Current File : //proc/self/root/var/softaculous/sitepad/editor/site-inc/js/dist/autop.js |
this["wp"] = this["wp"] || {}; this["wp"]["autop"] =
/******/ (function(modules) { // webpackBootstrap
/******/ // The module cache
/******/ var installedModules = {};
/******/
/******/ // The require function
/******/ function __webpack_require__(moduleId) {
/******/
/******/ // Check if module is in cache
/******/ if(installedModules[moduleId]) {
/******/ return installedModules[moduleId].exports;
/******/ }
/******/ // Create a new module (and put it into the cache)
/******/ var module = installedModules[moduleId] = {
/******/ i: moduleId,
/******/ l: false,
/******/ exports: {}
/******/ };
/******/
/******/ // Execute the module function
/******/ modules[moduleId].call(module.exports, module, module.exports, __webpack_require__);
/******/
/******/ // Flag the module as loaded
/******/ module.l = true;
/******/
/******/ // Return the exports of the module
/******/ return module.exports;
/******/ }
/******/
/******/
/******/ // expose the modules object (__webpack_modules__)
/******/ __webpack_require__.m = modules;
/******/
/******/ // expose the module cache
/******/ __webpack_require__.c = installedModules;
/******/
/******/ // define getter function for harmony exports
/******/ __webpack_require__.d = function(exports, name, getter) {
/******/ if(!__webpack_require__.o(exports, name)) {
/******/ Object.defineProperty(exports, name, { enumerable: true, get: getter });
/******/ }
/******/ };
/******/
/******/ // define __esModule on exports
/******/ __webpack_require__.r = function(exports) {
/******/ if(typeof Symbol !== 'undefined' && Symbol.toStringTag) {
/******/ Object.defineProperty(exports, Symbol.toStringTag, { value: 'Module' });
/******/ }
/******/ Object.defineProperty(exports, '__esModule', { value: true });
/******/ };
/******/
/******/ // create a fake namespace object
/******/ // mode & 1: value is a module id, require it
/******/ // mode & 2: merge all properties of value into the ns
/******/ // mode & 4: return value when already ns object
/******/ // mode & 8|1: behave like require
/******/ __webpack_require__.t = function(value, mode) {
/******/ if(mode & 1) value = __webpack_require__(value);
/******/ if(mode & 8) return value;
/******/ if((mode & 4) && typeof value === 'object' && value && value.__esModule) return value;
/******/ var ns = Object.create(null);
/******/ __webpack_require__.r(ns);
/******/ Object.defineProperty(ns, 'default', { enumerable: true, value: value });
/******/ if(mode & 2 && typeof value != 'string') for(var key in value) __webpack_require__.d(ns, key, function(key) { return value[key]; }.bind(null, key));
/******/ return ns;
/******/ };
/******/
/******/ // getDefaultExport function for compatibility with non-harmony modules
/******/ __webpack_require__.n = function(module) {
/******/ var getter = module && module.__esModule ?
/******/ function getDefault() { return module['default']; } :
/******/ function getModuleExports() { return module; };
/******/ __webpack_require__.d(getter, 'a', getter);
/******/ return getter;
/******/ };
/******/
/******/ // Object.prototype.hasOwnProperty.call
/******/ __webpack_require__.o = function(object, property) { return Object.prototype.hasOwnProperty.call(object, property); };
/******/
/******/ // __webpack_public_path__
/******/ __webpack_require__.p = "";
/******/
/******/
/******/ // Load entry module and return exports
/******/ return __webpack_require__(__webpack_require__.s = 195);
/******/ })
/************************************************************************/
/******/ ({
/***/ 195:
/***/ (function(module, __webpack_exports__, __webpack_require__) {
"use strict";
__webpack_require__.r(__webpack_exports__);
/* harmony export (binding) */ __webpack_require__.d(__webpack_exports__, "autop", function() { return autop; });
/* harmony export (binding) */ __webpack_require__.d(__webpack_exports__, "removep", function() { return removep; });
/* harmony import */ var _babel_runtime_helpers_esm_slicedToArray__WEBPACK_IMPORTED_MODULE_0__ = __webpack_require__(25);
/**
* The regular expression for an HTML element.
*
* @type {String}
*/
var htmlSplitRegex = function () {
/* eslint-disable no-multi-spaces */
var comments = '!' + // Start of comment, after the <.
'(?:' + // Unroll the loop: Consume everything until --> is found.
'-(?!->)' + // Dash not followed by end of comment.
'[^\\-]*' + // Consume non-dashes.
')*' + // Loop possessively.
'(?:-->)?'; // End of comment. If not found, match all input.
var cdata = '!\\[CDATA\\[' + // Start of comment, after the <.
'[^\\]]*' + // Consume non-].
'(?:' + // Unroll the loop: Consume everything until ]]> is found.
'](?!]>)' + // One ] not followed by end of comment.
'[^\\]]*' + // Consume non-].
')*?' + // Loop possessively.
'(?:]]>)?'; // End of comment. If not found, match all input.
var escaped = '(?=' + // Is the element escaped?
'!--' + '|' + '!\\[CDATA\\[' + ')' + '((?=!-)' + // If yes, which type?
comments + '|' + cdata + ')';
var regex = '(' + // Capture the entire match.
'<' + // Find start of element.
'(' + // Conditional expression follows.
escaped + // Find end of escaped element.
'|' + // ... else ...
'[^>]*>?' + // Find end of normal element.
')' + ')';
return new RegExp(regex);
/* eslint-enable no-multi-spaces */
}();
/**
* Separate HTML elements and comments from the text.
*
* @param {string} input The text which has to be formatted.
* @return {Array} The formatted text.
*/
function htmlSplit(input) {
var parts = [];
var workingInput = input;
var match;
while (match = workingInput.match(htmlSplitRegex)) {
parts.push(workingInput.slice(0, match.index));
parts.push(match[0]);
workingInput = workingInput.slice(match.index + match[0].length);
}
if (workingInput.length) {
parts.push(workingInput);
}
return parts;
}
/**
* Replace characters or phrases within HTML elements only.
*
* @param {string} haystack The text which has to be formatted.
* @param {Object} replacePairs In the form {from: 'to', ...}.
* @return {string} The formatted text.
*/
function replaceInHtmlTags(haystack, replacePairs) {
// Find all elements.
var textArr = htmlSplit(haystack);
var changed = false; // Extract all needles.
var needles = Object.keys(replacePairs); // Loop through delimiters (elements) only.
for (var i = 1; i < textArr.length; i += 2) {
for (var j = 0; j < needles.length; j++) {
var needle = needles[j];
if (-1 !== textArr[i].indexOf(needle)) {
textArr[i] = textArr[i].replace(new RegExp(needle, 'g'), replacePairs[needle]);
changed = true; // After one strtr() break out of the foreach loop and look at next element.
break;
}
}
}
if (changed) {
haystack = textArr.join('');
}
return haystack;
}
/**
* Replaces double line-breaks with paragraph elements.
*
* A group of regex replaces used to identify text formatted with newlines and
* replace double line-breaks with HTML paragraph tags. The remaining line-
* breaks after conversion become <<br />> tags, unless br is set to 'false'.
*
* @param {string} text The text which has to be formatted.
* @param {boolean} br Optional. If set, will convert all remaining line-
* breaks after paragraphing. Default true.
* @return {string} Text which has been converted into paragraph tags.
*/
function autop(text) {
var br = arguments.length > 1 && arguments[1] !== undefined ? arguments[1] : true;
var preTags = [];
if (text.trim() === '') {
return '';
} // Just to make things a little easier, pad the end.
text = text + '\n';
/*
* Pre tags shouldn't be touched by autop.
* Replace pre tags with placeholders and bring them back after autop.
*/
if (text.indexOf('<pre') !== -1) {
var textParts = text.split('</pre>');
var lastText = textParts.pop();
text = '';
for (var i = 0; i < textParts.length; i++) {
var textPart = textParts[i];
var start = textPart.indexOf('<pre'); // Malformed html?
if (start === -1) {
text += textPart;
continue;
}
var name = '<pre wp-pre-tag-' + i + '></pre>';
preTags.push([name, textPart.substr(start) + '</pre>']);
text += textPart.substr(0, start) + name;
}
text += lastText;
} // Change multiple <br>s into two line breaks, which will turn into paragraphs.
text = text.replace(/<br\s*\/?>\s*<br\s*\/?>/g, '\n\n');
var allBlocks = '(?:table|thead|tfoot|caption|col|colgroup|tbody|tr|td|th|div|dl|dd|dt|ul|ol|li|pre|form|map|area|blockquote|address|math|style|p|h[1-6]|hr|fieldset|legend|section|article|aside|hgroup|header|footer|nav|figure|figcaption|details|menu|summary)'; // Add a double line break above block-level opening tags.
text = text.replace(new RegExp('(<' + allBlocks + '[\s\/>])', 'g'), '\n\n$1'); // Add a double line break below block-level closing tags.
text = text.replace(new RegExp('(<\/' + allBlocks + '>)', 'g'), '$1\n\n'); // Standardize newline characters to "\n".
text = text.replace(/\r\n|\r/g, '\n'); // Find newlines in all elements and add placeholders.
text = replaceInHtmlTags(text, {
'\n': ' <!-- wpnl --> '
}); // Collapse line breaks before and after <option> elements so they don't get autop'd.
if (text.indexOf('<option') !== -1) {
text = text.replace(/\s*<option/g, '<option');
text = text.replace(/<\/option>\s*/g, '</option>');
}
/*
* Collapse line breaks inside <object> elements, before <param> and <embed> elements
* so they don't get autop'd.
*/
if (text.indexOf('</object>') !== -1) {
text = text.replace(/(<object[^>]*>)\s*/g, '$1');
text = text.replace(/\s*<\/object>/g, '</object>');
text = text.replace(/\s*(<\/?(?:param|embed)[^>]*>)\s*/g, '$1');
}
/*
* Collapse line breaks inside <audio> and <video> elements,
* before and after <source> and <track> elements.
*/
if (text.indexOf('<source') !== -1 || text.indexOf('<track') !== -1) {
text = text.replace(/([<\[](?:audio|video)[^>\]]*[>\]])\s*/g, '$1');
text = text.replace(/\s*([<\[]\/(?:audio|video)[>\]])/g, '$1');
text = text.replace(/\s*(<(?:source|track)[^>]*>)\s*/g, '$1');
} // Collapse line breaks before and after <figcaption> elements.
if (text.indexOf('<figcaption') !== -1) {
text = text.replace(/\s*(<figcaption[^>]*>)/, '$1');
text = text.replace(/<\/figcaption>\s*/, '</figcaption>');
} // Remove more than two contiguous line breaks.
text = text.replace(/\n\n+/g, '\n\n'); // Split up the contents into an array of strings, separated by double line breaks.
var texts = text.split(/\n\s*\n/).filter(Boolean); // Reset text prior to rebuilding.
text = ''; // Rebuild the content as a string, wrapping every bit with a <p>.
texts.forEach(function (textPiece) {
text += '<p>' + textPiece.replace(/^\n*|\n*$/g, '') + '</p>\n';
}); // Under certain strange conditions it could create a P of entirely whitespace.
text = text.replace(/<p>\s*<\/p>/g, ''); // Add a closing <p> inside <div>, <address>, or <form> tag if missing.
text = text.replace(/<p>([^<]+)<\/(div|address|form)>/g, '<p>$1</p></$2>'); // If an opening or closing block element tag is wrapped in a <p>, unwrap it.
text = text.replace(new RegExp('<p>\s*(<\/?' + allBlocks + '[^>]*>)\s*<\/p>', 'g'), '$1'); // In some cases <li> may get wrapped in <p>, fix them.
text = text.replace(/<p>(<li.+?)<\/p>/g, '$1'); // If a <blockquote> is wrapped with a <p>, move it inside the <blockquote>.
text = text.replace(/<p><blockquote([^>]*)>/gi, '<blockquote$1><p>');
text = text.replace(/<\/blockquote><\/p>/g, '</p></blockquote>'); // If an opening or closing block element tag is preceded by an opening <p> tag, remove it.
text = text.replace(new RegExp('<p>\s*(<\/?' + allBlocks + '[^>]*>)', 'g'), '$1'); // If an opening or closing block element tag is followed by a closing <p> tag, remove it.
text = text.replace(new RegExp('(<\/?' + allBlocks + '[^>]*>)\s*<\/p>', 'g'), '$1'); // Optionally insert line breaks.
if (br) {
// Replace newlines that shouldn't be touched with a placeholder.
text = text.replace(/<(script|style).*?<\/\\1>/g, function (match) {
return match[0].replace(/\n/g, '<WPPreserveNewline />');
}); // Normalize <br>
text = text.replace(/<br>|<br\/>/g, '<br />'); // Replace any new line characters that aren't preceded by a <br /> with a <br />.
text = text.replace(/(<br \/>)?\s*\n/g, function (a, b) {
return b ? a : '<br />\n';
}); // Replace newline placeholders with newlines.
text = text.replace(/<WPPreserveNewline \/>/g, '\n');
} // If a <br /> tag is after an opening or closing block tag, remove it.
text = text.replace(new RegExp('(<\/?' + allBlocks + '[^>]*>)\s*<br \/>', 'g'), '$1'); // If a <br /> tag is before a subset of opening or closing block tags, remove it.
text = text.replace(/<br \/>(\s*<\/?(?:p|li|div|dl|dd|dt|th|pre|td|ul|ol)[^>]*>)/g, '$1');
text = text.replace(/\n<\/p>$/g, '</p>'); // Replace placeholder <pre> tags with their original content.
preTags.forEach(function (preTag) {
var _preTag = Object(_babel_runtime_helpers_esm_slicedToArray__WEBPACK_IMPORTED_MODULE_0__[/* default */ "a"])(preTag, 2),
name = _preTag[0],
original = _preTag[1];
text = text.replace(name, original);
}); // Restore newlines in all elements.
if (-1 !== text.indexOf('<!-- wpnl -->')) {
text = text.replace(/\s?<!-- wpnl -->\s?/g, '\n');
}
return text;
}
/**
* Replaces <p> tags with two line breaks. "Opposite" of autop().
*
* Replaces <p> tags with two line breaks except where the <p> has attributes.
* Unifies whitespace. Indents <li>, <dt> and <dd> for better readability.
*
* @param {string} html The content from the editor.
* @return {string} The content with stripped paragraph tags.
*/
function removep(html) {
var blocklist = 'blockquote|ul|ol|li|dl|dt|dd|table|thead|tbody|tfoot|tr|th|td|h[1-6]|fieldset|figure';
var blocklist1 = blocklist + '|div|p';
var blocklist2 = blocklist + '|pre';
var preserve = [];
var preserveLinebreaks = false;
var preserveBr = false;
if (!html) {
return '';
} // Protect script and style tags.
if (html.indexOf('<script') !== -1 || html.indexOf('<style') !== -1) {
html = html.replace(/<(script|style)[^>]*>[\s\S]*?<\/\1>/g, function (match) {
preserve.push(match);
return '<wp-preserve>';
});
} // Protect pre tags.
if (html.indexOf('<pre') !== -1) {
preserveLinebreaks = true;
html = html.replace(/<pre[^>]*>[\s\S]+?<\/pre>/g, function (a) {
a = a.replace(/<br ?\/?>(\r\n|\n)?/g, '<wp-line-break>');
a = a.replace(/<\/?p( [^>]*)?>(\r\n|\n)?/g, '<wp-line-break>');
return a.replace(/\r?\n/g, '<wp-line-break>');
});
} // Remove line breaks but keep <br> tags inside image captions.
if (html.indexOf('[caption') !== -1) {
preserveBr = true;
html = html.replace(/\[caption[\s\S]+?\[\/caption\]/g, function (a) {
return a.replace(/<br([^>]*)>/g, '<wp-temp-br$1>').replace(/[\r\n\t]+/, '');
});
} // Normalize white space characters before and after block tags.
html = html.replace(new RegExp('\\s*</(' + blocklist1 + ')>\\s*', 'g'), '</$1>\n');
html = html.replace(new RegExp('\\s*<((?:' + blocklist1 + ')(?: [^>]*)?)>', 'g'), '\n<$1>'); // Mark </p> if it has any attributes.
html = html.replace(/(<p [^>]+>.*?)<\/p>/g, '$1</p#>'); // Preserve the first <p> inside a <div>.
html = html.replace(/<div( [^>]*)?>\s*<p>/gi, '<div$1>\n\n'); // Remove paragraph tags.
html = html.replace(/\s*<p>/gi, '');
html = html.replace(/\s*<\/p>\s*/gi, '\n\n'); // Normalize white space chars and remove multiple line breaks.
html = html.replace(/\n[\s\u00a0]+\n/g, '\n\n'); // Replace <br> tags with line breaks.
html = html.replace(/(\s*)<br ?\/?>\s*/gi, function (match, space) {
if (space && space.indexOf('\n') !== -1) {
return '\n\n';
}
return '\n';
}); // Fix line breaks around <div>.
html = html.replace(/\s*<div/g, '\n<div');
html = html.replace(/<\/div>\s*/g, '</div>\n'); // Fix line breaks around caption shortcodes.
html = html.replace(/\s*\[caption([^\[]+)\[\/caption\]\s*/gi, '\n\n[caption$1[/caption]\n\n');
html = html.replace(/caption\]\n\n+\[caption/g, 'caption]\n\n[caption'); // Pad block elements tags with a line break.
html = html.replace(new RegExp('\\s*<((?:' + blocklist2 + ')(?: [^>]*)?)\\s*>', 'g'), '\n<$1>');
html = html.replace(new RegExp('\\s*</(' + blocklist2 + ')>\\s*', 'g'), '</$1>\n'); // Indent <li>, <dt> and <dd> tags.
html = html.replace(/<((li|dt|dd)[^>]*)>/g, ' \t<$1>'); // Fix line breaks around <select> and <option>.
if (html.indexOf('<option') !== -1) {
html = html.replace(/\s*<option/g, '\n<option');
html = html.replace(/\s*<\/select>/g, '\n</select>');
} // Pad <hr> with two line breaks.
if (html.indexOf('<hr') !== -1) {
html = html.replace(/\s*<hr( [^>]*)?>\s*/g, '\n\n<hr$1>\n\n');
} // Remove line breaks in <object> tags.
if (html.indexOf('<object') !== -1) {
html = html.replace(/<object[\s\S]+?<\/object>/g, function (a) {
return a.replace(/[\r\n]+/g, '');
});
} // Unmark special paragraph closing tags.
html = html.replace(/<\/p#>/g, '</p>\n'); // Pad remaining <p> tags whit a line break.
html = html.replace(/\s*(<p [^>]+>[\s\S]*?<\/p>)/g, '\n$1'); // Trim.
html = html.replace(/^\s+/, '');
html = html.replace(/[\s\u00a0]+$/, '');
if (preserveLinebreaks) {
html = html.replace(/<wp-line-break>/g, '\n');
}
if (preserveBr) {
html = html.replace(/<wp-temp-br([^>]*)>/g, '<br$1>');
} // Restore preserved tags.
if (preserve.length) {
html = html.replace(/<wp-preserve>/g, function () {
return preserve.shift();
});
}
return html;
}
/***/ }),
/***/ 25:
/***/ (function(module, __webpack_exports__, __webpack_require__) {
"use strict";
// EXTERNAL MODULE: ./node_modules/@babel/runtime/helpers/esm/arrayWithHoles.js
var arrayWithHoles = __webpack_require__(35);
// CONCATENATED MODULE: ./node_modules/@babel/runtime/helpers/esm/iterableToArrayLimit.js
function _iterableToArrayLimit(arr, i) {
var _arr = [];
var _n = true;
var _d = false;
var _e = undefined;
try {
for (var _i = arr[Symbol.iterator](), _s; !(_n = (_s = _i.next()).done); _n = true) {
_arr.push(_s.value);
if (i && _arr.length === i) break;
}
} catch (err) {
_d = true;
_e = err;
} finally {
try {
if (!_n && _i["return"] != null) _i["return"]();
} finally {
if (_d) throw _e;
}
}
return _arr;
}
// EXTERNAL MODULE: ./node_modules/@babel/runtime/helpers/esm/nonIterableRest.js
var nonIterableRest = __webpack_require__(36);
// CONCATENATED MODULE: ./node_modules/@babel/runtime/helpers/esm/slicedToArray.js
/* harmony export (binding) */ __webpack_require__.d(__webpack_exports__, "a", function() { return _slicedToArray; });
function _slicedToArray(arr, i) {
return Object(arrayWithHoles["a" /* default */])(arr) || _iterableToArrayLimit(arr, i) || Object(nonIterableRest["a" /* default */])();
}
/***/ }),
/***/ 35:
/***/ (function(module, __webpack_exports__, __webpack_require__) {
"use strict";
/* harmony export (binding) */ __webpack_require__.d(__webpack_exports__, "a", function() { return _arrayWithHoles; });
function _arrayWithHoles(arr) {
if (Array.isArray(arr)) return arr;
}
/***/ }),
/***/ 36:
/***/ (function(module, __webpack_exports__, __webpack_require__) {
"use strict";
/* harmony export (binding) */ __webpack_require__.d(__webpack_exports__, "a", function() { return _nonIterableRest; });
function _nonIterableRest() {
throw new TypeError("Invalid attempt to destructure non-iterable instance");
}
/***/ })
/******/ });