mirror of
https://gerrit.wikimedia.org/r/mediawiki/extensions/MultimediaViewer
synced 2024-11-25 00:25:47 +00:00
d85e7bf32e
This takes care of several minor annoyances: * centralizes all the text processing functions which have been floating all around the code, and adds proper tests * filters out invisible elements (sometimes used for metadata) * avoids merging separate words on HTML->text transformation * adds caching since doing all this transformations could be processing-intensive for big chunks of HTML. (This might or might not be a good idea. I haven't done performance tests, so this might be premature optimization, and increases memory use. OTOH these functions are often called in situations where an immediate UI response is expected (such as selecting a size from the list) so even small delays would be perceivable. Bug: 63126 Change-Id: I1ef1e3a33efdfea17612df00da6b629bf39e07aa Mingle: https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/388 Mingle: https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/369
201 lines
6.4 KiB
JavaScript
201 lines
6.4 KiB
JavaScript
/*
|
|
* This file is part of the MediaWiki extension MediaViewer.
|
|
*
|
|
* MediaViewer 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 2 of the License, or
|
|
* (at your option) any later version.
|
|
*
|
|
* MediaViewer 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 MediaViewer. If not, see <http://www.gnu.org/licenses/>.
|
|
*/
|
|
|
|
( function( mw, $ ) {
|
|
var HUP, cache;
|
|
|
|
/**
|
|
* @member mw.mmv.HtmlUtils
|
|
* @private
|
|
* @static
|
|
* Shared cache between HtmlUtils instances to store the results of expensive text operations.
|
|
* @type {{text: Object.<string, string>, textWithLinks: Object.<string, string>}}
|
|
*/
|
|
cache = {
|
|
text: {},
|
|
textWithLinks: {}
|
|
};
|
|
|
|
/**
|
|
* @class mw.mmv.HtmlUtils
|
|
* Helper class that does various HTML-to-text transformations
|
|
* @constructor
|
|
*/
|
|
function HtmlUtils() {}
|
|
HUP = HtmlUtils.prototype;
|
|
|
|
/**
|
|
* Returns a jQuery node which contains the given HTML (wrapped into a `<div>` - this is
|
|
* necessary since an arbitrary HTML string might not have a jQuery representation).
|
|
* @param {string|HTMLElement|jQuery} html
|
|
* @return {jQuery}
|
|
*/
|
|
HUP.wrapAndJquerify = function ( html ) {
|
|
if ( html instanceof jQuery || html instanceof HTMLElement ) {
|
|
return $( '<div>' ).append( $( html ).clone() );
|
|
} else if ( typeof html === 'string' ) {
|
|
return $( '<div>' + html + '</div>' );
|
|
} else {
|
|
mw.log.warn( 'wrapAndJquerify: unknown type', html );
|
|
throw 'wrapAndJquerify: unknown type';
|
|
}
|
|
};
|
|
|
|
/**
|
|
* Filters display:none children of a node.
|
|
* The root element is never filtered, and generally ignored (i.e. whether the root element is
|
|
* visible won't affect the filtering).
|
|
* Works in place.
|
|
* @param {jQuery} $jq
|
|
*/
|
|
HUP.filterInvisible = function ( $jq ) {
|
|
// We are not using :visible because
|
|
// 1) it would require appending $jq to the document which makes things complicated;
|
|
// 2) the main difference is that it looks for CSS rules hiding the element;
|
|
// since this function is intended to be used on html originating from a different
|
|
// document, possibly a different site, that would probably have unexpected results.
|
|
$jq
|
|
.find( '[style]' )
|
|
.filter( function() { return this.style.display === 'none'; } )
|
|
.remove();
|
|
};
|
|
|
|
/**
|
|
* Discards all nodes which do not match the whitelist,
|
|
* but keeps the text and whitelisted nodes inside them.
|
|
* Works in-place.
|
|
* @param {jQuery} $el
|
|
* @param {string} whitelist a jQuery selector string such as 'a, span, br'
|
|
*/
|
|
HUP.whitelistHtml = function ( $el, whitelist ) {
|
|
var child, $prev, $child = $el.children().first();
|
|
|
|
while ( $child && $child.length ) {
|
|
child = $child.get( 0 );
|
|
|
|
if ( child.nodeType !== child.ELEMENT_NODE ) {
|
|
return;
|
|
}
|
|
|
|
this.whitelistHtml( $child, whitelist );
|
|
|
|
if ( !$child.is( whitelist ) ) {
|
|
$prev = $child.prev();
|
|
$child.replaceWith( $child.contents() );
|
|
} else {
|
|
$prev = $child;
|
|
}
|
|
|
|
if ( $prev && $prev.length === 1 ) {
|
|
$child = $prev.next();
|
|
} else {
|
|
$child = $el.children().first();
|
|
}
|
|
}
|
|
};
|
|
|
|
/**
|
|
* Adds a whitespace to block elements. This is useful if you want to convert the contents
|
|
* to text and don't want words that are visually separate (e.g. table cells) to be fused.
|
|
* Works in-place.
|
|
* @param {jQuery} $el
|
|
*/
|
|
HUP.appendWhitespaceToBlockElements = function ( $el ) {
|
|
// the list of what elements to add whitespace to is somewhat ad-hoc (not all of these
|
|
// are technically block-level elements, and a lot of block-level elements are missing)
|
|
// but will hopefully cover the common cases where text is fused together.
|
|
$el
|
|
.find( 'blockquote, dd, dl, dt, li, td' )
|
|
.before( ' ' )
|
|
.after( ' ' );
|
|
$el
|
|
.find( 'br, tr, p' )
|
|
.before( '\n' )
|
|
.after( '\n' );
|
|
};
|
|
|
|
/**
|
|
* Returns the HTML code for a jQuery element (only the first one if passed a set of elements).
|
|
* Unlike .html(), this includes HTML code for the outermost element; compare
|
|
* - `$('<div>').html() // ''`
|
|
* - `mw.mmv.HtmlUtils.jqueryToHtml( $('<div>') ) // '<div></div>'`
|
|
* @param {jQuery} $el
|
|
* @return {string}
|
|
*/
|
|
HUP.jqueryToHtml = function ( $el ) {
|
|
// There are two possible implementations for this:
|
|
// 1) load innto a wrapper element and get its innerHTML;
|
|
// 2) use outerHTML.
|
|
// We go with 2) because it handles the case when a jQuery object contains something
|
|
// that is not an element (this can happen with e.g. $x.children() which returns text
|
|
// nodes as well).
|
|
return $( '<div>' ).append( $el ).html();
|
|
};
|
|
|
|
/**
|
|
* @protected
|
|
* Cleans up superfluous whitespace.
|
|
* Given that the results will be displayed in a HTML environment, this doesn't have any real
|
|
* effect. It is mostly there to make testing easier.
|
|
* @param {string} html a HTML (or plaintext) string
|
|
* @return {string}
|
|
*/
|
|
HUP.mergeWhitespace = function ( html ) {
|
|
html = html.replace( /^\s+|\s+$/g, '' );
|
|
html = html.replace( /\s*\n\s*/g, '\n' );
|
|
html = html.replace( / {2,}/g, ' ' );
|
|
return html;
|
|
};
|
|
|
|
/**
|
|
* Returns the text content of a html string.
|
|
* Tries to give an approximation of what would be visible if the HTML would be displayed.
|
|
* @param {string} html
|
|
* @return {string}
|
|
*/
|
|
HUP.htmlToText = function ( html ) {
|
|
var $html;
|
|
if ( !cache.text[html] ) {
|
|
$html = this.wrapAndJquerify( html );
|
|
this.filterInvisible( $html );
|
|
this.appendWhitespaceToBlockElements( $html );
|
|
cache.text[html] = this.mergeWhitespace( $html.text() );
|
|
}
|
|
return cache.text[html];
|
|
};
|
|
|
|
/**
|
|
* Returns the text content of a html string, with the `<a>` tags left intact.
|
|
* Tries to give an approximation of what would be visible if the HTML would be displayed.
|
|
* @param {string} html
|
|
* @return {string}
|
|
*/
|
|
HUP.htmlToTextWithLinks = function ( html ) {
|
|
var $html;
|
|
if ( !cache.textWithLinks[html] ) {
|
|
$html = this.wrapAndJquerify( html );
|
|
this.filterInvisible( $html );
|
|
this.appendWhitespaceToBlockElements( $html );
|
|
this.whitelistHtml( $html, 'a' );
|
|
cache.textWithLinks[html] = this.mergeWhitespace( $html.html() );
|
|
}
|
|
return cache.textWithLinks[html];
|
|
};
|
|
|
|
mw.mmv.HtmlUtils = HtmlUtils;
|
|
}( mediaWiki, jQuery ) );
|