuaModule#HTTP_1.0_support).
When `status == 0` (i.e., `ngx.OK`), it will only quit the current phase handler (or the content handler if the [content_by_lua](http://wiki.nginx.org/HttpLuaModule#content_by_lua) directive is used) and continue to run laster phases (if any) for the current request.
Calculates the CRC-32 (Cyclic Redundancy Code) digest for the `str` argument.
This method performs better on relatively short `str` inputs (i.e., less than 30 ~ 60 bytes), as compared to [ngx.crc32_long](http://wiki.nginx.org/HttpLuaModule#ngx.crc32_long). The result is exactly the same as [ngx.crc32_long](http://wiki.nginx.org/HttpLuaModule#ngx.crc32_long).
Behind the scene, it is just a thin wrapper around the `ngx_crc32_short` function defined in the Nginx core.
This API was first introduced in the `v0.3.1rc8` release.
Calculates the CRC-32 (Cyclic Redundancy Code) digest for the `str` argument.
This method performs better on relatively long `str` inputs (i.e., longer than 30 ~ 60 bytes), as compared to [ngx.crc32_short](http://wiki.nginx.org/HttpLuaModule#ngx.crc32_short). The result is exactly the same as [ngx.crc32_short](http://wiki.nginx.org/HttpLuaModule#ngx.crc32_short).
Behind the scene, it is just a thin wrapper around the `ngx_crc32_long` function defined in the Nginx core.
This API was first introduced in the `v0.3.1rc8` release.
Returns the current timestamp (in the format `yyyy-mm-dd hh:mm:ss`) of the nginx cached time (no syscall involved unlike Lua's [os.date](http://www.lua.org/manual/5.1/manual.html#pdf-os.date) function).
Returns the current timestamp (in the format `yyyy-mm-dd hh:mm:ss`) of the nginx cached time (no syscall involved unlike Lua's [os.date](http://www.lua.org/manual/5.1/manual.html#pdf-os.date) function).
Returns a formated string can be used as the cookie expiration time. The parameter `sec` is the timestamp in seconds (like those returned from [ngx.time](http://wiki.nginx.org/HttpLuaModule#ngx.time)).
Returns a formated string can be used as the http header time (for example, being used in `Last-Modified` header). The parameter `sec` is the timestamp in seconds (like those returned from [ngx.time](http://wiki.nginx.org/HttpLuaModule#ngx.time)).
Parse the http time string (as returned by [ngx.http_time](http://wiki.nginx.org/HttpLuaModule#ngx.http_time)) into seconds. Returns the seconds or `nil` if the input string is in bad forms.
Only the first occurrence of the match is returned, or `nil` if no match is found. In case of fatal errors, like seeing bad `UTF-8` sequences in `UTF-8` mode, a Lua exception will be raised.
When a match is found, a Lua table `captures` is returned, where `captures[0]` holds the whole substring being matched, and `captures[1]` holds the first parenthesized subpattern's capturing, `captures[2]` the second, and so on. Here's some examples:
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 take `nil` values in their `captures` table fields. For instance,
local m = ngx.re.match("hello, world", "(world)|(hello)")
Escaping sequences in Perl-compatible regular expressions like `\d`, `\s`, and `\w`, require special care when specifying them in Lua string literals, because the backslash character, `\`, needs to be escaped in Lua string literals too, for example,
? m = ngx.re.match("hello, 1234", "\d+")
won't work as expected and won't match at all. Intead, you should escape the backslash itself and write
m = ngx.re.match("hello, 1234", "\\d+")
When you put the Lua code snippet in your `nginx.conf` file, you have to escape the backslash one more time, because your Lua code is now in an nginx string literal, and backslashes in nginx string literals require escaping as well. For instance,
location /test {
content_by_lua '
local m = ngx.re.match("hello, 1234", "\\\\d+")
if m then ngx.say(m[0]) else ngx.say("not matched!") end
The `o` regex option is good for performance tuning, because the regex in question will only be compiled once, cached in the worker-process level, and shared among all the requests in the current Nginx worker process. You can tune the upper limit of the regex cache via the [lua_regex_cache_max_entries](http://wiki.nginx.org/HttpLuaModule#lua_regex_cache_max_entries) directive.
The optional fourth argument, `ctx`, can be a Lua table holding an optional `pos` field. When the `pos` field in the `ctx` table argument is specified, `ngx.re.match` will start matching from that offset. Regardless of the presence of the `pos` field in the `ctx` table, `ngx.re.match` will always set this `pos` field to the position *after* the substring matched by the whole pattern in case of a successful match. When match fails, the `ctx` table will leave intact. Here is some examples,
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 `ctx` table argument combined with the `a` regex modifier can be used to construct a lexer atop `ngx.re.match`.
Note that, the `options` argument is not optional when the `ctx` argument is specified; use the empty Lua string (`""`) as the placeholder for `options` if you do not want to specify any regex options.
Similar to [ngx.re.match](http://wiki.nginx.org/HttpLuaModule#ngx.re.match), but returns a Lua iterator instead, so as to let the user programmer iterate all the matches over the `<subject>` string argument with the Perl-compatible regular expression `regex`.
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.
This feature was first introduced in the `v0.2.1rc12` release.
Substitutes the first match of the Perl-compatible regular expression `regex` on the `subject` argument string with the string or function argument `replace`. The optional `options` argument has exactly the same meaning as in [ngx.re.match](http://wiki.nginx.org/HttpLuaModule#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 `<replace>` string argument, for example).
When the `replace` 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 `$0` referring to the whole substring matched by the pattern and `$1` referring to the first parenthesized capturing substring.
You can also use curly braces to disambiguate variable names from the background string literals:
local newstr, n = ngx.re.sub("hello, 1234", "[0-9]", "${0}00")
-- newstr == "hello, 10034"
-- n == 1
Literal dollar sign characters (`$`) in the `replace` 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 won't work as expected.
When the `replace` 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 `replace` function is exactly the same as the return value of [ngx.re.match](http://wiki.nginx.org/HttpLuaModule#ngx.re.match). Here is an example:
Fetching the shm-based Lua dictionary object for the shared memory zone named `DICT` defined by the [lua_shared_dict](http://wiki.nginx.org/HttpLuaModule#lua_shared_dict) directive.
The resulting object `dict` has the following methods:
You will consistently get the output `8` when accessing `/get` regardless how many Nginx workers there are because the `dogs` 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 `HUP` signal to the Nginx process or by using the `-s reload` command-line option).
Retrieving the value in the dictionary [ngx.shared.DICT](http://wiki.nginx.org/HttpLuaModule#ngx.shared.DICT) for the key `key`. If the key does not exist or has been expired, then `nil` 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,
Unconditionally sets a key-value pair into the shm-based dictionary [ngx.shared.DICT](http://wiki.nginx.org/HttpLuaModule#ngx.shared.DICT). Returns three values:
*`success`: boolean value to indicate whether the key-value pair is stored or not.
*`err`: textual error message, can be `"no memory"`.
*`forcible`: a boolean value to indicate whether other valid items have been removed forcibly when out of storage in the shared memory zone.
The `value` argument inserted can be Lua booleans, numbers, strings, or `nil`. 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](http://wiki.nginx.org/HttpLuaModule#ngx.shared.DICT.get) method.
The optional `exptime` argument specifies expiration time (in seconds) for the inserted key-value pair. The time resolution is `0.001` seconds. If the `exptime` takes the value `0` (which is the default), then the item will never be expired.
When it fails to allocate memory for the current key-value item, then `set` 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](http://wiki.nginx.org/HttpLuaModule#lua_shared_dict) or memory segmentation), then the `err` return value will be `no memory` and `success` will be `false`.
If this method succeeds in storing the current item by forcibly removing other not-yet-expired items in the dictionary via LRU, the `forcible` return value will be `true`. If it stores the item without forcibly removing other valid items, then the return value `forcible` will be `false`.
Just like the [set](http://wiki.nginx.org/HttpLuaModule#ngx.shared.DICT.set) method, but only stores the key-value pair into the dictionary [ngx.shared.DICT](http://wiki.nginx.org/HttpLuaModule#ngx.shared.DICT) if the key does *not* exist.
If the `key` argument already exists in the dictionary (and not expired for sure), the `success` return value will be `false` and the `err` return value will be `"exists"`.
This feature was first introduced in the `v0.3.1rc22` release.
See also [ngx.shared.DICT](http://wiki.nginx.org/HttpLuaModule#ngx.shared.DICT).
Just like the [set](http://wiki.nginx.org/HttpLuaModule#ngx.shared.DICT.set) method, but only stores the key-value pair into the dictionary [ngx.shared.DICT](http://wiki.nginx.org/HttpLuaModule#ngx.shared.DICT) if the key *does* exist.
If the `key` argument does *not* exist in the dictionary (or expired already), the `success` return value will be `false` and the `err` return value will be `"not found"`.
This feature was first introduced in the `v0.3.1rc22` release.
See also [ngx.shared.DICT](http://wiki.nginx.org/HttpLuaModule#ngx.shared.DICT).
Increments the (numerical) value for `key` in the shm-based dictionary [ngx.shared.DICT](http://wiki.nginx.org/HttpLuaModule#ngx.shared.DICT) by the step value `value`. Returns the new resulting number if the operation is successfully completed or `nil` and an error message otherwise.
The key must already exist in the dictionary, otherwise it will return `nil` and `"not found"`.
If the original value is not a valid Lua number in the dictionary, it will return `nil` and `"not a number"`.
The `value` argument can be any valid Lua numbers, like negative numbers or floating-point numbers.
This feature was first introduced in the `v0.3.1rc22` release.
This mechanism allows calling other nginx C modules' directives that are implemented by [Nginx Devel Kit](https://github.com/simpl/ngx_devel_kit) (NDK)'s set_var submodule's `ndk_set_var_value`.
Similarly, the following directives provided by [HttpEncryptedSessionModule](http://wiki.nginx.org/HttpEncryptedSessionModule) can be invoked from within Lua too:
**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.
It's 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
* 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.
Actually the possibilities are unlimited, the Lua language is plays the role of glueing existing stuffs in Nginx.
This module gives the flexibility of scripting but also offers performance comparable with C (in terms of both CPU time and memory footprint), especially when LuaJIT 2.0 is enabled, which other scripting language implementations are hard to catch up.
1. Install Lua or LuaJIT into your system. At least Lua 5.1 is required. Lua can be obtained freely from its project [homepage](http://www.lua.org/). For Ubuntu/Debian users, just install the liblua5.1-0-dev package (or something like that).
1. Download the latest version of the release tarball of the ngx_devel_kit (NDK) module from lua-nginx-module [file list](http://github.com/simpl/ngx_devel_kit/tags).
1. Download the latest version of the release tarball of this module from lua-nginx-module [file list](http://github.com/chaoslawful/lua-nginx-module/tags).
1. Grab the nginx source code from [nginx.org](http://nginx.org/), for example, the version 1.0.5 (see nginx compatibility), and then build the source with this module:
Although a lot of effort has been put into testing and code tuning, there must be some serious bugs lurking somewhere in this module. So whenever you are bitten by any quirks, please don't hesitate to
1. create a ticket on the [issue tracking interface](http://github.com/chaoslawful/lua-nginx-module/issues) provided by GitHub,
1. or send a bug report or even patches to the [nginx mailing list](http://mailman.nginx.org/mailman/listinfo/nginx).
Source Repository
=================
Available on github at [chaoslawful/lua-nginx-module](http://github.com/chaoslawful/lua-nginx-module).
* add `ignore_resp_headers`, `ignore_resp_body`, and `ignore_resp` options to [ngx.location.capture](http://wiki.nginx.org/HttpLuaModule#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](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture) and [ngx.location.capture_multi](http://wiki.nginx.org/HttpLuaModule#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](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture) and [ngx.location.capture_multi](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture_multi) so as to specify the [ngx.ctx](http://wiki.nginx.org/HttpLuaModule#ngx.ctx) table for subrequests.
* expose nginx's shared memory facility to the Lua land.
* add the `lua_require` 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.
* As ngx_lua's predefined Nginx I/O APIs use coroutine yielding/resuming mechanism, the user code should not call any Lua modules that use coroutine API to prevent obfuscating the predefined Nginx APIs like [ngx.location.capture](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture) (actually coroutine modules have been masked off in [content_by_lua](http://wiki.nginx.org/HttpLuaModule#content_by_lua) directives and others). This limitation is a little crucial, but don't worry, we're working on an alternative coroutine implementation that can fit into the Nginx event model. When it is done, the user code will be able to use the Lua coroutine mechanism freely as in standard Lua again!
* Lua's `dofile` builtin is implemented as a C function in both Lua 5.1 and LuaJIT 2.0. And when you call [ngx.location.capture](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture), [ngx.exec](http://wiki.nginx.org/HttpLuaModule#ngx.exec), [ngx.exit](http://wiki.nginx.org/HttpLuaModule#ngx.exit) or [ngx.req.read_body](http://wiki.nginx.org/HttpLuaModule#ngx.req.read_body) or something like those in the .lua file to be loaded, it'll effectively initiate a coroutine yield and that yield will run across C function boundary, which is disallowed, and usually result in error messages like `lua handler aborted: runtime error: attempt to yield across C-call boundary`. You should use Lua's `require` builtin instead.
* Because the standard Lua 5.1 interpreter's VM is not fully resumable, the methods [ngx.location.capture](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture), [ngx.location.capture_multi](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture_multi), [ngx.redirect](http://wiki.nginx.org/HttpLuaModule#ngx.redirect), [ngx.exec](http://wiki.nginx.org/HttpLuaModule#ngx.exec), and [ngx.exit](http://wiki.nginx.org/HttpLuaModule#ngx.exit) cannot be used within the context of a Lua [pcall()](http://www.lua.org/manual/5.1/manual.html#pdf-pcall) or [xpcall()](http://www.lua.org/manual/5.1/manual.html#pdf-xpcall) when the standard Lua 5.1 interpreter is used; you'll get the error `attempt to yield across metamethod/C-call boundary`. To fix this, please use LuaJIT 2.0 instead, because LuaJIT 2.0 supports a fully resume-able VM.
* The [ngx.location.capture](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture) and [ngx.location.capture_multi](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture_multi) Lua methods cannot capture locations configured by [HttpEchoModule](http://wiki.nginx.org/HttpEchoModule)'s [echo_location](http://wiki.nginx.org/HttpEchoModule#echo_location), [echo_location_async](http://wiki.nginx.org/HttpEchoModule#echo_location_async), [echo_subrequest](http://wiki.nginx.org/HttpEchoModule#echo_subrequest), or [echo_subrequest_async](http://wiki.nginx.org/HttpEchoModule#echo_subrequest_async) directives. This won't be fixed in the future due to technical problems.
* **WATCH OUT: Globals WON'T persist between requests**, because of the one-coroutine-per-request isolation design. Especially watch yourself when using `require()` to import modules, and use this form:
The old form will cause module unusable in requests for the reason told previously. If you have to stick with the old form, you can always force loading module for every request by clean `package.loaded.<module>`, like this:
* It's recommended to always put the following piece of code at the end of your Lua modules using [ngx.location.capture](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture) or [ngx.location.capture_multi](http://wiki.nginx.org/HttpLuaModule#ngx.location.capture_multi) to prevent casual use of module-level global variables that are shared among *all* requests, which is usually not what you want:
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 `foo.bar`. This will guarantee that you have declared your Lua functions' local Lua variables as "local" in your Lua modules, or bad race conditions while accessing these variables under load will tragically happen. See the `Data Sharing within an Nginx Worker` for the reasons of this danger.
* added the [header_filter_by_lua](http://wiki.nginx.org/HttpLuaModule#header_filter_by_lua) and [header_filter_by_lua_file](http://wiki.nginx.org/HttpLuaModule#header_filter_by_lua_file) directives. thanks Liseen Wan (万珣新).
* implemented the PCRE regex API for Lua: [ngx.re.match](http://wiki.nginx.org/HttpLuaModule#ngx.re.match), [ngx.re.gmatch](http://wiki.nginx.org/HttpLuaModule#ngx.re.gmatch), [ngx.re.sub](http://wiki.nginx.org/HttpLuaModule#ngx.re.sub), and [ngx.re.gsub](http://wiki.nginx.org/HttpLuaModule#ngx.re.gsub).
* now we add the `ngx` and `ndk` table into `package.loaded` such that the user can write `local ngx = require 'ngx'` and `local ndk = require 'ndk'`. thanks @Lance.
* added new directive [lua_regex_cache_max_entries](http://wiki.nginx.org/HttpLuaModule#lua_regex_cache_max_entries) to control the upper limit of the worker-process-level compiled-regex cache enabled by the `o` regex option.
* implemented the special [ngx.ctx](http://wiki.nginx.org/HttpLuaModule#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](http://wiki.nginx.org/HttpLuaModule#ngx.print) and [ngx.say](http://wiki.nginx.org/HttpLuaModule#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](http://www.savarese.com/software/ltp/).
* implemented the [ngx.req.get_post_args](http://wiki.nginx.org/HttpLuaModule#ngx.req.get_post_args) method for fetching url-encoded POST query arguments from within Lua.
* implemented the [ngx.req.get_uri_args](http://wiki.nginx.org/HttpLuaModule#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](http://wiki.nginx.org/HttpLuaModule#ngx.parse_http_time), thanks James Hurst.
* now we allow Lua boolean and `nil` values in arguments to [ngx.say](http://wiki.nginx.org/HttpLuaModule#ngx.say), [ngx.print](http://wiki.nginx.org/HttpLuaModule#ngx.print), [ngx.log](http://wiki.nginx.org/HttpLuaModule#ngx.log) and [print](http://wiki.nginx.org/HttpLuaModule#print).
* added support for user C macros `LUA_DEFAULT_PATH` and `LUA_DEFAULT_CPATH`. for now we can only define them in `ngx_lua`'s `config` file because nginx `configure`'s `--with-cc-opt` option hates values with double-quotes in them. sigh. [ngx_openresty](http://openresty.org/) is already using this feature to bundle 3rd-party Lua libraries.
**Bug fixes**
* worked-around the "stack overflow" issue while using `luarocks.loader` and disabling [lua_code_cache](http://wiki.nginx.org/HttpLuaModule#lua_code_cache), as described as github issue #27. thanks Patrick Crosby.
* fixed the `zero size buf in output` alert while combining [lua_need_request_body](http://wiki.nginx.org/HttpLuaModule#lua_need_request_body) on + [access_by_lua](http://wiki.nginx.org/HttpLuaModule#access_by_lua)/[rewrite_by_lua](http://wiki.nginx.org/HttpLuaModule#rewrite_by_lua) + [proxy_pass](http://wiki.nginx.org/HttpProxyModule#proxy_pass)/[fastcgi_pass](http://wiki.nginx.org/HttpFcgiModule#fastcgi_pass). thanks Liseen Wan (万珣新).
* fixed issues with HTTP 1.0 HEAD requests.
* made setting `ngx.header.HEADER` 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 `ngx_http_lua_exception` in a single compilation unit. thanks @姜大炮.
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.