Name

ngx_replace_filter - Streaming regular expression replacement in response bodies.

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

Table of Contents

• Name
• Status
• Synopsis
• Description
• Directives
  • replace_filter
  • replace_filter_types
  • replace_filter_max_buffered_size
  • replace_filter_last_modified
  • replace_filter_skip
• Installation
• Trouble Shooting
• TODO
• Community
  • English Mailing List
  • Chinese Mailing List
• Bugs and Patches
• Author
• Copyright and License
• See Also

Status

This module is already quite usable though still at the early phase of development and is considered experimental.

Synopsis

    location /t {
        default_type text/html;
        echo abc;
        replace_filter 'ab|abc' X;
    }

    location / {
        # proxy_pass/fastcgi_pass/...

        # caseless global substitution:
        replace_filter '\d+' 'blah blah' 'ig';
        replace_filter_types text/plain text/css;
    }

    location /a {
        # proxy_pass/fastcgi_pass/root/...

        # remove line-leading spaces and line-trailing spaces,
        # as well as blank lines:
        replace_filter '^\s+|\s+$' '' g;
    }

    location /b {
        # proxy_pass/fastcgi_pass/root/...

        # only remove line-leading spaces and line-trailing spaces:
        replace_filter '^[ \f\t]+|[ \f\t]+$' '' g;
    }

    location ~ '\.cpp$' {
        # proxy_pass/fastcgi_pass/root/...

        replace_filter_types text/plain;

        # skip C/C++ string literals:
        replace_filter "'(?:\\\\[^\n]|[^'\n])*'" $& g;
        replace_filter '"(?:\\\\[^\n]|[^"\n])*"' $& g;

        # remove all those ugly C/C++ comments:
        replace_filter '/\*.*?\*/|//[^\n]*' '' g;
    }

Description

This Nginx output filter module tries to do regular expression substitutions in a non-buffered manner wherever possible.

This module does not use traditional backtracking regular expression engines like PCRE, rather, it uses the new sregex library implemented by the author himself, which was designed with streaming processing in mind from the very beginning:

A good common subset of Perl 5 regular expressions is supported by sregex. For the complete feature list, check out sregex's documentation:

https://github.com/agentzh/sregex#syntax-supported

Response body data is only buffered when absolutely necessary, like facing an incomplete capture that belongs to a possible match near the data chunk boundaries.

Back to TOC

Directives

Back to TOC

replace_filter

syntax: replace_filter <regex> <replace>

syntax: replace_filter <regex> <replace> <options>

default: no

context: http, server, location, location if

phase: output body filter

Specifies the regex pattern and text to be replaced, with optional regex flags.

By default, the filter stops matching after the first match is found. This behavior can be changed by specifying the g regex option.

The following regex options are supported:

• g

  for global search and substitution (default off)

• i

  for case-insensitive matching (default off)

Multiple options can be combined in a single string argument, for example:

    replace_filter hello hiya ig;

Nginx variables can be interpolated into the text to be replaced, for example:

    replace_filter \w+ "[$foo,$bar]";

If you want to use the literal dollar sign character ($), use the $$ sequence for that, for instance:

    replace_filter \w "$$";

Use of submatch capturing variables like $&, $1, $2, and etc are also supported, for example,

    replace_filter [bc]|d [$&-$1-$2] g;

The semantics of the submatch capturing variables is exactly the same as in the Perl 5 language.

Multiple replace_filter directives in the same scope is also supported. All the patterns will be applied at the same time as in a tokenizer. We will not use the longest token match semantics, but rather, patterns will be prioritized according to their order in the configure file.

Here is an example for removing all the C/C++ comments from a C/C++ source code file:

    replace_filter "'(?:\\\\[^\n]|[^'\n])*'" $& g;
    replace_filter '"(?:\\\\[^\n]|[^"\n])*"' $& g;
    replace_filter '/\*.*?\*/|//[^\n]*' '' g;

When the Content-Encoding response header is not empty (like gzip), the response body will always remain intact. So usually you want to disable the gzip compression in your backend servers' responses by adding the following line to your nginx.conf if you are the ngx_proxy module:

    proxy_set_header Accept-Encoding '';

Your responses can still be gzip compressed on the Nginx server level though.

Back to TOC

replace_filter_types

syntax: replace_filter_types <mime-type> ...

default: replace_filter_types text/html

context: http, server, location, location if

phase: output body filter

Specify one or more MIME types (in the Content-Type response header) to be processed.

By default, only text/html typed responses are processed.

Back to TOC

replace_filter_max_buffered_size

syntax: replace_filter_max_buffered_size <size>

default: replace_filter_max_buffered_size 8k

context: http, server, location, location if

phase: output body filter

Limits the total size of the data buffered by the module at runtime. Default to 8k.

When the limit is reached, replace_filter will immediately stop processing and leave all the remaining response body data intact.

Back to TOC

replace_filter_last_modified

syntax: replace_filter_last_modifiled keep | clear

default: replace_filter_last_modified clear

context: http, server, location, location if

phase: output body filter

Controls how to deal with the existing Last-Modified response header.

By default, this module will clear the Last-Modified response header if there is any. You can specify

    replace_filter_last_modified keep;

to always keep the original Last-Modified response header.

Back to TOC

replace_filter_skip

syntax: replace_filter_skip $var

default: no

context: http, server, location, location if

phase: output header filter

This directive controls whether to skip all the replace_filter rules on a per-request basis.

Both constant values or strings containing NGINX variables are supported.

When the value is evaluated to an empty value ("") or the value "0" in the request output header phase, no replace_filter rules will be skipped for the current request. Otherwise all the replace_filter rules will be skipped for the current request.

Below is a trivial example for this:

set $skip '';
location /t {
    content_by_lua '
        ngx.var.skip = 1
        ngx.say("abcabd")
    ';
    replace_filter_skip $skip;
    replace_filter abcabd X;
}

Back to TOC

Installation

You need to install the sregex library first:

https://github.com/agentzh/sregex

And then rebuild your Nginx like this:

    ./configure --add-module=/path/to/replace-filter-nginx-module

If sregex is not installed to the default prefix (i.e., /usr/local), then you should specify the locations of your sregex installation via the SREGEX_INC and SREGEX_LIB environments before running the ./configure script, as in

    export SREGEX_INC=/opt/sregex/include
    export SREGEX_LIB=/opt/sregex/lib

assuming that your sregex is installed to the prefix /opt/sregex.

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 your nginx.conf via the load_module directive, for example,

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

Back to TOC

Trouble Shooting

• If you are seeing the error "error while loading shared libraries: libsregex.so.0: cannot open shared object file: No such file or directory" while starting nginx, then it means that the installation path of your libsregex library is not in your system's default library search path. You can solve this issue by passing the option --with-ld-opt='-Wl,-rpath,/usr/local/lib' to nginx's ./configure command. Alternatively, you can just add the path of your libsregex.so.0 to the LD_LIBRARY_PATH environment value before starting your nginx server.

Back to TOC

TODO

• optimize the special case for verbatim substitutions, i.e., replace_filter <regex> $&;.
• implement the replace_filter_skip $var directive to control whether to enable the filter on the fly.
• reduce the amount of data that has to be buffered for when an partial match is already found.
• recycle the memory blocks used to buffer the pending capture data and "complex values" for replacement.
• allow use of inlined Lua code as the replacement argument of the replace_filter directive to generate the text to be replaced on-the-fly.

Back to TOC

Community

Back to TOC

English Mailing List

The openresty-en mailing list is for English speakers.

Back to TOC

Chinese Mailing List

The openresty mailing list is for Chinese speakers.

Back to TOC

Bugs and Patches

Please submit bug reports, wishlists, or patches by

1. creating a ticket on the GitHub Issue Tracker,
2. or posting to the OpenResty community.

Back to TOC

Author

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

Back to TOC

Copyright and License

This module is licensed under the BSD license.

Copyright (C) 2012-2017, by Yichun "agentzh" Zhang (章亦春), OpenResty Inc.

All rights reserved.

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

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

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

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

Back to TOC

See Also

• agentzh's sregex library: https://github.com/agentzh/sregex
• The standard ngx_sub_filter module: http://nginx.org/en/docs/http/ngx_http_sub_module.html
• Slides for my talk "sregex: matching Perl 5 regexes on data streams": http://agentzh.org/misc/slides/yapc-na-2013-sregex.pdf Back to TOC