Sanitize untrusted HTML (to prevent XSS) with a configuration specified by a Whitelist.

---

xss is a module used to filter input from users to prevent XSS attacks. (What is XSS attack?)

Project Homepage: http://jsxss.com

Try Online: http://jsxss.com/en/try.html

中文版文档

---

Features

• Specifies HTML tags and their attributes allowed with whitelist
• Handle any tags or attributes using custom function.

Reference

• XSS Filter Evasion Cheat Sheet
• Data URI scheme
• XSS with Data URI Scheme

Benchmark (for references only)

• the xss module: 22.53 MB/s
• xss() function from module validator@0.3.7: 6.9 MB/s

For test code please refer to benchmark directory.

They are using xss module

• nodeclub - A Node.js bbs using MongoDB - https://github.com/cnodejs/nodeclub
• cnpmjs.org - Private npm registry and web for Enterprise - https://github.com/cnpm/cnpmjs.org
• cocalc.com - Collaborative Calculation and Data Science - https://cocalc.com

Install

NPM

1npm install xss

Bower

1bower install xss

Or

1bower install https://github.com/leizongmin/js-xss.git

Usages

On Node.js

1var xss = require("xss");
2var html = xss('<script>alert("xss");</script>');
3console.log(html);

On Browser

Shim mode (reference file test/test.html):

1<script src="https://rawgit.com/leizongmin/js-xss/master/dist/xss.js"></script>
2<script>
3  // apply function filterXSS in the same way
4  var html = filterXSS('<script>alert("xss");</scr' + "ipt>");
5  alert(html);
6</script>

AMD mode - shim:

1<script>
2  require.config({
3    baseUrl: "./",
4    paths: {
5      xss: "https://rawgit.com/leizongmin/js-xss/master/dist/xss.js",
6    },
7    shim: {
8      xss: { exports: "filterXSS" },
9    },
10  });
11  require(["xss"], function (xss) {
12    var html = xss('<script>alert("xss");</scr' + "ipt>");
13    alert(html);
14  });
15</script>

Notes: please don't use the URL https://rawgit.com/leizongmin/js-xss/master/dist/xss.js in production environment.

Command Line Tool

Process File

You can use the xss command line tool to process a file. Usage:

1xss -i <input_file> -o <output_file>

Example:

1xss -i origin.html -o target.html

Active Test

Run the following command, them you can type HTML code in the command-line, and check the filtered output:

1xss -t

For more details, please run $ xss -h to see it.

Custom filter rules

When using the xss() function, the second parameter could be used to specify custom rules:

1options = {}; // Custom rules
2html = xss('<script>alert("xss");</script>', options);

To avoid passing options every time, you can also do it in a faster way by creating a FilterXSS instance:

1options = {}; // Custom rules
2myxss = new xss.FilterXSS(options);
3// then apply myxss.process()
4html = myxss.process('<script>alert("xss");</script>');

Details of parameters in options would be described below.

Whitelist

By specifying a whiteList, e.g. { 'tagName': [ 'attr-1', 'attr-2' ] }. Tags and attributes not in the whitelist would be filter out. For example:

1// only tag a and its attributes href, title, target are allowed
2var options = {
3  whiteList: {
4    a: ["href", "title", "target"],
5  },
6};
7// With the configuration specified above, the following HTML:
8// <a href="#" onclick="hello()"><i>Hello</i></a>
9// would become:
10// <a href="#">&lt;i&gt;Hello&lt;/i&gt;</a>

For the default whitelist, please refer xss.whiteList.

allowList is also supported, and has the same function as whiteList.

Customize the handler function for matched tags

By specifying the handler function with onTag:

1function onTag(tag, html, options) {
2  // tag is the name of current tag, e.g. 'a' for tag <a>
3  // html is the HTML of this tag, e.g. '<a>' for tag <a>
4  // options is some addition informations:
5  //   isWhite    boolean, whether the tag is in whitelist
6  //   isClosing  boolean, whether the tag is a closing tag, e.g. true for </a>
7  //   position        integer, the position of the tag in output result
8  //   sourcePosition  integer, the position of the tag in input HTML source
9  // If a string is returned, the current tag would be replaced with the string
10  // If return nothing, the default measure would be taken:
11  //   If in whitelist: filter attributes using onTagAttr, as described below
12  //   If not in whitelist: handle by onIgnoreTag, as described below
13}

Customize the handler function for attributes of matched tags

By specifying the handler function with onTagAttr:

1function onTagAttr(tag, name, value, isWhiteAttr) {
2  // tag is the name of current tag, e.g. 'a' for tag <a>
3  // name is the name of current attribute, e.g. 'href' for href="#"
4  // isWhiteAttr whether the attribute is in whitelist
5  // If a string is returned, the attribute would be replaced with the string
6  // If return nothing, the default measure would be taken:
7  //   If in whitelist: filter the value using safeAttrValue as described below
8  //   If not in whitelist: handle by onIgnoreTagAttr, as described below
9}

Customize the handler function for tags not in the whitelist

By specifying the handler function with onIgnoreTag:

1function onIgnoreTag(tag, html, options) {
2  // Parameters are the same with onTag
3  // If a string is returned, the tag would be replaced with the string
4  // If return nothing, the default measure would be taken (specifies using
5  // escape, as described below)
6}

Customize the handler function for attributes not in the whitelist

By specifying the handler function with onIgnoreTagAttr:

1function onIgnoreTagAttr(tag, name, value, isWhiteAttr) {
2  // Parameters are the same with onTagAttr
3  // If a string is returned, the value would be replaced with this string
4  // If return nothing, then keep default (remove the attribute)
5}

Customize escaping function for HTML

By specifying the handler function with escapeHtml. Following is the default function (Modification is not recommended):

1function escapeHtml(html) {
2  return html.replace(/</g, "&lt;").replace(/>/g, "&gt;");
3}

Customize escaping function for value of attributes

By specifying the handler function with safeAttrValue:

1function safeAttrValue(tag, name, value) {
2  // Parameters are the same with onTagAttr (without options)
3  // Return the value as a string
4}

Customize output attribute value syntax for HTML

By specifying a singleQuotedAttributeValue. Use true for '. Otherwise default " will be used

1var options = {
2  singleQuotedAttributeValue: true,
3};
4// With the configuration specified above, the following HTML:
5// <a href="#">Hello</a>
6// would become:
7// <a href='#'>Hello</a>

Customize CSS filter

If you allow the attribute style, the value will be processed by cssfilter module. The cssfilter module includes a default css whitelist. You can specify the options for cssfilter module like this:

1myxss = new xss.FilterXSS({
2  css: {
3    whiteList: {
4      position: /^fixed|relative$/,
5      top: true,
6      left: true,
7    },
8  },
9});
10html = myxss.process('<script>alert("xss");</script>');

If you don't want to filter out the style content, just specify false to the css option:

1myxss = new xss.FilterXSS({
2  css: false,
3});

For more help, please see https://github.com/leizongmin/js-css-filter

Quick Start

Filter out tags not in the whitelist

By using stripIgnoreTag parameter:

• true filter out tags not in the whitelist
• false: by default: escape the tag using configured escape function

Example:

If stripIgnoreTag = true is set, the following code:

1code:
2<script>
3  alert(/xss/);
4</script>

would output filtered:

1code:alert(/xss/);

Filter out tags and tag bodies not in the whitelist

By using stripIgnoreTagBody parameter:

• false|null|undefined by default: do nothing
• '*'|true: filter out all tags not in the whitelist
• ['tag1', 'tag2']: filter out only specified tags not in the whitelist

Example:

If stripIgnoreTagBody = ['script'] is set, the following code:

1code:
2<script>
3  alert(/xss/);
4</script>

would output filtered:

1code:

Filter out HTML comments

By using allowCommentTag parameter:

• true: do nothing
• false by default: filter out HTML comments

Example:

If allowCommentTag = false is set, the following code:

1code:<!-- something -->
2END

would output filtered:

1code: END

Examples

Allow attributes of whitelist tags start with data-

1var source = '<div a="1" b="2" data-a="3" data-b="4">hello</div>';
2var html = xss(source, {
3  onIgnoreTagAttr: function (tag, name, value, isWhiteAttr) {
4    if (name.substr(0, 5) === "data-") {
5      // escape its value using built-in escapeAttrValue function
6      return name + '="' + xss.escapeAttrValue(value) + '"';
7    }
8  },
9});
10
11console.log("%s\nconvert to:\n%s", source, html);

Result:

1<div a="1" b="2" data-a="3" data-b="4">hello</div>
2convert to:
3<div data-a="3" data-b="4">hello</div>

Allow tags start with x-

1var source = "<x><x-1>he<x-2 checked></x-2>wwww</x-1><a>";
2var html = xss(source, {
3  onIgnoreTag: function (tag, html, options) {
4    if (tag.substr(0, 2) === "x-") {
5      // do not filter its attributes
6      return html;
7    }
8  },
9});
10
11console.log("%s\nconvert to:\n%s", source, html);

Result:

1<x
2  ><x-1>he<x-2 checked></x-2>wwww</x-1
3  ><a>
4    convert to: &lt;x&gt;<x-1>he<x-2 checked></x-2>wwww</x-1><a></a></a
5></x>

Parse images in HTML

1var source =
2  '<img src="img1">a<img src="img2">b<img src="img3">c<img src="img4">d';
3var list = [];
4var html = xss(source, {
5  onTagAttr: function (tag, name, value, isWhiteAttr) {
6    if (tag === "img" && name === "src") {
7      // Use the built-in friendlyAttrValue function to escape attribute
8      // values. It supports converting entity tags such as &lt; to printable
9      // characters such as <
10      list.push(xss.friendlyAttrValue(value));
11    }
12    // Return nothing, means keep the default handling measure
13  },
14});
15
16console.log("image list:\n%s", list.join(", "));

Result:

1image list: img1, img2, img3, img4

Filter out HTML tags (keeps only plain text)

1var source = "<strong>hello</strong><script>alert(/xss/);</script>end";
2var html = xss(source, {
3  whiteList: {}, // empty, means filter out all tags
4  stripIgnoreTag: true, // filter out all HTML not in the whitelist
5  stripIgnoreTagBody: ["script"], // the script tag is a special case, we need
6  // to filter out its content
7});
8
9console.log("text: %s", html);

Result:

1text: helloend

License

1Copyright (c) 2012-2018 Zongmin Lei(雷宗民) <leizongmin@gmail.com>
2http://ucdok.com
3
4The MIT License
5
6Permission is hereby granted, free of charge, to any person obtaining
7a copy of this software and associated documentation files (the
8"Software"), to deal in the Software without restriction, including
9without limitation the rights to use, copy, modify, merge, publish,
10distribute, sublicense, and/or sell copies of the Software, and to
11permit persons to whom the Software is furnished to do so, subject to
12the following conditions:
13
14The above copyright notice and this permission notice shall be
15included in all copies or substantial portions of the Software.
16
17THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
18EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
19MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
20NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
21LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
22OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
23WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.