Repository files navigation

ngx_echo - Brings "echo", "sleep", "time", "exec" and more shell-style goodies to Nginx config file.

This module is not distributed with the Nginx source. Seethe installation instructions.

• Name
• Status
• Version
• Synopsis
• Description
• Content Handler Directives
  • echo
  • echo_duplicate
  • echo_flush
  • echo_sleep
  • echo_blocking_sleep
  • echo_reset_timer
  • echo_read_request_body
  • echo_location_async
  • echo_location
  • echo_subrequest_async
  • echo_subrequest
  • echo_foreach_split
  • echo_end
  • echo_request_body
  • echo_exec
  • echo_status
• Filter Directives
  • echo_before_body
  • echo_after_body
• Variables
  • $echo_it
  • $echo_timer_elapsed
  • $echo_request_body
  • $echo_request_method
  • $echo_client_request_method
  • $echo_client_request_headers
  • $echo_cacheable_request_uri
  • $echo_request_uri
  • $echo_incr
  • $echo_response_status
• Installation
• Compatibility
• Modules that use this module for testing
• Community
  • English Mailing List
  • Chinese Mailing List
• Report Bugs
• Source Repository
• Changes
• Test Suite
• TODO
• Getting involved
• Author
• Copyright & License
• See Also

This module is production ready.

This document describes ngx_echov0.63 released on 1 August, 2022.

location /hello{     echo"hello, world!";}

location /hello{     echo -n"hello, ";     echo"world!";}

location /timed_hello{     echo_reset_timer;     echo hello world;     echo"'hello world' takes about $echo_timer_elapsed sec.";     echo hiya igor;     echo"'hiya igor' takes about $echo_timer_elapsed sec.";}

location /echo_with_sleep{     echo hello;     echo_flush;  # ensure the client can see previous output immediately     echo_sleep   2.5;  # in sec     echo world;}

   # in the following example, accessing /echo yields   #   hello   #   world   #   blah   #   hiya   #   igorlocation /echo{       echo_before_body hello;       echo_before_body world;proxy_pass$scheme://127.0.0.1:$server_port$request_uri/more;       echo_after_body hiya;       echo_after_body igor;}location /echo/more{       echo blah;}

   # the output of /main might be   #   hello   #   world   #   took 0.000 sec for total.   # and the whole request would take about 2 sec to complete.location /main{       echo_reset_timer;       # subrequests in parallel       echo_location_async /sub1;       echo_location_async /sub2;       echo"took $echo_timer_elapsed sec for total.";}location /sub1{       echo_sleep 2;       echo hello;}location /sub2{       echo_sleep 1;       echo world;}

   # the output of /main might be   #   hello   #   world   #   took 3.003 sec for total.   # and the whole request would take about 3 sec to complete.location /main{       echo_reset_timer;       # subrequests in series (chained by CPS)       echo_location /sub1;       echo_location /sub2;       echo"took $echo_timer_elapsed sec for total.";}location /sub1{       echo_sleep 2;       echo hello;}location /sub2{       echo_sleep 1;       echo world;}

   # Accessing /dup gives   #   ------ END ------location /dup{     echo_duplicate 3"--";     echo_duplicate 1" END ";     echo_duplicate 3"--";     echo;}

   # /bighello will generate 1000,000,000 hello's.location /bighello{     echo_duplicate1000_000_000'hello';}

   # echo back the client requestlocation /echoback{     echo_duplicate 1$echo_client_request_headers;     echo"\r";     echo_read_request_body;     echo_request_body;}

   # GET /multi will yields   #   querystring: foo=Foo   #   method: POST   #   body: hi   #   content length: 2   #   ///   #   querystring: bar=Bar   #   method: PUT   #   body: hello   #   content length: 5   #   ///location /multi{       echo_subrequest_async POST'/sub' -q 'foo=Foo' -b 'hi';       echo_subrequest_async PUT'/sub' -q 'bar=Bar' -b 'hello';}location /sub{       echo"querystring: $query_string";       echo"method: $echo_request_method";       echo"body: $echo_request_body";       echo"content length: $http_content_length";       echo'///';}

   # GET /merge?/foo.js&/bar/blah.js&/yui/baz.js will merge the .js resources togetherlocation /merge{default_type'text/javascript';       echo_foreach_split'&'$query_string;           echo"/* JS File $echo_it */";           echo_location_async$echo_it;           echo;       echo_end;}

   # accessing /if?val=abc yields the "hit" output   # while /if?val=bcd yields "miss":location ^~ /if{set$res miss;if($arg_val~*'^a'){set$res hit;           echo$res;}       echo$res;}

Back to TOC

This module wraps lots of Nginx internal APIs for streaming input and output, parallel/sequential subrequests, timers and sleeping, as well as various meta data accessing.

Basically it provides various utilities that help testing and debugging of other modules by trivially emulating different kinds of faked subrequest locations.

People will also find it useful in real-world applications that need to

1. serve static contents directly from memory (loading from the Nginx config file).
2. wrap the upstream response with custom header and footer (kinda like theaddition module but with contents read directly from the config file and Nginx variables).
3. merge contents of various "Nginx locations" (i.e., subrequests) together in a single main request (usingecho_location and its friends).

This is a special dual-role module that canlazily serve as a content handler or register itself as an output filter only upon demand. By default, this module does not do anything at all.

Technically, this module has also demonstrated the following techniques that might be helpful for module writers:

1. Issue parallel subrequests directly from content handler.
2. Issue chained subrequests directly from content handler, by passing continuation along the subrequest chain.
3. Issue subrequests with all HTTP 1.1 methods and even an optional faked HTTP request body.
4. Interact with the Nginx event model directly from content handler using custom events and timers, and resume the content handler back if necessary.
5. Dual-role module that can (lazily) serve as a content handler or an output filter or both.
6. Nginx config file variable creation and interpolation.
7. Streaming output control using output_chain, flush and its friends.
8. Read client request body from the content handler, and returns back (asynchronously) to the content handler after completion.
9. Use Perl-based declarativetest suite to drive the development of Nginx C modules.

Back to TOC

Use of the following directives register this module to the current Nginx location as a content handler. If you want to use another module, like thestandard proxy module, as the content handler, use thefilter directives provided by this module.

All the content handler directives can be mixed together in a single Nginx location and they're supposed to run sequentially just as in the Bash scripting language.

Every content handler directive supports variable interpolation in its arguments (if any).

The MIME type set by thestandard default_type directive is respected by this module, as in:

location /hello{default_type text/plain;     echo hello;}

Then on the client side:

   $ curl -I'http://localhost/echo'   HTTP/1.1 200 OK   Server: nginx/0.8.20   Date: Sat, 17 Oct 2009 03:40:19 GMT   Content-Type: text/plain   Connection: keep-alive

Since thev0.22 release, all of the directives are allowed in therewrite module'sif directive block, for instance:

location ^~ /if{set$res miss;if($arg_val~*'^a'){set$res hit;         echo$res;}     echo$res;}

Back to TOC

syntax:echo [options] <string>...

default:no

context:location, location if

phase:content

Sends arguments joined by spaces, along with a trailing newline, out to the client.

Note that the data might be buffered by Nginx's underlying buffer. To force the output data flushed immediately, use theecho_flush command just afterecho, as in

    echo hello world;    echo_flush;

When no argument is specified,echo emits the trailing newline alone, just like theecho command in shell.

Variables may appear in the arguments. An example is

    echo The current request uri is$request_uri;

where$request_uri is a variable exposed by thengx_http_core_module.

This command can be used multiple times in a single location configuration, as in

location /echo{     echo hello;     echo world;}

The output on the client side looks like this

 $ curl'http://localhost/echo' hello world

Special characters like newlines (\n) and tabs (\t) can be escaped using C-style escaping sequences. But a notable exception is the dollar sign ($). As of Nginx 0.8.20, there's still no clean way to escape this character. (A work-around might be to use a$echo_dollor variable that is always evaluated to the constant$ character. This feature will possibly be introduced in a future version of this module.)

As of the echov0.28 release, one can suppress the trailing newline character in the output by using the-n option, as in

location /echo{     echo -n"hello, ";     echo"world";}

Accessing/echo gives

 $ curl'http://localhost/echo' hello, world

Leading-n in variable values won't take effect and will be emitted literally, as in

location /echo{set$opt -n;     echo$opt"hello,";     echo"world";}

This gives the following output

 $ curl'http://localhost/echo' -n hello, world

One can output leading-n literals and other options using the special-- option like this

location /echo{     echo -- -n is an option;}

which yields

 $ curl'http://localhost/echo' -n is an option

Use this form when you want to output anything leading with a dash (-).

Back to TOC

syntax:echo_duplicate <count> <string>

default:no

context:location, location if

phase:content

Outputs duplication of a string indicated by the second argument, using the count specified in the first argument.

For instance,

location /dup{       echo_duplicate 3"abc";}

will lead to the output of"abcabcabc".

Underscores are allowed in the count number, just like in Perl. For example, to emit 1000,000,000 instances of"hello, world":

location /many_hellos{       echo_duplicate1000_000_000"hello, world";}

Thecount argument could be zero, but not negative. The secondstring argument could be an empty string ("") likewise.

Unlike theecho directive, no trailing newline is appended to the result. So it's possible to "abuse" this directive as a no-trailing-newline version ofecho by using "count" 1, as in

location /echo_art{       echo_duplicate 2'---';       echo_duplicate 1' END ';  # we don't want a trailing newline here       echo_duplicate 2'---';       echo;  # we want a trailing newline here...}

You get

   ------ END ------

But use of the-n option inecho is more appropriate for this purpose.

This directive was first introduced inversion 0.11.

Back to TOC

syntax:echo_flush

default:no

context:location, location if

phase:content

Forces the data potentially buffered by underlying Nginx output filters to send immediately to the client side via socket.

Note that techically the command just emits a ngx_buf_t object withflush slot set to 1, so certain weird third-party output filter module could still block it before it reaches Nginx's (last) write filter.

This directive does not take any argument.

Consider the following example:

location /flush{      echo hello;      echo_flush;      echo_sleep 1;      echo world;}

Then on the client side, using curl to access/flush, you'll see the "hello" line immediately, but only after 1 second, the last "world" line. Without callingecho_flush in the example above, you'll most likely see no output until 1 second is elapsed due to the internal buffering of Nginx.

This directive will fail to flush the output buffer in case of subrequests get involved. Consider the following example:

location /main{       echo_location_async /sub;       echo hello;       echo_flush;}location /sub{       echo_sleep 1;}

Then the client won't see "hello" appear even ifecho_flush has been executed before the subrequest to/sub has actually started executing. The outputs of/main that are sentafterecho_location_async will be postponed and buffered firmly.

This doesnot apply to outputs sent before the subrequest initiated. For a modified version of the example given above:

location /main{       echo hello;       echo_flush;       echo_location_async /sub;}location /sub{       echo_sleep 1;}

The client will immediately see "hello" before/sub enters sleeping.

See alsoecho,echo_sleep, andecho_location_async.

Back to TOC

syntax:echo_sleep <seconds>

default:no

context:location, location if

phase:content

Sleeps for the time period specified by the argument, which is in seconds.

This operation is non-blocking on server side, so unlike theecho_blocking_sleep directive, it won't block the whole Nginx worker process.

The period might takes three digits after the decimal point and must be greater than 0.001.

An example is

location /echo_after_sleep{        echo_sleep 1.234;        echo resumed!;}

Behind the scene, it sets up a per-request "sleep" ngx_event_t object, and adds a timer using that custom event to the Nginx event model and just waits for a timeout on that event. Because the "sleep" event is per-request, this directive can work in parallel subrequests.

Back to TOC

syntax:echo_blocking_sleep <seconds>

default:no

context:location, location if

phase:content

This is a blocking version of theecho_sleep directive.

See the documentation ofecho_sleep for more detail.

Behind the curtain, it calls the ngx_msleep macro provided by the Nginx core which maps to usleep on POSIX-compliant systems.

Note that this directive will block the current Nginx worker process completely while being executed, so never use it in production environment.

Back to TOC

syntax:echo_reset_timer

default:no

context:location, location if

phase:content

Reset the timer begin time tonow, i.e., the time when this command is executed during request.

The timer begin time is default to the starting time of the current request and can be overridden by this directive, potentially multiple times in a single location. For example:

location /timed_sleep{       echo_sleep 0.03;       echo"$echo_timer_elapsed sec elapsed.";       echo_reset_timer;       echo_sleep 0.02;       echo"$echo_timer_elapsed sec elapsed.";}

The output on the client side might be

 $ curl'http://localhost/timed_sleep' 0.032 sec elapsed. 0.020 sec elapsed.

The actual figures you get on your side may vary a bit due to your system's current activities.

Invocation of this directive will force the underlying Nginx timer to get updated to the current system time (regardless the timer resolution specified elsewhere in the config file). Furthermore, references of the$echo_timer_elapsed variable will also trigger timer update forcibly.

See alsoecho_sleep and$echo_timer_elapsed.

Back to TOC

syntax:echo_read_request_body

default:no

context:location, location if

phase:content

Explicitly reads request body so that the$request_body variable will always have non-empty values (unless the body is so big that it has been saved by Nginx to a local temporary file).

Note that this might not be the original client request body because the current request might be a subrequest with a "artificial" body specified by its parent.

This directive does not generate any output itself, just likeecho_sleep.

Here's an example for echo'ing back the original HTTP client request (both headers and body are included):

location /echoback{     echo_duplicate 1$echo_client_request_headers;     echo"\r";     echo_read_request_body;     echo$request_body;}

The content of/echoback looks like this on my side (I was using Perl's LWP utility to access this location on the server):

   $ (echo hello;echo world)| lwp-request -m POST'http://localhost/echoback'   POST /echoback HTTP/1.1   TE: deflate,gzip;q=0.3   Connection: TE, close   Host: localhost   User-Agent: lwp-request/5.818 libwww-perl/5.820   Content-Length: 12   Content-Type: application/x-www-form-urlencoded   hello   world

Because/echoback is the main request,$request_body holds the original client request body.

Before Nginx 0.7.56, it makes no sense to use this directive because$request_body was first introduced in Nginx 0.7.58.

This directive itself was first introduced in the echo module'sv0.14 release.

Back to TOC

syntax:echo_location_async <location> [<url_args>]

default:no

context:location, location if

phase:content

Issue GET subrequest to the location specified (first argument) with optional url arguments specified in the second argument.

As of Nginx 0.8.20, thelocation argument doesnot support named location, due to a limitation in thengx_http_subrequest function. The same is true for its brother, theecho_location directive.

A very simple example is

location /main{     echo_location_async /sub;     echo world;}location /sub{     echo hello;}

Accessing/main gets

   hello   world

Calling multiple locations in parallel is also possible:

location /main{     echo_reset_timer;     echo_location_async /sub1;     echo_location_async /sub2;     echo"took $echo_timer_elapsed sec for total.";}location /sub1{     echo_sleep 2; # sleeps 2 sec     echo hello;}location /sub2{     echo_sleep 1; # sleeps 1 sec     echo world;}

Accessing/main yields

   $time curl'http://localhost/main'   hello   world   took 0.000 secfor total.   real  0m2.006s   user  0m0.000s   sys   0m0.004s

You can see that the main handler/main doesnot wait the subrequests/sub1 and/sub2 to complete and quickly goes on, hence the "0.000 sec" timing result. The whole request, however takes approximately 2 sec in total to complete because/sub1 and/sub2 run in parallel (or "concurrently" to be more accurate).

If you useecho_blocking_sleep in the previous example instead, then you'll get the same output, but with 3 sec total response time, because "blocking sleep" blocks the whole Nginx worker process.

Locations can also take an optional querystring argument, for instance

location /main{     echo_location_async /sub'foo=Foo&bar=Bar';}location /sub{     echo$arg_foo$arg_bar;}

Accessing/main yields

   $ curl'http://localhost/main'   Foo Bar

Querystrings isnot allowed to be concatenated onto thelocation argument with "?" directly, for example,/sub?foo=Foo&bar=Bar is an invalid location, and shouldn't be fed as the first argument to this directive.

Technically speaking, this directive is an example that Nginx content handler issues one or more subrequests directly. AFAIK, thefancyindex module also does such kind of things ;)

Nginx named locations like@foo isnot supported here.

This directive is logically equivalent to the GET version ofecho_subrequest_async. For example,

   echo_location_async /foo'bar=Bar';

is logically equivalent to

   echo_subrequest_async GET /foo -q'bar=Bar';

But calling this directive is slightly faster than callingecho_subrequest_async usingGET because we don't have to parse the HTTP method names likeGET and options like-q.

This directive is first introduced inversion 0.09 of this module and requires at least Nginx 0.7.46.

Back to TOC

syntax:echo_location <location> [<url_args>]

default:no

context:location, location if

phase:content

Just like theecho_location_async directive, butecho_location issues subrequestsin series rather than in parallel. That is, the content handler directives following this directive won't be executed until the subrequest issued by this directive completes.

The final response body is almost always equivalent to the case whenecho_location_async is used instead, only if timing variables is used in the outputs.

Consider the following example:

location /main{     echo_reset_timer;     echo_location /sub1;     echo_location /sub2;     echo"took $echo_timer_elapsed sec for total.";}location /sub1{     echo_sleep 2;     echo hello;}location /sub2{     echo_sleep 1;     echo world;}

The location/main above will take for total 3 sec to complete (compared to 2 sec ifecho_location_async is used instead here). Here's the result in action on my machine:

   $ curl'http://localhost/main'   hello   world   took 3.003 secfor total.   real  0m3.027s   user  0m0.020s   sys   0m0.004s

This directive is logically equivalent to the GET version ofecho_subrequest. For example,

   echo_location /foo'bar=Bar';

is logically equivalent to

   echo_subrequest GET /foo -q'bar=Bar';

But calling this directive is slightly faster than callingecho_subrequest usingGET because we don't have to parse the HTTP method names likeGET and options like-q.

Behind the scene, it creates anngx_http_post_subrequest_t object as acontinuation and passes it into thengx_http_subrequest function call. Nginx will later reopen this "continuation" in the subrequest'sngx_http_finalize_request function call. We resumes the execution of the parent-request's content handler and starts to run the next directive (command) if any.

Nginx named locations like@foo isnot supported here.

This directive was first introduced in therelease v0.12.

See alsoecho_location_async for more details about the meaning of the arguments.

Back to TOC

syntax:echo_subrequest_async <HTTP_method> <location> [-q <url_args>] [-b <request_body>] [-f <request_body_path>]

default:no

context:location, location if

phase:content

Initiate an asynchronous subrequest using HTTP method, an optional url arguments (or querystring) and an optional request body which can be defined as a string or as a path to a file which contains the body.

This directive is very much like a generalized version of theecho_location_async directive.

Here's a small example demonstrating its usage:

location /multi{     # body defined as string     echo_subrequest_async POST'/sub' -q 'foo=Foo' -b 'hi';     # body defined as path to a file, relative to nginx prefix path if not absolute     echo_subrequest_async PUT'/sub' -q 'bar=Bar' -f '/tmp/hello.txt';}location /sub{     echo"querystring: $query_string";     echo"method: $echo_request_method";     echo"body: $echo_request_body";     echo"content length: $http_content_length";     echo'///';}

Then on the client side:

   $echo -n hello> /tmp/hello.txt   $ curl'http://localhost/multi'   querystring: foo=Foo   method: POST   body: hi   content length: 2   ///   querystring: bar=Bar   method: PUT   body: hello   content length: 5   ///

Here's more funny example using the standardproxy module to handle the subrequest:

location /main{     echo_subrequest_async POST /sub -b'hello, world';}location /sub{proxy_pass$scheme://127.0.0.1:$server_port/proxied;}location /proxied{     echo"method: $echo_request_method.";     # we need to read body explicitly here...or $echo_request_body     #   will evaluate to empty ("")     echo_read_request_body;     echo"body: $echo_request_body.";}

Then on the client side, we can see that

   $ curl'http://localhost/main'   method: POST.   body: hello, world.

Nginx named locations like@foo isnot supported here.

This directive takes several options:

-q <url_args>        Specify the URL arguments (or URL querystring) for the subrequest.-f <path>            Specify the path for the file whose content will be serve as the                     subrequest's request body.-b <data>            Specify the request body data

This directive was first introduced in therelease v0.15.

The-f option to define a file path for the body was introduced in therelease v0.35.

See also theecho_subrequest andecho_location_async directives.

Back to TOC

syntax:echo_subrequest <HTTP_method> <location> [-q <url_args>] [-b <request_body>] [-f <request_body_path>]

default:no

context:location, location if

phase:content

This is the synchronous version of theecho_subrequest_async directive. And just likeecho_location, it does not block the Nginx worker process (whileecho_blocking_sleep does), rather, it uses continuation to pass control along the subrequest chain.

Seeecho_subrequest_async for more details.

Nginx named locations like@foo isnot supported here.

This directive was first introduced in therelease v0.15.

Back to TOC

syntax:echo_foreach_split <delimiter> <string>

default:no

context:location, location if

phase:content

Split the second argumentstring using the delimiter specified in the first argument, and then iterate through the resulting items. For instance:

location /loop{     echo_foreach_split','$arg_list;       echo"item: $echo_it";     echo_end;}

Accessing /main yields

   $ curl'http://localhost/loop?list=cat,dog,mouse'   item: cat   item: dog   item: mouse

As seen in the previous example, this directive should always be accompanied by anecho_end directive.

Parallelecho_foreach_split loops are allowed, but nested ones are currently forbidden.

Thedelimiter argument could containmultiple arbitrary characters, like

   # this outputs "cat\ndog\nmouse\n"   echo_foreach_split --'-a-' 'cat-a-dog-a-mouse';     echo$echo_it;   echo_end;

Logically speaking, this looping structure is just theforeach loop combined with asplit function call in Perl (using the previous example):

foreach (split',',$arg_list) {print"item$_\n";    }

People will also find it useful in merging multiple.js or.css resources into a whole. Here's an example:

location /merge{default_type'text/javascript';       echo_foreach_split'&'$query_string;           echo"/* JS File $echo_it */";           echo_location_async$echo_it;           echo;       echo_end;}

Then accessing /merge to merge the.js resources specified in the query string:

   $ curl'http://localhost/merge?/foo/bar.js&/yui/blah.js&/baz.js'

One can also use third-party Nginx cache module to cache the merged response generated by the/merge location in the previous example.

This directive was first introduced in therelease v0.17.

Back to TOC

syntax:echo_end

default:no

context:location, location if

phase:content

This directive is used to terminate the body of looping and conditional control structures likeecho_foreach_split.

This directive was first introduced in therelease v0.17.

Back to TOC

syntax:echo_request_body

default:no

context:location, location if

phase:content

Outputs the contents of the request body previous read.

Behind the scene, it's implemented roughly like this:

if (r->request_body&&r->request_body->bufs) {returnngx_http_output_filter(r,r->request_body->bufs);   }

Unlike the$echo_request_body and $request_body variables, this directive will show the whole request body even if some parts or all parts of it are saved in temporary files on the disk.

It is a "no-op" if no request body has been read yet.

This directive was first introduced in therelease v0.18.

See alsoecho_read_request_body and thechunkin module.

Back to TOC

syntax:echo_exec <location> [<query_string>]

syntax:echo_exec <named_location>

default:no

context:location, location if

phase:content

Does an internal redirect to the location specified. An optional query string can be specified for normal locations, as in

location /foo{       echo_exec /bar weight=5;}location /bar{       echo$arg_weight;}

Or equivalently

location /foo{       echo_exec /bar?weight=5;}location /bar{       echo$arg_weight;}

Named locations are also supported. Here's an example:

location /foo{       echo_exec @bar;}location @bar{       # you'll get /foo rather than @bar       #  due to a potential bug in nginx.       echo$echo_request_uri;}

But query string (if any) will always be ignored for named location redirects due to a limitation in thengx_http_named_location function.

Never try to echo things before theecho_exec directive or you won't see the proper response of the location you want to redirect to. Because any echoing will cause the original location handler to send HTTP headers before the redirection happens.

Technically speaking, this directive exposes the Nginx internal API functionsngx_http_internal_redirect andngx_http_named_location.

This directive was first introduced in thev0.21 release.

Back to TOC

syntax:echo_status <status-num>

default:echo_status 200

context:location, location if

phase:content

Specify the default response status code. Default to200. This directive is declarative and the relative order with other echo-like directives is not important.

Here is an example,

location = /bad{     echo_status404;     echo"Something is missing...";}

then we get a response like this:

HTTP/1.1 404 Not FoundServer: nginx/1.2.1Date: Sun, 24 Jun 2012 03:58:18 GMTContent-Type: text/plainTransfer-Encoding: chunkedConnection: keep-aliveSomething is missing...

This directive was first introduced in thev0.40 release.

Back to TOC

Use of the following directives trigger the filter registration of this module. By default, no filter will be registered by this module.

Every filter directive supports variable interpolation in its arguments (if any).

Back to TOC

syntax:echo_before_body [options] [argument]...

default:no

context:location, location if

phase:output filter

It's the filter version of theecho directive, and prepends its output to the beginning of the original outputs generated by the underlying content handler.

An example is

location /echo{     echo_before_body hello;proxy_pass$scheme://127.0.0.1:$server_port$request_uri/more;}location /echo/more{     echo world}

Accessing/echo from the client side yields

   hello   world

In the previous sample, we borrow thestandard proxy module to serve as the underlying content handler that generates the "main contents".

Multiple instances of this filter directive are also allowed, as in:

location /echo{     echo_before_body hello;     echo_before_body world;     echo !;}

On the client side, the output is like

   $ curl'http://localhost/echo'   hello   world!

In this example, we also use thecontent handler directives provided by this module as the underlying content handler.

This directive also supports the-n and-- options like theecho directive.

This directive can be mixed with its brother directiveecho_after_body.

Back to TOC

syntax:echo_after_body [argument]...

default:no

context:location, location if

phase:output filter

It's very much like theecho_before_body directive, butappends its output to the end of the original outputs generated by the underlying content handler.

Here's a simple example:

location /echo{     echo_after_body hello;proxy_passhttp://127.0.0.1:$server_port$request_uri/more;}location /echo/more{     echo world}

Accessing/echo from the client side yields

  world  hello

Multiple instances are allowed, as in:

location /echo{     echo_after_body hello;     echo_after_body world;     echo i;     echo say;}

The output on the client side while accessing the/echo location looks like

  i  say  hello  world

This directive also supports the-n and-- options like theecho directive.

This directive can be mixed with its brother directiveecho_before_body.

Back to TOC

Back to TOC

This is a "topic variable" used byecho_foreach_split, just like the$_ variable in Perl.

Back to TOC

This variable holds the seconds elapsed since the start of the current request (might be a subrequest though) or the last invocation of theecho_reset_timer command.

The timing result takes three digits after the decimal point.

References of this variable will force the underlying Nginx timer to update to the current system time, regardless the timer resolution settings elsewhere in the config file, just like theecho_reset_timer directive.

Back to TOC

Evaluates to the current (sub)request's request body previously read if no part of the body has been saved to a temporary file. To always show the request body even if it's very large, use theecho_request_body directive.

Back to TOC

Evaluates to the HTTP request method of the current request (it can be a subrequest).

Behind the scene, it just takes the string data stored inr->method_name.

Compare it to the$echo_client_request_method variable.

At least for Nginx 0.8.20 and older, the$request_method variable provided by thehttp core module is actually doing what our$echo_client_request_method is doing.

This variable was first introduced in ourv0.15 release.

Back to TOC

Always evaluates to the main request's HTTP method even if the current request is a subrequest.

Behind the scene, it just takes the string data stored inr->main->method_name.

Compare it to the$echo_request_method variable.

This variable was first introduced in ourv0.15 release.

Back to TOC

Evaluates to the original client request's headers.

Just as the name suggests, it will always take the main request (or the client request) even if it's currently executed in a subrequest.

A simple example is below:

location /echoback{      echo"headers are:"      echo$echo_client_request_headers;}

Accessing/echoback yields

   $ curl'http://localhost/echoback'   headers are   GET /echoback HTTP/1.1   User-Agent: curl/7.18.2 (i486-pc-linux-gnu) libcurl/7.18.2 OpenSSL/0.9.8g   Host: localhost:1984   Accept:*/*

Behind the scene, it recoversr->main->header_in (or the large header buffers, if any) on the C level and does not construct the headers itself by traversing parsed results in the request object.

This varible is always evaluated to an empty value in HTTP/2 requests for now due to the current implementation.

This variable was first introduced inversion 0.15.

Back to TOC

Evaluates to the parsed form of the URI (usually led by/) of the current (sub-)request. Unlike the$echo_request_uri variable, it is cacheable.

See$echo_request_uri for more details.

This variable was first introduced inversion 0.17.

Back to TOC

Evaluates to the parsed form of the URI (usually led by/) of the current (sub-)request. Unlike the$echo_cacheable_request_uri variable, it isnot cacheable.

This is quite different from the$request_uri variable exported by thengx_http_core_module, because$request_uri is theunparsed form of the current request's URI.

This variable was first introduced inversion 0.17.

Back to TOC

It is a counter that always generate the current counting number, starting from 1. The counter is always associated with the main request even if it is accessed within a subrequest.

Consider the following example

location /main{     echo"main pre: $echo_incr";     echo_location_async /sub;     echo_location_async /sub;     echo"main post: $echo_incr";}location /sub{     echo"sub: $echo_incr";}

Accessing/main yields

main pre: 1sub: 3sub: 4main post: 2

This directive was first introduced in thev0.18 release.

Back to TOC

Evaluates to the status code of the current (sub)request, null if not any.

Behind the scene, it's just the textual representation ofr->headers_out->status.

This directive was first introduced in thev0.23 release.

Back to TOC

You're recommended to install this module (as well as the Nginx core and many other goodies) via theOpenResty bundle. Seethe detailed instructions for downloading and installing OpenResty into your system. This is the easiest and most safe way to set things up.

Alternatively, you can install this module manually with the Nginx source:

Grab the nginx source code fromnginx.org, for example,the version 1.11.2 (seenginx compatibility), and then build the source with this module:

 $ wget'http://nginx.org/download/nginx-1.11.2.tar.gz' $ tar -xzvf nginx-1.11.2.tar.gz $cd nginx-1.11.2/# Here we assume you would install you nginx under /opt/nginx/. $ ./configure --prefix=/opt/nginx \     --add-module=/path/to/echo-nginx-module $ make -j2 $ make install

Download the latest version of the release tarball of this module fromecho-nginx-module file list.

Starting from NGINX 1.9.11, you can also compile this module as a dynamic module, by using the--add-dynamic-module=PATH option instead of--add-module=PATH on the./configure command line above. And then you can explicitly load the module in yournginx.conf via theload_moduledirective, for example,

load_module /path/to/modules/ngx_http_echo_module.so;

Also, this module is included and enabled by default in theOpenResty bundle.

Back to TOC

The following versions of Nginx should work with this module:

• 1.16.x
• 1.15.x (last tested: 1.15.8)
• 1.14.x
• 1.13.x (last tested: 1.13.6)
• 1.12.x
• 1.11.x (last tested: 1.11.2)
• 1.10.x
• 1.9.x (last tested: 1.9.15)
• 1.8.x
• 1.7.x (last tested: 1.7.10)
• 1.6.x
• 1.5.x (last tested: 1.5.12)
• 1.4.x (last tested: 1.4.4)
• 1.3.x (last tested: 1.3.7)
• 1.2.x (last tested: 1.2.9)
• 1.1.x (last tested: 1.1.5)
• 1.0.x (last tested: 1.0.11)
• 0.9.x (last tested: 0.9.4)
• 0.8.x (last tested: 0.8.54)
• 0.7.x >= 0.7.21 (last tested: 0.7.68)

In particular,

• the directiveecho_location_async and its brotherecho_subrequest_async donot work with0.7.x < 0.7.46.
• theecho_after_body directive doesnot work at all with nginx< 0.8.7.
• theecho_sleep directive cannot be used afterecho_location orecho_subrequest for nginx< 0.8.11.

Earlier versions of Nginx like 0.6.x and 0.5.x willnot work at all.

If you find that any particular version of Nginx above 0.7.21 does not work with this module, please considerreporting a bug.

Back to TOC

The following modules take advantage of thisecho module in their test suite:

• Thememc module that supports almost the whole memcached TCP protocol.
• Thechunkin module that adds HTTP 1.1 chunked input support to Nginx.
• Theheaders_more module that allows you to add, set, and clear input and output headers under the conditions that you specify.
• Theecho module itself.

Please mail me other modules that useecho in any form and I'll add them to the list above :)

Back to TOC

Back to TOC

Theopenresty-en mailing list is for English speakers.

Back to TOC

Theopenresty mailing list is for Chinese speakers.

Back to TOC

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 theissue tracking interface provided by GitHub,
2. or send a bug report, questions, or even patches to theOpenResty Community.

Back to TOC

Available on github atopenresty/echo-nginx-module.

Back to TOC

The changes of every release of this module can be obtained from the OpenResty bundle's change logs:

http://openresty.org/#Changes

Back to TOC

This module comes with a Perl-driven test suite. Thetest cases aredeclarative too. Thanks to theTest::Nginx module in the Perl world.

To run it on your side:

 $ PATH=/path/to/your/nginx-with-echo-module:$PATH prove -r t

You need to terminate any Nginx processes before running the test suite if you have changed the Nginx server binary.

Because a single nginx server (by default,localhost:1984) is used across all the test scripts (.t files), it's meaningless to run the test suite in parallel by specifying-jN when invoking theprove utility.

Some parts of the test suite requires standard modulesproxy,rewrite andSSI to be enabled as well when building Nginx.

Back to TOC

• Fix theecho_after_body directive in subrequests.
• Add directivesecho_read_client_request_body andecho_request_headers.
• Add new directiveecho_log to use Nginx's logging facility directly from the config file and specific loglevel can be specified, as in

   echo_log debug"I am being called.";

• Add support for options-h and-t toecho_subrequest_async andecho_subrequest. For example

   echo_subrequest POST /sub -q'foo=Foo&bar=Bar' -b 'hello' -t 'text/plan' -h 'X-My-Header: blah blah'

• Add options to control whether a subrequest should inherit cached variables from its parent request (i.e. the current request that is calling the subrequest in question). Currently none of the subrequests issued by this module inherit the cached variables from the parent request.
• Add new variable$echo_active_subrequests to showr->main->count - 1.
• Add theecho_file andecho_cached_file directives.
• Add new varaible$echo_request_headers to accompany the existing$echo_client_request_headers variable.
• Add new directiveecho_foreach, as in

   echo_foreach'cat' 'dog' 'mouse';     echo_location_async"/animals/$echo_it";   echo_end;

• Add new directiveecho_foreach_range, as in

   echo_foreach_range'[1..100]' '[a-zA-z0-9]';     echo_location_async"/item/$echo_it";   echo_end;

• Add new directiveecho_repeat, as in

   echo_repeat10$i{       echo"Page $i";       echo_location"/path/to/page/$i";}

This is just another way of saying

   echo_foreach_range$i [1..10];       echo"Page $i";       echo_location"/path/to/page/$i";   echo_end;

Thanks Marcus Clyne for providing this idea.

• Add new variable $echo_random which always returns a random non-negative integer with the lower/upper limit specified by the new directivesecho_random_min andecho_random_max. For example,

   echo_random_min10   echo_random_max200   echo"random number: $echo_random";

Thanks Marcus Clyne for providing this idea.

Back to TOC

You'll be very welcomed to submit patches to theauthor or just ask for a commit bit to thesource repository on GitHub.

Back to TOC

Yichun "agentzh" Zhang (章亦春)<agentzh@gmail.com>, OpenResty Inc.

This wiki page is also maintained by the author himself, and everybody is encouraged to improve this page as well.

Back to TOC

Copyright (c) 2009-2018, Yichun "agentzh" Zhang (章亦春)agentzh@gmail.com, OpenResty Inc.

This module is licensed under the terms of the BSD license.

Redistribution and use in source and binary forms, with or withoutmodification, are permitted provided that the following conditionsare 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 NOTLIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FORA PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHTHOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITEDTO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, ORPROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OFLIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDINGNEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THISSOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

Back to TOC

• The originalblog post about this module's initial development.
• The standardaddition filter module.
• The standardproxy module.
• TheOpenResty bundle.

Back to TOC