File : core/htmlSanitizer/HTMLSanitizer.js

/*
Copyright - 2017 2023 - wwwouaiebe - Contact: https://www.ouaie.be/

This  program is free software;
you can redistribute it and/or modify it under the terms of the
GNU General Public License as published by the Free Software Foundation;
either version 3 of the License, or any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
*/

/*
Changes:
    - v2.0.0:
        - created
    - v3.0.0:
        - Issue ♯175 : Private and static fields and methods are coming
    - v3.1.0:
        - Issue ♯2 : Set all properties as private and use accessors.
Doc reviewed 20210914
Tests ...
*/

import HTMLSanitizerData from './HTMLSanitizerData.js';
import HtmlStringValidationResult from './HtmlStringValidationResult.js';
import UrlValidationResult from './UrlValidationResult.js';
import { SVG_NS, ZERO, NOT_FOUND } from '../../main/Constants.js';

/* ------------------------------------------------------------------------------------------------------------------------- */
/**
This class contains methods to sanitize url and string, filtering html tags and attributes
present in the string.

See theHTMLSanitizer for the one and only one instance of this class
*/
/* ------------------------------------------------------------------------------------------------------------------------- */

class HTMLSanitizer {

    /**
    the results of the #stringify method
    @type {String}
    */

    #stringifiedHTML = '';

    /**
    the errors detected by the #stringify method
    @type {String}
    */

    #stringifyErrors = '';

    /**
    An instance of the HTMLSanitizerData
    @type {HTMLSanitizerData}
    */

    static #htmlSanitizerData = new HTMLSanitizerData ( );

    /**
    Replace < >' " and nbsp chars with htmlEntities
    @param {String} htmlString the string to transform
    @return {String} a string with htmlEntities
    */

    #addHtmlEntities ( htmlString ) {
        const newHtmlString = htmlString
            .replaceAll ( /\u003c/g, '<' )
            .replaceAll ( /\u003e/g, '>' )
            .replaceAll ( /\u0022/g, '"' )
            .replaceAll ( /\u0027/g, ''' )
            .replaceAll ( /\u0a00/g, ' ' );

        return newHtmlString;
    }

    /**
    Helper method for the #stringify method. Validate an url present in a htmlString
    @param {String} url The url to validate
    @param {String} attributeName The attribute name in witch the url was found
    */

    #stringifyUrl ( url, attributeName ) {
        const validUrl = this.sanitizeToUrl ( url, attributeName ).url;
        if ( '' === validUrl && '' !== url ) {
            this.#stringifyErrors +=
                '\nAn invalid url (' +     url + ') was removed from a ' + attributeName + ' attribute';
        }
        else {
            this.#stringifiedHTML += ' ' + attributeName + '="' + validUrl + '"';
        }
    }

    /**
    Helper method for the #stringify method.  Validate and stringify the attributes of a svg node
    @param {SVGElement} currentNode The svg node for witch the attributes are stringified.
    @param {String} nodeName the name of the currentNode
    */

    #stringifySvgAttributes ( currentNode, nodeName ) {
        HTMLSanitizer.#htmlSanitizerData.getValidAttributesNames ( nodeName ).forEach (
            validAttributeName => {
                if ( currentNode.hasAttributeNS ( null, validAttributeName ) ) {
                    this.#stringifiedHTML += ' ' + validAttributeName + '="' +
                        this.#addHtmlEntities ( currentNode.getAttributeNS ( null, validAttributeName ) ) +
                        '"';
                    currentNode.removeAttributeNS ( null, validAttributeName );
                }
            }
        );
    }

    /**
    Helper method for the #stringify method.  Validate and stringify the attributes of a HTML node
    @param {HTMLElement} currentNode The HTML node for witch the attributes are stringified.
    @param {String} nodeName the name of the currentNode
    */

    #stringifyHTMLAttributes ( currentNode, nodeName ) {
        if ( currentNode.hasAttribute ( 'target' ) ) {
            this.#stringifiedHTML += ' rel="noopener noreferrer"';
        }
        HTMLSanitizer.#htmlSanitizerData.getValidAttributesNames ( nodeName ).forEach (
            validAttributeName => {
                if ( currentNode.hasAttribute ( validAttributeName ) ) {
                    if ( 'href' === validAttributeName || 'src' === validAttributeName ) {
                        this.#stringifyUrl ( currentNode.getAttribute ( validAttributeName ), validAttributeName );
                    }
                    else {
                        this.#stringifiedHTML += ' ' + validAttributeName + '="' +
                        this.#addHtmlEntities ( currentNode.getAttribute ( validAttributeName ) ) +
                        '"';
                    }
                    currentNode.removeAttribute ( validAttributeName );
                }
            }
        );
    }

    /**
    Helper method for the #stringify method.  Add the removed attributes to the error string
    @param {HTMLElement} currentNode The HTML node for witch the attributes are stringified.
    */

    #addStringifyErrors ( currentNode ) {
        for ( let attCounter = ZERO; attCounter < currentNode.attributes.length; attCounter ++ ) {
            if ( 'rel' !== currentNode.attributes [ attCounter ].name ) {
                this.#stringifyErrors +=
                    '\nAn unsecure attribute ' +
                    currentNode.attributes [ attCounter ].name +
                    '="' +
                    currentNode.attributes [ attCounter ].value +
                    '" was removed.';
            }
        }
    }

    /**
    Transform a node and it's descendants into a string, removing all the invalid tags, invalid atrributes,
    invalid texts and invalid url's
    @param {HTMLElement} sourceNode The node to stringify
    */

    #stringify ( sourceNode ) {
        const childs = sourceNode.childNodes;
        for ( let nodeCounter = 0; nodeCounter < childs.length; nodeCounter ++ ) {
            const currentNode = sourceNode.childNodes [ nodeCounter ];
            const nodeName = HTMLSanitizer.#htmlSanitizerData.getValidNodeName ( currentNode.nodeName );
            if ( '' === nodeName ) {
                this.#stringifyErrors += '\nAn invalid tag ' + currentNode.nodeName + ' was removed';
            }
            else if ( '\u0023text' === nodeName ) {
                this.#stringifiedHTML += this.#addHtmlEntities ( currentNode.nodeValue );
            }
            else {
                this.#stringifiedHTML += '<' + nodeName;
                if ( 'svg' === nodeName || 'text' === nodeName || 'polyline' === nodeName ) {
                    this.#stringifySvgAttributes ( currentNode, nodeName );
                }
                else {
                    this.#stringifyHTMLAttributes ( currentNode, nodeName );
                }
                this.#stringifiedHTML += '>';
                this.#stringify ( currentNode );
                this.#stringifiedHTML += '</' + nodeName + '>';
                if ( currentNode.attributes ) {
                    this.#addStringifyErrors ( currentNode );
                }
            }
        }
    }

    /**
    Helper function for the #cloneNode method. Clone a svg node
    @param {SVGElement} currentNode The svg node to clone
    @param {String} nodeName The name of the currentNode
    */

    #cloneSvg ( currentNode, nodeName ) {
        const newChildNode = document.createElementNS ( SVG_NS, nodeName );
        HTMLSanitizer.#htmlSanitizerData.getValidAttributesNames ( nodeName ).forEach (
            validAttributeName => {
                if ( currentNode.hasAttributeNS ( null, validAttributeName ) ) {
                    newChildNode.setAttributeNS (
                        null,
                        validAttributeName,
                        currentNode.getAttributeNS ( null, validAttributeName )
                    );
                    currentNode.removeAttributeNS ( null, validAttributeName );
                }
            }
        );
        return newChildNode;
    }

    /**
    Helper function for the #cloneNode method. Clone a HTML node
    @param {HTMLElement} currentNode The html node to clone
    @param {String} nodeName The name of the currentNode
    */

    #cloneHTML ( currentNode, nodeName ) {
        const newChildNode = document.createElement ( nodeName );
        HTMLSanitizer.#htmlSanitizerData.getValidAttributesNames ( nodeName ).forEach (
            validAttributeName => {
                if ( currentNode.hasAttribute ( validAttributeName ) ) {
                    if ( 'href' === validAttributeName || 'src' === validAttributeName ) {
                        const attributeValue = this.sanitizeToUrl (
                            currentNode.getAttribute ( validAttributeName ),
                            validAttributeName
                        ).url;
                        if ( '' !== attributeValue ) {
                            newChildNode.setAttribute ( validAttributeName, attributeValue );
                        }
                    }
                    else {
                        newChildNode.setAttribute (
                            validAttributeName,
                            currentNode.getAttribute ( validAttributeName )
                        );
                    }
                }
            }
        );
        if ( currentNode.hasAttribute ( 'target' ) ) {
            newChildNode.setAttribute ( 'rel', 'noopener noreferrer' );
        }
        return newChildNode;
    }

    /**
    Deep clone the contains of an HTML node into another node. Only valid tags, valid attributes, valid url's
    and valid texts are cloned

    @param {HTMLElement} clonedNode The node to clone
    @param {HTMLElement} newNode The destination node
    */

    #cloneNode ( clonedNode, newNode ) {
        const childs = clonedNode.childNodes;
        for ( let nodeCounter = 0; nodeCounter < childs.length; nodeCounter ++ ) {
            const currentNode = clonedNode.childNodes [ nodeCounter ];
            const nodeName = HTMLSanitizer.#htmlSanitizerData.getValidNodeName ( currentNode.nodeName );
            if ( '\u0023text' === nodeName ) {
                newNode.appendChild ( document.createTextNode ( currentNode.nodeValue ) );
            }
            else if ( '' !== nodeName ) {
                const newChildNode =
                    'svg' === nodeName || 'text' === nodeName || 'polyline' === nodeName
                        ?
                        this.#cloneSvg ( currentNode, nodeName )
                        :
                        this.#cloneHTML ( currentNode, nodeName );

                newNode.appendChild ( newChildNode );
                this.#cloneNode ( currentNode, newChildNode );
            }
        }
    }

    /**
    The constructor
    */

    constructor ( ) {
        Object.freeze ( this );
    }

    /**
    This method transform a string containing html and svg tags into html and svg elements and copy these elements
    as child nodes of the targetNode. Only tags and attributes present in the HTMLSanitizerData.#validityMap variable
    are copied in the targetNode. Url in the href and src attributes must be valid url (see sanitizeToUrl method)
    @param {String} htmlString the string to transform
    @param {HTMLElement} targetNode the node in witch the created elements are placed
    */

    sanitizeToHtmlElement ( htmlString, targetNode ) {

        const parseResult = new DOMParser ( ).parseFromString ( '<div>' + htmlString + '</div>', 'text/html' );

        const docFragment = new DocumentFragment ( );
        if ( parseResult && '\u0023document' === parseResult.nodeName ) {
            this.#cloneNode ( parseResult.body.firstChild, docFragment );
            targetNode.appendChild ( docFragment );
        }
        else {
            targetNode.textContent = '';
        }
    }

    /**
    This method clone a DOM node, removing all invalid childs and attributes
    @param {HTMLElement} htmlElement The node to clone
    @return {HTMLElement} The cloned node
    */

    clone ( htmlElement ) {
        const clone = document.createElement ( htmlElement.tagName );
        this.#cloneNode ( htmlElement, clone );

        return clone;
    }

    /**
    This method transform a string containing html and svg tags. Tags and attributes not present in the
    HTMLSanitizerData.#validityMap variable are removed. Invalid Url in the href and src attributes are
    also removed (see sanitizeToUrl method)
    @param {String} htmlString the string to transform
    @return {HtmlStringValidationResult} a HtmlStringValidationResult with the result of the validation
    */

    sanitizeToHtmlString ( htmlString ) {

        // ! don't use XMLSerializer. Problems with ", &apos and   and xmlns

        this.#stringifiedHTML = '';
        this.#stringifyErrors = '';

        const parseResult =
            new DOMParser ( ).parseFromString ( '<div>' + htmlString.replace ( ' ', '\u0a00' ) + '</div>', 'text/html' );
        if ( parseResult && '\u0023document' === parseResult.nodeName ) {
            this.#stringify ( parseResult.body.firstChild );
            return new HtmlStringValidationResult ( this.#stringifiedHTML, this.#stringifyErrors );
        }
        return new HtmlStringValidationResult ( '', 'Parsing error' );
    }

    /**
    This method verify that a string contains a valid url.

    A valid url must not contains html tags or html entities or invalid characters
    and must start with a valid protocol.

    Valid protocols are http: and https:. For href attributes mailto:, sms: and tel: are also valid
    and for src attributes, data: is also valid.

    sms: and tel: url's  must start with a + and contains only digits, *, # or space
    @param {String} urlString The url to validate
    @param {String} attributeName The attribute name in witch the url will be placed. must be 'src' or
    null (in this case 'href' is used as default)
    @return {UrlValidationResult} a UrlValidationResult with the result of the validation
    */

    sanitizeToUrl ( urlString, attributeName ) {

        const tmpAttributeName = attributeName || 'href';

        // set the url inside a div and then parsing...
        const parseResult = new DOMParser ( ).parseFromString ( '<div>' + urlString + '</div>', 'text/html' );
        if ( ! parseResult || '\u0023document' !== parseResult.nodeName ) {

            // strange: no result or not a document. We return an empty string
            return new UrlValidationResult ( '', 'Parsing error' );
        }

        // Taking the first child node of the pasing and concatenate the childnodes of this node...
        const resultNode = parseResult.body.firstChild;
        let newUrlString = '';
        for ( let nodeCounter = 0; nodeCounter < resultNode.childNodes.length; nodeCounter ++ ) {
            if ( '\u0023text' === resultNode.childNodes [ nodeCounter ].nodeName ) {

                // ...  if only text nodes are found
                newUrlString += resultNode.childNodes [ nodeCounter ].nodeValue;
            }
            else {

                // otherwise returning an empty string
                return new UrlValidationResult ( '', 'Invalid characters found in the url' );
            }
        }

        // removing < > " ' characters in a copy url ...
        newUrlString = newUrlString
            .replaceAll ( /</g, '' )
            .replaceAll ( />/g, '' )
            .replaceAll ( /"/g, '' )
            .replaceAll ( /\u0027/g, '' )
            .replaceAll ( /</g, '' )
            .replaceAll ( />/g, '' )
            .replaceAll ( /"/g, '' )
            .replaceAll ( /'/g, '' )
            .replaceAll ( /%3C/g, '' )
            .replaceAll ( /%3c/g, '' )
            .replaceAll ( /%3E/g, '' )
            .replaceAll ( /%3e/g, '' )
            .replaceAll ( /%22/g, '' )
            .replaceAll ( /%27/g, '' );

        // and comparing the result with the url
        if ( newUrlString !== urlString ) {

            // < > " ' characters found i the url. Returning an empty string
            return new UrlValidationResult ( '', 'Invalid characters found in the url' );
        }

        // creating a list of valid protocols for the url
        const validProtocols = [ 'https:' ];
        if ( 'http:' === window.location.protocol || 'href' === tmpAttributeName ) {
            validProtocols.push ( 'http:' );
        }
        if ( 'href' === tmpAttributeName ) {
            validProtocols.push ( 'mailto:' );
            validProtocols.push ( 'sms:' );
            validProtocols.push ( 'tel:' );

            // the url contains only letters and numbers chars and start with a hash. It's a link to the document itself
            const urlHash = newUrlString.match ( /^\u0023\w*/ );
            if ( urlHash && newUrlString === urlHash [ ZERO ] ) {
                return new UrlValidationResult ( newUrlString, '' );
            }
        }
        if ( 'src' === tmpAttributeName ) {
            validProtocols.push ( 'data:' );
        }

        // We try to create a url object from the url string
        let url = null;
        try {
            url = new URL ( newUrlString );
        }
        catch ( err ) {

            // not possible to create an url. Returning an empty string
            return new UrlValidationResult ( '', 'Invalid url string' );
        }
        if ( NOT_FOUND === validProtocols.indexOf ( url.protocol ) ) {

            // the url protocol is not in the list of valid protocol. Returning an empty string
            return new UrlValidationResult ( '', 'Invalid protocol ' + url.protocol );
        }
        if ( NOT_FOUND !== [ 'sms:', 'tel:' ].indexOf ( url.protocol ) ) {

            // sms and tel url must start with a + and contains only numbers, hash or star
            if ( url.pathname.match ( /^\+[0-9,*,\u0023]*$/ ) ) {
                return new UrlValidationResult ( newUrlString, '' );
            }

            return new UrlValidationResult ( '', 'Invalid sms: or tel: url' );

        }

        // try the encodeURIComponent function on the href part of the url
        try {
            encodeURIComponent ( url.href );
        }
        catch ( err ) {
            return new UrlValidationResult ( '', 'Invalid character in url' );
        }
        return new UrlValidationResult ( newUrlString, '' );
    }

    /**
    Remove all html tags from a string and replace htmlEntities and < > ' " and nbsp chars with others similar unicode chars
    @param {String} stringToSanitize the string to transform
    @return {String} a string with html tags removed and htmlEntities and < >' " and nbsp chars replaced
    */

    sanitizeToJsString ( stringToSanitize ) {

        // Parsing the string inside a div...
        const parseResult = new DOMParser ( ).parseFromString ( '<div>' + stringToSanitize + '</div>', 'text/html' );
        if ( ! parseResult || '\u0023document' !== parseResult.nodeName ) {

            // Bad results from the parsing... Returning an empty string
            return '';
        }
        const resultNode = parseResult.body.firstChild;
        let sanitizedString = '';
        for ( let nodeCounter = 0; nodeCounter < resultNode.childNodes.length; nodeCounter ++ ) {
            if ( '\u0023text' === resultNode.childNodes [ nodeCounter ].nodeName ) {
                sanitizedString += resultNode.childNodes [ nodeCounter ].nodeValue;
            }
            else {

                // The parsing contains others nodes than text string... returning an empty string
                return '';
            }
        }

        // replacing <>'" with others similar chars
        sanitizedString = sanitizedString
            .replaceAll ( /</g, '\u227a' )
            .replaceAll ( />/g, '\u227b' )
            .replaceAll ( /"/g, '\u2033' )
            .replaceAll ( /\u0027/g, '\u2032' );

        return sanitizedString;
    }

    /**
    This method verify that a string describe a css color. A valid css color must start with a hash followed by 6 hex numbers
    @param {String} colorString the string to test
    @return {String} the verified color or null if the given color is invalid
    */

    sanitizeToColor ( colorString ) {
        const newColor = colorString.match ( /^\u0023[0-9,A-F,a-f]{6}$/ );
        if ( newColor ) {
            return newColor [ ZERO ];
        }
        return null;
    }
}

/* ------------------------------------------------------------------------------------------------------------------------- */
/**
The one and only one instance of HTMLSanitizer class
@type {HTMLSanitizer}
*/
/* ------------------------------------------------------------------------------------------------------------------------- */

const theHTMLSanitizer = new HTMLSanitizer ( );

export default theHTMLSanitizer;

/* --- End of file --------------------------------------------------------------------------------------------------------- */


1	/*
2	Copyright - 2017 2023 - wwwouaiebe - Contact: https://www.ouaie.be/
3
4	This program is free software;
5	you can redistribute it and/or modify it under the terms of the
6	GNU General Public License as published by the Free Software Foundation;
7	either version 3 of the License, or any later version.
8
9	This program is distributed in the hope that it will be useful,
10	but WITHOUT ANY WARRANTY; without even the implied warranty of
11	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12	GNU General Public License for more details.
13
14	You should have received a copy of the GNU General Public License
15	along with this program; if not, write to the Free Software
16	Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
17	*/
18
19	/*
20	Changes:
21	- v2.0.0:
22	- created
23	- v3.0.0:
24	- Issue ♯175 : Private and static fields and methods are coming
25	- v3.1.0:
26	- Issue ♯2 : Set all properties as private and use accessors.
27	Doc reviewed 20210914
28	Tests ...
29	*/
30
31	import HTMLSanitizerData from './HTMLSanitizerData.js';
32	import HtmlStringValidationResult from './HtmlStringValidationResult.js';
33	import UrlValidationResult from './UrlValidationResult.js';
34	import { SVG_NS, ZERO, NOT_FOUND } from '../../main/Constants.js';
35
36	/* ------------------------------------------------------------------------------------------------------------------------- */
37	/**
38	This class contains methods to sanitize url and string, filtering html tags and attributes
39	present in the string.
40
41	See theHTMLSanitizer for the one and only one instance of this class
42	*/
43	/* ------------------------------------------------------------------------------------------------------------------------- */
44
45	class HTMLSanitizer {
46
47	/**
48	the results of the #stringify method
49	@type {String}
50	*/
51
52	#stringifiedHTML = '';
53
54	/**
55	the errors detected by the #stringify method
56	@type {String}
57	*/
58
59	#stringifyErrors = '';
60
61	/**
62	An instance of the HTMLSanitizerData
63	@type {HTMLSanitizerData}
64	*/
65
66	static #htmlSanitizerData = new HTMLSanitizerData ( );
67
68	/**
69	Replace < >' " and nbsp chars with htmlEntities
70	@param {String} htmlString the string to transform
71	@return {String} a string with htmlEntities
72	*/
73
74	#addHtmlEntities ( htmlString ) {
75	const newHtmlString = htmlString
76	.replaceAll ( /\u003c/g, '<' )
77	.replaceAll ( /\u003e/g, '>' )
78	.replaceAll ( /\u0022/g, '"' )
79	.replaceAll ( /\u0027/g, ''' )
80	.replaceAll ( /\u0a00/g, ' ' );
81
82	return newHtmlString;
83	}
84
85	/**
86	Helper method for the #stringify method. Validate an url present in a htmlString
87	@param {String} url The url to validate
88	@param {String} attributeName The attribute name in witch the url was found
89	*/
90
91	#stringifyUrl ( url, attributeName ) {
92	const validUrl = this.sanitizeToUrl ( url, attributeName ).url;
93	if ( '' === validUrl && '' !== url ) {
94	this.#stringifyErrors +=
95	'\nAn invalid url (' + url + ') was removed from a ' + attributeName + ' attribute';
96	}
97	else {
98	this.#stringifiedHTML += ' ' + attributeName + '="' + validUrl + '"';
99	}
100	}
101
102	/**
103	Helper method for the #stringify method. Validate and stringify the attributes of a svg node
104	@param {SVGElement} currentNode The svg node for witch the attributes are stringified.
105	@param {String} nodeName the name of the currentNode
106	*/
107
108	#stringifySvgAttributes ( currentNode, nodeName ) {
109	HTMLSanitizer.#htmlSanitizerData.getValidAttributesNames ( nodeName ).forEach (
110	validAttributeName => {
111	if ( currentNode.hasAttributeNS ( null, validAttributeName ) ) {
112	this.#stringifiedHTML += ' ' + validAttributeName + '="' +
113	this.#addHtmlEntities ( currentNode.getAttributeNS ( null, validAttributeName ) ) +
114	'"';
115	currentNode.removeAttributeNS ( null, validAttributeName );
116	}
117	}
118	);
119	}
120
121	/**
122	Helper method for the #stringify method. Validate and stringify the attributes of a HTML node
123	@param {HTMLElement} currentNode The HTML node for witch the attributes are stringified.
124	@param {String} nodeName the name of the currentNode
125	*/
126
127	#stringifyHTMLAttributes ( currentNode, nodeName ) {
128	if ( currentNode.hasAttribute ( 'target' ) ) {
129	this.#stringifiedHTML += ' rel="noopener noreferrer"';
130	}
131	HTMLSanitizer.#htmlSanitizerData.getValidAttributesNames ( nodeName ).forEach (
132	validAttributeName => {
133	if ( currentNode.hasAttribute ( validAttributeName ) ) {
134	if ( 'href' === validAttributeName \|\| 'src' === validAttributeName ) {
135	this.#stringifyUrl ( currentNode.getAttribute ( validAttributeName ), validAttributeName );
136	}
137	else {
138	this.#stringifiedHTML += ' ' + validAttributeName + '="' +
139	this.#addHtmlEntities ( currentNode.getAttribute ( validAttributeName ) ) +
140	'"';
141	}
142	currentNode.removeAttribute ( validAttributeName );
143	}
144	}
145	);
146	}
147
148	/**
149	Helper method for the #stringify method. Add the removed attributes to the error string
150	@param {HTMLElement} currentNode The HTML node for witch the attributes are stringified.
151	*/
152
153	#addStringifyErrors ( currentNode ) {
154	for ( let attCounter = ZERO; attCounter < currentNode.attributes.length; attCounter ++ ) {
155	if ( 'rel' !== currentNode.attributes [ attCounter ].name ) {
156	this.#stringifyErrors +=
157	'\nAn unsecure attribute ' +
158	currentNode.attributes [ attCounter ].name +
159	'="' +
160	currentNode.attributes [ attCounter ].value +
161	'" was removed.';
162	}
163	}
164	}
165
166	/**
167	Transform a node and it's descendants into a string, removing all the invalid tags, invalid atrributes,
168	invalid texts and invalid url's
169	@param {HTMLElement} sourceNode The node to stringify
170	*/
171
172	#stringify ( sourceNode ) {
173	const childs = sourceNode.childNodes;
174	for ( let nodeCounter = 0; nodeCounter < childs.length; nodeCounter ++ ) {
175	const currentNode = sourceNode.childNodes [ nodeCounter ];
176	const nodeName = HTMLSanitizer.#htmlSanitizerData.getValidNodeName ( currentNode.nodeName );
177	if ( '' === nodeName ) {
178	this.#stringifyErrors += '\nAn invalid tag ' + currentNode.nodeName + ' was removed';
179	}
180	else if ( '\u0023text' === nodeName ) {
181	this.#stringifiedHTML += this.#addHtmlEntities ( currentNode.nodeValue );
182	}
183	else {
184	this.#stringifiedHTML += '<' + nodeName;
185	if ( 'svg' === nodeName \|\| 'text' === nodeName \|\| 'polyline' === nodeName ) {
186	this.#stringifySvgAttributes ( currentNode, nodeName );
187	}
188	else {
189	this.#stringifyHTMLAttributes ( currentNode, nodeName );
190	}
191	this.#stringifiedHTML += '>';
192	this.#stringify ( currentNode );
193	this.#stringifiedHTML += '</' + nodeName + '>';
194	if ( currentNode.attributes ) {
195	this.#addStringifyErrors ( currentNode );
196	}
197	}
198	}
199	}
200
201	/**
202	Helper function for the #cloneNode method. Clone a svg node
203	@param {SVGElement} currentNode The svg node to clone
204	@param {String} nodeName The name of the currentNode
205	*/
206
207	#cloneSvg ( currentNode, nodeName ) {
208	const newChildNode = document.createElementNS ( SVG_NS, nodeName );
209	HTMLSanitizer.#htmlSanitizerData.getValidAttributesNames ( nodeName ).forEach (
210	validAttributeName => {
211	if ( currentNode.hasAttributeNS ( null, validAttributeName ) ) {
212	newChildNode.setAttributeNS (
213	null,
214	validAttributeName,
215	currentNode.getAttributeNS ( null, validAttributeName )
216	);
217	currentNode.removeAttributeNS ( null, validAttributeName );
218	}
219	}
220	);
221	return newChildNode;
222	}
223
224	/**
225	Helper function for the #cloneNode method. Clone a HTML node
226	@param {HTMLElement} currentNode The html node to clone
227	@param {String} nodeName The name of the currentNode
228	*/
229
230	#cloneHTML ( currentNode, nodeName ) {
231	const newChildNode = document.createElement ( nodeName );
232	HTMLSanitizer.#htmlSanitizerData.getValidAttributesNames ( nodeName ).forEach (
233	validAttributeName => {
234	if ( currentNode.hasAttribute ( validAttributeName ) ) {
235	if ( 'href' === validAttributeName \|\| 'src' === validAttributeName ) {
236	const attributeValue = this.sanitizeToUrl (
237	currentNode.getAttribute ( validAttributeName ),
238	validAttributeName
239	).url;
240	if ( '' !== attributeValue ) {
241	newChildNode.setAttribute ( validAttributeName, attributeValue );
242	}
243	}
244	else {
245	newChildNode.setAttribute (
246	validAttributeName,
247	currentNode.getAttribute ( validAttributeName )
248	);
249	}
250	}
251	}
252	);
253	if ( currentNode.hasAttribute ( 'target' ) ) {
254	newChildNode.setAttribute ( 'rel', 'noopener noreferrer' );
255	}
256	return newChildNode;
257	}
258
259	/**
260	Deep clone the contains of an HTML node into another node. Only valid tags, valid attributes, valid url's
261	and valid texts are cloned
262
263	@param {HTMLElement} clonedNode The node to clone
264	@param {HTMLElement} newNode The destination node
265	*/
266
267	#cloneNode ( clonedNode, newNode ) {
268	const childs = clonedNode.childNodes;
269	for ( let nodeCounter = 0; nodeCounter < childs.length; nodeCounter ++ ) {
270	const currentNode = clonedNode.childNodes [ nodeCounter ];
271	const nodeName = HTMLSanitizer.#htmlSanitizerData.getValidNodeName ( currentNode.nodeName );
272	if ( '\u0023text' === nodeName ) {
273	newNode.appendChild ( document.createTextNode ( currentNode.nodeValue ) );
274	}
275	else if ( '' !== nodeName ) {
276	const newChildNode =
277	'svg' === nodeName \|\| 'text' === nodeName \|\| 'polyline' === nodeName
278	?
279	this.#cloneSvg ( currentNode, nodeName )
280	:
281	this.#cloneHTML ( currentNode, nodeName );
282
283	newNode.appendChild ( newChildNode );
284	this.#cloneNode ( currentNode, newChildNode );
285	}
286	}
287	}
288
289	/**
290	The constructor
291	*/
292
293	constructor ( ) {
294	Object.freeze ( this );
295	}
296
297	/**
298	This method transform a string containing html and svg tags into html and svg elements and copy these elements
299	as child nodes of the targetNode. Only tags and attributes present in the HTMLSanitizerData.#validityMap variable
300	are copied in the targetNode. Url in the href and src attributes must be valid url (see sanitizeToUrl method)
301	@param {String} htmlString the string to transform
302	@param {HTMLElement} targetNode the node in witch the created elements are placed
303	*/
304
305	sanitizeToHtmlElement ( htmlString, targetNode ) {
306
307	const parseResult = new DOMParser ( ).parseFromString ( '<div>' + htmlString + '</div>', 'text/html' );
308
309	const docFragment = new DocumentFragment ( );
310	if ( parseResult && '\u0023document' === parseResult.nodeName ) {
311	this.#cloneNode ( parseResult.body.firstChild, docFragment );
312	targetNode.appendChild ( docFragment );
313	}
314	else {
315	targetNode.textContent = '';
316	}
317	}
318
319	/**
320	This method clone a DOM node, removing all invalid childs and attributes
321	@param {HTMLElement} htmlElement The node to clone
322	@return {HTMLElement} The cloned node
323	*/
324
325	clone ( htmlElement ) {
326	const clone = document.createElement ( htmlElement.tagName );
327	this.#cloneNode ( htmlElement, clone );
328
329	return clone;
330	}
331
332	/**
333	This method transform a string containing html and svg tags. Tags and attributes not present in the
334	HTMLSanitizerData.#validityMap variable are removed. Invalid Url in the href and src attributes are
335	also removed (see sanitizeToUrl method)
336	@param {String} htmlString the string to transform
337	@return {HtmlStringValidationResult} a HtmlStringValidationResult with the result of the validation
338	*/
339
340	sanitizeToHtmlString ( htmlString ) {
341
342	// ! don't use XMLSerializer. Problems with ", &apos and and xmlns
343
344	this.#stringifiedHTML = '';
345	this.#stringifyErrors = '';
346
347	const parseResult =
348	new DOMParser ( ).parseFromString ( '<div>' + htmlString.replace ( ' ', '\u0a00' ) + '</div>', 'text/html' );
349	if ( parseResult && '\u0023document' === parseResult.nodeName ) {
350	this.#stringify ( parseResult.body.firstChild );
351	return new HtmlStringValidationResult ( this.#stringifiedHTML, this.#stringifyErrors );
352	}
353	return new HtmlStringValidationResult ( '', 'Parsing error' );
354	}
355
356	/**
357	This method verify that a string contains a valid url.
358
359	A valid url must not contains html tags or html entities or invalid characters
360	and must start with a valid protocol.
361
362	Valid protocols are http: and https:. For href attributes mailto:, sms: and tel: are also valid
363	and for src attributes, data: is also valid.
364
365	sms: and tel: url's must start with a + and contains only digits, *, # or space
366	@param {String} urlString The url to validate
367	@param {String} attributeName The attribute name in witch the url will be placed. must be 'src' or
368	null (in this case 'href' is used as default)
369	@return {UrlValidationResult} a UrlValidationResult with the result of the validation
370	*/
371
372	sanitizeToUrl ( urlString, attributeName ) {
373
374	const tmpAttributeName = attributeName \|\| 'href';
375
376	// set the url inside a div and then parsing...
377	const parseResult = new DOMParser ( ).parseFromString ( '<div>' + urlString + '</div>', 'text/html' );
378	if ( ! parseResult \|\| '\u0023document' !== parseResult.nodeName ) {
379
380	// strange: no result or not a document. We return an empty string
381	return new UrlValidationResult ( '', 'Parsing error' );
382	}
383
384	// Taking the first child node of the pasing and concatenate the childnodes of this node...
385	const resultNode = parseResult.body.firstChild;
386	let newUrlString = '';
387	for ( let nodeCounter = 0; nodeCounter < resultNode.childNodes.length; nodeCounter ++ ) {
388	if ( '\u0023text' === resultNode.childNodes [ nodeCounter ].nodeName ) {
389
390	// ... if only text nodes are found
391	newUrlString += resultNode.childNodes [ nodeCounter ].nodeValue;
392	}
393	else {
394
395	// otherwise returning an empty string
396	return new UrlValidationResult ( '', 'Invalid characters found in the url' );
397	}
398	}
399
400	// removing < > " ' characters in a copy url ...
401	newUrlString = newUrlString
402	.replaceAll ( /</g, '' )
403	.replaceAll ( />/g, '' )
404	.replaceAll ( /"/g, '' )
405	.replaceAll ( /\u0027/g, '' )
406	.replaceAll ( /</g, '' )
407	.replaceAll ( />/g, '' )
408	.replaceAll ( /"/g, '' )
409	.replaceAll ( /'/g, '' )
410	.replaceAll ( /%3C/g, '' )
411	.replaceAll ( /%3c/g, '' )
412	.replaceAll ( /%3E/g, '' )
413	.replaceAll ( /%3e/g, '' )
414	.replaceAll ( /%22/g, '' )
415	.replaceAll ( /%27/g, '' );
416
417	// and comparing the result with the url
418	if ( newUrlString !== urlString ) {
419
420	// < > " ' characters found i the url. Returning an empty string
421	return new UrlValidationResult ( '', 'Invalid characters found in the url' );
422	}
423
424	// creating a list of valid protocols for the url
425	const validProtocols = [ 'https:' ];
426	if ( 'http:' === window.location.protocol \|\| 'href' === tmpAttributeName ) {
427	validProtocols.push ( 'http:' );
428	}
429	if ( 'href' === tmpAttributeName ) {
430	validProtocols.push ( 'mailto:' );
431	validProtocols.push ( 'sms:' );
432	validProtocols.push ( 'tel:' );
433
434	// the url contains only letters and numbers chars and start with a hash. It's a link to the document itself
435	const urlHash = newUrlString.match ( /^\u0023\w*/ );
436	if ( urlHash && newUrlString === urlHash [ ZERO ] ) {
437	return new UrlValidationResult ( newUrlString, '' );
438	}
439	}
440	if ( 'src' === tmpAttributeName ) {
441	validProtocols.push ( 'data:' );
442	}
443
444	// We try to create a url object from the url string
445	let url = null;
446	try {
447	url = new URL ( newUrlString );
448	}
449	catch ( err ) {
450
451	// not possible to create an url. Returning an empty string
452	return new UrlValidationResult ( '', 'Invalid url string' );
453	}
454	if ( NOT_FOUND === validProtocols.indexOf ( url.protocol ) ) {
455
456	// the url protocol is not in the list of valid protocol. Returning an empty string
457	return new UrlValidationResult ( '', 'Invalid protocol ' + url.protocol );
458	}
459	if ( NOT_FOUND !== [ 'sms:', 'tel:' ].indexOf ( url.protocol ) ) {
460
461	// sms and tel url must start with a + and contains only numbers, hash or star
462	if ( url.pathname.match ( /^\+[0-9,,\u0023]$/ ) ) {
463	return new UrlValidationResult ( newUrlString, '' );
464	}
465
466	return new UrlValidationResult ( '', 'Invalid sms: or tel: url' );
467
468	}
469
470	// try the encodeURIComponent function on the href part of the url
471	try {
472	encodeURIComponent ( url.href );
473	}
474	catch ( err ) {
475	return new UrlValidationResult ( '', 'Invalid character in url' );
476	}
477	return new UrlValidationResult ( newUrlString, '' );
478	}
479
480	/**
481	Remove all html tags from a string and replace htmlEntities and < > ' " and nbsp chars with others similar unicode chars
482	@param {String} stringToSanitize the string to transform
483	@return {String} a string with html tags removed and htmlEntities and < >' " and nbsp chars replaced
484	*/
485
486	sanitizeToJsString ( stringToSanitize ) {
487
488	// Parsing the string inside a div...
489	const parseResult = new DOMParser ( ).parseFromString ( '<div>' + stringToSanitize + '</div>', 'text/html' );
490	if ( ! parseResult \|\| '\u0023document' !== parseResult.nodeName ) {
491
492	// Bad results from the parsing... Returning an empty string
493	return '';
494	}
495	const resultNode = parseResult.body.firstChild;
496	let sanitizedString = '';
497	for ( let nodeCounter = 0; nodeCounter < resultNode.childNodes.length; nodeCounter ++ ) {
498	if ( '\u0023text' === resultNode.childNodes [ nodeCounter ].nodeName ) {
499	sanitizedString += resultNode.childNodes [ nodeCounter ].nodeValue;
500	}
501	else {
502
503	// The parsing contains others nodes than text string... returning an empty string
504	return '';
505	}
506	}
507
508	// replacing <>'" with others similar chars
509	sanitizedString = sanitizedString
510	.replaceAll ( /</g, '\u227a' )
511	.replaceAll ( />/g, '\u227b' )
512	.replaceAll ( /"/g, '\u2033' )
513	.replaceAll ( /\u0027/g, '\u2032' );
514
515	return sanitizedString;
516	}
517
518	/**
519	This method verify that a string describe a css color. A valid css color must start with a hash followed by 6 hex numbers
520	@param {String} colorString the string to test
521	@return {String} the verified color or null if the given color is invalid
522	*/
523
524	sanitizeToColor ( colorString ) {
525	const newColor = colorString.match ( /^\u0023[0-9,A-F,a-f]{6}$/ );
526	if ( newColor ) {
527	return newColor [ ZERO ];
528	}
529	return null;
530	}
531	}
532
533	/* ------------------------------------------------------------------------------------------------------------------------- */
534	/**
535	The one and only one instance of HTMLSanitizer class
536	@type {HTMLSanitizer}
537	*/
538	/* ------------------------------------------------------------------------------------------------------------------------- */
539
540	const theHTMLSanitizer = new HTMLSanitizer ( );
541
542	export default theHTMLSanitizer;
543
544	/* --- End of file --------------------------------------------------------------------------------------------------------- */
545