| 1 | /**
|
| 2 | * @license Copyright (c) 2003-2024, CKSource Holding sp. z o.o. All rights reserved.
|
| 3 | * For licensing, see LICENSE.md or https://ckeditor.com/legal/ckeditor-oss-license
|
| 4 | */
|
| 5 | /**
|
| 6 | * @module paste-from-office/filters/space
|
| 7 | */
|
| 8 | /**
|
| 9 | * Replaces last space preceding elements closing tag with ` `. Such operation prevents spaces from being removed
|
| 10 | * during further DOM/View processing (see especially {@link module:engine/view/domconverter~DomConverter#_processDomInlineNodes}).
|
| 11 | * This method also takes into account Word specific `<o:p></o:p>` empty tags.
|
| 12 | * Additionally multiline sequences of spaces and new lines between tags are removed (see #39 and #40).
|
| 13 | *
|
| 14 | * @param htmlString HTML string in which spacing should be normalized.
|
| 15 | * @returns Input HTML with spaces normalized.
|
| 16 | */
|
| 17 | export function normalizeSpacing(htmlString) {
|
| 18 | // Run normalizeSafariSpaceSpans() two times to cover nested spans.
|
| 19 | return normalizeSafariSpaceSpans(normalizeSafariSpaceSpans(htmlString))
|
| 20 | // Remove all \r\n from "spacerun spans" so the last replace line doesn't strip all whitespaces.
|
| 21 | .replace(/(<span\s+style=['"]mso-spacerun:yes['"]>[^\S\r\n]*?)[\r\n]+([^\S\r\n]*<\/span>)/g, '$1$2')
|
| 22 | .replace(/<span\s+style=['"]mso-spacerun:yes['"]><\/span>/g, '')
|
| 23 | .replace(/(<span\s+style=['"]letter-spacing:[^'"]+?['"]>)[\r\n]+(<\/span>)/g, '$1 $2')
|
| 24 | .replace(/ <\//g, '\u00A0</')
|
| 25 | .replace(/ <o:p><\/o:p>/g, '\u00A0<o:p></o:p>')
|
| 26 | // Remove <o:p> block filler from empty paragraph. Safari uses \u00A0 instead of .
|
| 27 | .replace(/<o:p>( |\u00A0)<\/o:p>/g, '')
|
| 28 | // Remove all whitespaces when they contain any \r or \n.
|
| 29 | .replace(/>([^\S\r\n]*[\r\n]\s*)</g, '><');
|
| 30 | }
|
| 31 | /**
|
| 32 | * Normalizes spacing in special Word `spacerun spans` (`<span style='mso-spacerun:yes'>\s+</span>`) by replacing
|
| 33 | * all spaces with ` ` pairs. This prevents spaces from being removed during further DOM/View processing
|
| 34 | * (see especially {@link module:engine/view/domconverter~DomConverter#_processDomInlineNodes}).
|
| 35 | *
|
| 36 | * @param htmlDocument Native `Document` object in which spacing should be normalized.
|
| 37 | */
|
| 38 | export function normalizeSpacerunSpans(htmlDocument) {
|
| 39 | htmlDocument.querySelectorAll('span[style*=spacerun]').forEach(el => {
|
| 40 | const htmlElement = el;
|
| 41 | const innerTextLength = htmlElement.innerText.length || 0;
|
| 42 | htmlElement.innerText = Array(innerTextLength + 1).join('\u00A0 ').substr(0, innerTextLength);
|
| 43 | });
|
| 44 | }
|
| 45 | /**
|
| 46 | * Normalizes specific spacing generated by Safari when content pasted from Word (`<span class="Apple-converted-space"> </span>`)
|
| 47 | * by replacing all spaces sequences longer than 1 space with ` ` pairs. This prevents spaces from being removed during
|
| 48 | * further DOM/View processing (see especially {@link module:engine/view/domconverter~DomConverter#_processDataFromDomText}).
|
| 49 | *
|
| 50 | * This function is similar to {@link module:clipboard/utils/normalizeclipboarddata normalizeClipboardData util} but uses
|
| 51 | * regular spaces / sequence for replacement.
|
| 52 | *
|
| 53 | * @param htmlString HTML string in which spacing should be normalized
|
| 54 | * @returns Input HTML with spaces normalized.
|
| 55 | */
|
| 56 | function normalizeSafariSpaceSpans(htmlString) {
|
| 57 | return htmlString.replace(/<span(?: class="Apple-converted-space"|)>(\s+)<\/span>/g, (fullMatch, spaces) => {
|
| 58 | return spaces.length === 1 ? ' ' : Array(spaces.length + 1).join('\u00A0 ').substr(0, spaces.length);
|
| 59 | });
|
| 60 | }
|