@promptbook/markdown-utils 0.77.1 → 0.78.0-0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/umd/index.umd.js CHANGED
@@ -1,8 +1,8 @@
1
1
  (function (global, factory) {
2
- typeof exports === 'object' && typeof module !== 'undefined' ? factory(exports, require('spacetrim'), require('prettier'), require('prettier/parser-html'), require('path'), require('waitasecond'), require('crypto-js'), require('crypto-js/enc-hex'), require('mime-types'), require('papaparse')) :
3
- typeof define === 'function' && define.amd ? define(['exports', 'spacetrim', 'prettier', 'prettier/parser-html', 'path', 'waitasecond', 'crypto-js', 'crypto-js/enc-hex', 'mime-types', 'papaparse'], factory) :
4
- (global = typeof globalThis !== 'undefined' ? globalThis : global || self, factory(global["promptbook-markdown-utils"] = {}, global.spaceTrim, global.prettier, global.parserHtml, global.path, global.waitasecond, global.cryptoJs, global.hexEncoder, global.mimeTypes, global.papaparse));
5
- })(this, (function (exports, spaceTrim, prettier, parserHtml, path, waitasecond, cryptoJs, hexEncoder, mimeTypes, papaparse) { 'use strict';
2
+ typeof exports === 'object' && typeof module !== 'undefined' ? factory(exports, require('spacetrim'), require('prettier'), require('prettier/parser-html'), require('waitasecond'), require('path'), require('crypto-js'), require('crypto-js/enc-hex'), require('mime-types'), require('papaparse')) :
3
+ typeof define === 'function' && define.amd ? define(['exports', 'spacetrim', 'prettier', 'prettier/parser-html', 'waitasecond', 'path', 'crypto-js', 'crypto-js/enc-hex', 'mime-types', 'papaparse'], factory) :
4
+ (global = typeof globalThis !== 'undefined' ? globalThis : global || self, factory(global["promptbook-markdown-utils"] = {}, global.spaceTrim, global.prettier, global.parserHtml, global.waitasecond, global.path, global.cryptoJs, global.hexEncoder, global.mimeTypes, global.papaparse));
5
+ })(this, (function (exports, spaceTrim, prettier, parserHtml, waitasecond, path, cryptoJs, hexEncoder, mimeTypes, papaparse) { 'use strict';
6
6
 
7
7
  function _interopDefaultLegacy (e) { return e && typeof e === 'object' && 'default' in e ? e : { 'default': e }; }
8
8
 
@@ -22,7 +22,7 @@
22
22
  *
23
23
  * @see https://github.com/webgptorg/promptbook
24
24
  */
25
- var PROMPTBOOK_ENGINE_VERSION = '0.77.0';
25
+ var PROMPTBOOK_ENGINE_VERSION = '0.77.1';
26
26
  /**
27
27
  * TODO: string_promptbook_version should be constrained to the all versions of Promptbook engine
28
28
  * Note: [💞] Ignore a discrepancy between file name and entity name
@@ -1564,415 +1564,26 @@
1564
1564
  return new (SimplePipelineCollection.bind.apply(SimplePipelineCollection, __spreadArray([void 0], __read(promptbooks), false)))();
1565
1565
  }
1566
1566
 
1567
- var defaultDiacriticsRemovalMap = [
1568
- {
1569
- base: 'A',
1570
- letters: '\u0041\u24B6\uFF21\u00C0\u00C1\u00C2\u1EA6\u1EA4\u1EAA\u1EA8\u00C3\u0100\u0102\u1EB0\u1EAE\u1EB4\u1EB2\u0226\u01E0\u00C4\u01DE\u1EA2\u00C5\u01FA\u01CD\u0200\u0202\u1EA0\u1EAC\u1EB6\u1E00\u0104\u023A\u2C6F',
1571
- },
1572
- { base: 'AA', letters: '\uA732' },
1573
- { base: 'AE', letters: '\u00C6\u01FC\u01E2' },
1574
- { base: 'AO', letters: '\uA734' },
1575
- { base: 'AU', letters: '\uA736' },
1576
- { base: 'AV', letters: '\uA738\uA73A' },
1577
- { base: 'AY', letters: '\uA73C' },
1578
- {
1579
- base: 'B',
1580
- letters: '\u0042\u24B7\uFF22\u1E02\u1E04\u1E06\u0243\u0182\u0181',
1581
- },
1582
- {
1583
- base: 'C',
1584
- letters: '\u0043\u24B8\uFF23\u0106\u0108\u010A\u010C\u00C7\u1E08\u0187\u023B\uA73E',
1585
- },
1586
- {
1587
- base: 'D',
1588
- letters: '\u0044\u24B9\uFF24\u1E0A\u010E\u1E0C\u1E10\u1E12\u1E0E\u0110\u018B\u018A\u0189\uA779\u00D0',
1589
- },
1590
- { base: 'DZ', letters: '\u01F1\u01C4' },
1591
- { base: 'Dz', letters: '\u01F2\u01C5' },
1592
- {
1593
- base: 'E',
1594
- letters: '\u0045\u24BA\uFF25\u00C8\u00C9\u00CA\u1EC0\u1EBE\u1EC4\u1EC2\u1EBC\u0112\u1E14\u1E16\u0114\u0116\u00CB\u1EBA\u011A\u0204\u0206\u1EB8\u1EC6\u0228\u1E1C\u0118\u1E18\u1E1A\u0190\u018E',
1595
- },
1596
- { base: 'F', letters: '\u0046\u24BB\uFF26\u1E1E\u0191\uA77B' },
1597
- {
1598
- base: 'G',
1599
- letters: '\u0047\u24BC\uFF27\u01F4\u011C\u1E20\u011E\u0120\u01E6\u0122\u01E4\u0193\uA7A0\uA77D\uA77E',
1600
- },
1601
- {
1602
- base: 'H',
1603
- letters: '\u0048\u24BD\uFF28\u0124\u1E22\u1E26\u021E\u1E24\u1E28\u1E2A\u0126\u2C67\u2C75\uA78D',
1604
- },
1605
- {
1606
- base: 'I',
1607
- letters: '\u0049\u24BE\uFF29\u00CC\u00CD\u00CE\u0128\u012A\u012C\u0130\u00CF\u1E2E\u1EC8\u01CF\u0208\u020A\u1ECA\u012E\u1E2C\u0197',
1608
- },
1609
- { base: 'J', letters: '\u004A\u24BF\uFF2A\u0134\u0248' },
1610
- {
1611
- base: 'K',
1612
- letters: '\u004B\u24C0\uFF2B\u1E30\u01E8\u1E32\u0136\u1E34\u0198\u2C69\uA740\uA742\uA744\uA7A2',
1613
- },
1614
- {
1615
- base: 'L',
1616
- letters: '\u004C\u24C1\uFF2C\u013F\u0139\u013D\u1E36\u1E38\u013B\u1E3C\u1E3A\u0141\u023D\u2C62\u2C60\uA748\uA746\uA780',
1617
- },
1618
- { base: 'LJ', letters: '\u01C7' },
1619
- { base: 'Lj', letters: '\u01C8' },
1620
- { base: 'M', letters: '\u004D\u24C2\uFF2D\u1E3E\u1E40\u1E42\u2C6E\u019C' },
1621
- {
1622
- base: 'N',
1623
- letters: '\u004E\u24C3\uFF2E\u01F8\u0143\u00D1\u1E44\u0147\u1E46\u0145\u1E4A\u1E48\u0220\u019D\uA790\uA7A4',
1624
- },
1625
- { base: 'NJ', letters: '\u01CA' },
1626
- { base: 'Nj', letters: '\u01CB' },
1627
- {
1628
- base: 'O',
1629
- letters: '\u004F\u24C4\uFF2F\u00D2\u00D3\u00D4\u1ED2\u1ED0\u1ED6\u1ED4\u00D5\u1E4C\u022C\u1E4E\u014C\u1E50\u1E52\u014E\u022E\u0230\u00D6\u022A\u1ECE\u0150\u01D1\u020C\u020E\u01A0\u1EDC\u1EDA\u1EE0\u1EDE\u1EE2\u1ECC\u1ED8\u01EA\u01EC\u00D8\u01FE\u0186\u019F\uA74A\uA74C',
1630
- },
1631
- { base: 'OI', letters: '\u01A2' },
1632
- { base: 'OO', letters: '\uA74E' },
1633
- { base: 'OU', letters: '\u0222' },
1634
- { base: 'OE', letters: '\u008C\u0152' },
1635
- { base: 'oe', letters: '\u009C\u0153' },
1636
- {
1637
- base: 'P',
1638
- letters: '\u0050\u24C5\uFF30\u1E54\u1E56\u01A4\u2C63\uA750\uA752\uA754',
1639
- },
1640
- { base: 'Q', letters: '\u0051\u24C6\uFF31\uA756\uA758\u024A' },
1641
- {
1642
- base: 'R',
1643
- letters: '\u0052\u24C7\uFF32\u0154\u1E58\u0158\u0210\u0212\u1E5A\u1E5C\u0156\u1E5E\u024C\u2C64\uA75A\uA7A6\uA782',
1644
- },
1645
- {
1646
- base: 'S',
1647
- letters: '\u0053\u24C8\uFF33\u1E9E\u015A\u1E64\u015C\u1E60\u0160\u1E66\u1E62\u1E68\u0218\u015E\u2C7E\uA7A8\uA784',
1648
- },
1649
- {
1650
- base: 'T',
1651
- letters: '\u0054\u24C9\uFF34\u1E6A\u0164\u1E6C\u021A\u0162\u1E70\u1E6E\u0166\u01AC\u01AE\u023E\uA786',
1652
- },
1653
- { base: 'TZ', letters: '\uA728' },
1654
- {
1655
- base: 'U',
1656
- letters: '\u0055\u24CA\uFF35\u00D9\u00DA\u00DB\u0168\u1E78\u016A\u1E7A\u016C\u00DC\u01DB\u01D7\u01D5\u01D9\u1EE6\u016E\u0170\u01D3\u0214\u0216\u01AF\u1EEA\u1EE8\u1EEE\u1EEC\u1EF0\u1EE4\u1E72\u0172\u1E76\u1E74\u0244',
1657
- },
1658
- { base: 'V', letters: '\u0056\u24CB\uFF36\u1E7C\u1E7E\u01B2\uA75E\u0245' },
1659
- { base: 'VY', letters: '\uA760' },
1660
- {
1661
- base: 'W',
1662
- letters: '\u0057\u24CC\uFF37\u1E80\u1E82\u0174\u1E86\u1E84\u1E88\u2C72',
1663
- },
1664
- { base: 'X', letters: '\u0058\u24CD\uFF38\u1E8A\u1E8C' },
1665
- {
1666
- base: 'Y',
1667
- letters: '\u0059\u24CE\uFF39\u1EF2\u00DD\u0176\u1EF8\u0232\u1E8E\u0178\u1EF6\u1EF4\u01B3\u024E\u1EFE',
1668
- },
1669
- {
1670
- base: 'Z',
1671
- letters: '\u005A\u24CF\uFF3A\u0179\u1E90\u017B\u017D\u1E92\u1E94\u01B5\u0224\u2C7F\u2C6B\uA762',
1672
- },
1673
- {
1674
- base: 'a',
1675
- letters: '\u0061\u24D0\uFF41\u1E9A\u00E0\u00E1\u00E2\u1EA7\u1EA5\u1EAB\u1EA9\u00E3\u0101\u0103\u1EB1\u1EAF\u1EB5\u1EB3\u0227\u01E1\u00E4\u01DF\u1EA3\u00E5\u01FB\u01CE\u0201\u0203\u1EA1\u1EAD\u1EB7\u1E01\u0105\u2C65\u0250',
1676
- },
1677
- { base: 'aa', letters: '\uA733' },
1678
- { base: 'ae', letters: '\u00E6\u01FD\u01E3' },
1679
- { base: 'ao', letters: '\uA735' },
1680
- { base: 'au', letters: '\uA737' },
1681
- { base: 'av', letters: '\uA739\uA73B' },
1682
- { base: 'ay', letters: '\uA73D' },
1683
- {
1684
- base: 'b',
1685
- letters: '\u0062\u24D1\uFF42\u1E03\u1E05\u1E07\u0180\u0183\u0253',
1686
- },
1687
- {
1688
- base: 'c',
1689
- letters: '\u0063\u24D2\uFF43\u0107\u0109\u010B\u010D\u00E7\u1E09\u0188\u023C\uA73F\u2184',
1690
- },
1691
- {
1692
- base: 'd',
1693
- letters: '\u0064\u24D3\uFF44\u1E0B\u010F\u1E0D\u1E11\u1E13\u1E0F\u0111\u018C\u0256\u0257\uA77A',
1694
- },
1695
- { base: 'dz', letters: '\u01F3\u01C6' },
1696
- {
1697
- base: 'e',
1698
- letters: '\u0065\u24D4\uFF45\u00E8\u00E9\u00EA\u1EC1\u1EBF\u1EC5\u1EC3\u1EBD\u0113\u1E15\u1E17\u0115\u0117\u00EB\u1EBB\u011B\u0205\u0207\u1EB9\u1EC7\u0229\u1E1D\u0119\u1E19\u1E1B\u0247\u025B\u01DD',
1699
- },
1700
- { base: 'f', letters: '\u0066\u24D5\uFF46\u1E1F\u0192\uA77C' },
1701
- {
1702
- base: 'g',
1703
- letters: '\u0067\u24D6\uFF47\u01F5\u011D\u1E21\u011F\u0121\u01E7\u0123\u01E5\u0260\uA7A1\u1D79\uA77F',
1704
- },
1705
- {
1706
- base: 'h',
1707
- letters: '\u0068\u24D7\uFF48\u0125\u1E23\u1E27\u021F\u1E25\u1E29\u1E2B\u1E96\u0127\u2C68\u2C76\u0265',
1708
- },
1709
- { base: 'hv', letters: '\u0195' },
1710
- {
1711
- base: 'i',
1712
- letters: '\u0069\u24D8\uFF49\u00EC\u00ED\u00EE\u0129\u012B\u012D\u00EF\u1E2F\u1EC9\u01D0\u0209\u020B\u1ECB\u012F\u1E2D\u0268\u0131',
1713
- },
1714
- { base: 'j', letters: '\u006A\u24D9\uFF4A\u0135\u01F0\u0249' },
1715
- {
1716
- base: 'k',
1717
- letters: '\u006B\u24DA\uFF4B\u1E31\u01E9\u1E33\u0137\u1E35\u0199\u2C6A\uA741\uA743\uA745\uA7A3',
1718
- },
1719
- {
1720
- base: 'l',
1721
- letters: '\u006C\u24DB\uFF4C\u0140\u013A\u013E\u1E37\u1E39\u013C\u1E3D\u1E3B\u017F\u0142\u019A\u026B\u2C61\uA749\uA781\uA747',
1722
- },
1723
- { base: 'lj', letters: '\u01C9' },
1724
- { base: 'm', letters: '\u006D\u24DC\uFF4D\u1E3F\u1E41\u1E43\u0271\u026F' },
1725
- {
1726
- base: 'n',
1727
- letters: '\u006E\u24DD\uFF4E\u01F9\u0144\u00F1\u1E45\u0148\u1E47\u0146\u1E4B\u1E49\u019E\u0272\u0149\uA791\uA7A5',
1728
- },
1729
- { base: 'nj', letters: '\u01CC' },
1730
- {
1731
- base: 'o',
1732
- letters: '\u006F\u24DE\uFF4F\u00F2\u00F3\u00F4\u1ED3\u1ED1\u1ED7\u1ED5\u00F5\u1E4D\u022D\u1E4F\u014D\u1E51\u1E53\u014F\u022F\u0231\u00F6\u022B\u1ECF\u0151\u01D2\u020D\u020F\u01A1\u1EDD\u1EDB\u1EE1\u1EDF\u1EE3\u1ECD\u1ED9\u01EB\u01ED\u00F8\u01FF\u0254\uA74B\uA74D\u0275',
1733
- },
1734
- { base: 'oi', letters: '\u01A3' },
1735
- { base: 'ou', letters: '\u0223' },
1736
- { base: 'oo', letters: '\uA74F' },
1737
- {
1738
- base: 'p',
1739
- letters: '\u0070\u24DF\uFF50\u1E55\u1E57\u01A5\u1D7D\uA751\uA753\uA755',
1740
- },
1741
- { base: 'q', letters: '\u0071\u24E0\uFF51\u024B\uA757\uA759' },
1742
- {
1743
- base: 'r',
1744
- letters: '\u0072\u24E1\uFF52\u0155\u1E59\u0159\u0211\u0213\u1E5B\u1E5D\u0157\u1E5F\u024D\u027D\uA75B\uA7A7\uA783',
1745
- },
1746
- {
1747
- base: 's',
1748
- letters: '\u0073\u24E2\uFF53\u00DF\u015B\u1E65\u015D\u1E61\u0161\u1E67\u1E63\u1E69\u0219\u015F\u023F\uA7A9\uA785\u1E9B',
1749
- },
1750
- {
1751
- base: 't',
1752
- letters: '\u0074\u24E3\uFF54\u1E6B\u1E97\u0165\u1E6D\u021B\u0163\u1E71\u1E6F\u0167\u01AD\u0288\u2C66\uA787',
1753
- },
1754
- { base: 'tz', letters: '\uA729' },
1755
- {
1756
- base: 'u',
1757
- letters: '\u0075\u24E4\uFF55\u00F9\u00FA\u00FB\u0169\u1E79\u016B\u1E7B\u016D\u00FC\u01DC\u01D8\u01D6\u01DA\u1EE7\u016F\u0171\u01D4\u0215\u0217\u01B0\u1EEB\u1EE9\u1EEF\u1EED\u1EF1\u1EE5\u1E73\u0173\u1E77\u1E75\u0289',
1758
- },
1759
- { base: 'v', letters: '\u0076\u24E5\uFF56\u1E7D\u1E7F\u028B\uA75F\u028C' },
1760
- { base: 'vy', letters: '\uA761' },
1761
- {
1762
- base: 'w',
1763
- letters: '\u0077\u24E6\uFF57\u1E81\u1E83\u0175\u1E87\u1E85\u1E98\u1E89\u2C73',
1764
- },
1765
- { base: 'x', letters: '\u0078\u24E7\uFF58\u1E8B\u1E8D' },
1766
- {
1767
- base: 'y',
1768
- letters: '\u0079\u24E8\uFF59\u1EF3\u00FD\u0177\u1EF9\u0233\u1E8F\u00FF\u1EF7\u1E99\u1EF5\u01B4\u024F\u1EFF',
1769
- },
1770
- {
1771
- base: 'z',
1772
- letters: '\u007A\u24E9\uFF5A\u017A\u1E91\u017C\u017E\u1E93\u1E95\u01B6\u0225\u0240\u2C6C\uA763',
1773
- },
1774
- ];
1775
1567
  /**
1776
- * Map of letters from diacritic variant to diacritless variant
1777
- * Contains lowercase and uppercase separatelly
1778
- *
1779
- * > "á" => "a"
1780
- * > "ě" => "e"
1781
- * > "Ă" => "A"
1782
- * > ...
1568
+ * This error type indicates that some tools are missing for pipeline execution or preparation
1783
1569
  *
1784
- * @public exported from `@promptbook/utils`
1570
+ * @public exported from `@promptbook/core`
1785
1571
  */
1786
- var DIACRITIC_VARIANTS_LETTERS = {};
1787
- // tslint:disable-next-line: prefer-for-of
1788
- for (var i = 0; i < defaultDiacriticsRemovalMap.length; i++) {
1789
- var letters = defaultDiacriticsRemovalMap[i].letters;
1790
- // tslint:disable-next-line: prefer-for-of
1791
- for (var j = 0; j < letters.length; j++) {
1792
- DIACRITIC_VARIANTS_LETTERS[letters[j]] = defaultDiacriticsRemovalMap[i].base;
1572
+ var MissingToolsError = /** @class */ (function (_super) {
1573
+ __extends(MissingToolsError, _super);
1574
+ function MissingToolsError(message) {
1575
+ var _this = _super.call(this, spaceTrim.spaceTrim(function (block) { return "\n ".concat(block(message), "\n\n Note: You have probbably forgot to provide some tools for pipeline execution or preparation\n\n "); })) || this;
1576
+ _this.name = 'MissingToolsError';
1577
+ Object.setPrototypeOf(_this, MissingToolsError.prototype);
1578
+ return _this;
1793
1579
  }
1794
- }
1795
- // <- TODO: [🍓] Put to maker function to save execution time if not needed
1796
- /*
1797
- @see https://stackoverflow.com/questions/990904/remove-accents-diacritics-in-a-string-in-javascript
1798
- Licensed under the Apache License, Version 2.0 (the "License");
1799
- you may not use this file except in compliance with the License.
1800
- You may obtain a copy of the License at
1801
-
1802
- http://www.apache.org/licenses/LICENSE-2.0
1803
-
1804
- Unless required by applicable law or agreed to in writing, software
1805
- distributed under the License is distributed on an "AS IS" BASIS,
1806
- WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
1807
- See the License for the specific language governing permissions and
1808
- limitations under the License.
1809
- */
1580
+ return MissingToolsError;
1581
+ }(Error));
1810
1582
 
1811
1583
  /**
1812
- * @@@
1584
+ * This error indicates errors during the execution of the pipeline
1813
1585
  *
1814
- * @param input @@@
1815
- * @returns @@@
1816
- * @public exported from `@promptbook/utils`
1817
- */
1818
- function removeDiacritics(input) {
1819
- /*eslint no-control-regex: "off"*/
1820
- return input.replace(/[^\u0000-\u007E]/g, function (a) {
1821
- return DIACRITIC_VARIANTS_LETTERS[a] || a;
1822
- });
1823
- }
1824
- /**
1825
- * TODO: [Ж] Variant for cyrillic (and in general non-latin) letters
1826
- */
1827
-
1828
- /**
1829
- * @@@
1830
- *
1831
- * @param text @@@
1832
- * @returns @@@
1833
- * @example 'hello-world'
1834
- * @example 'i-love-promptbook'
1835
- * @public exported from `@promptbook/utils`
1836
- */
1837
- function normalizeToKebabCase(text) {
1838
- var e_1, _a;
1839
- text = removeDiacritics(text);
1840
- var charType;
1841
- var lastCharType = 'OTHER';
1842
- var normalizedName = '';
1843
- try {
1844
- for (var text_1 = __values(text), text_1_1 = text_1.next(); !text_1_1.done; text_1_1 = text_1.next()) {
1845
- var char = text_1_1.value;
1846
- var normalizedChar = void 0;
1847
- if (/^[a-z]$/.test(char)) {
1848
- charType = 'LOWERCASE';
1849
- normalizedChar = char;
1850
- }
1851
- else if (/^[A-Z]$/.test(char)) {
1852
- charType = 'UPPERCASE';
1853
- normalizedChar = char.toLowerCase();
1854
- }
1855
- else if (/^[0-9]$/.test(char)) {
1856
- charType = 'NUMBER';
1857
- normalizedChar = char;
1858
- }
1859
- else {
1860
- charType = 'OTHER';
1861
- normalizedChar = '-';
1862
- }
1863
- if (charType !== lastCharType &&
1864
- !(lastCharType === 'UPPERCASE' && charType === 'LOWERCASE') &&
1865
- !(lastCharType === 'NUMBER') &&
1866
- !(charType === 'NUMBER')) {
1867
- normalizedName += '-';
1868
- }
1869
- normalizedName += normalizedChar;
1870
- lastCharType = charType;
1871
- }
1872
- }
1873
- catch (e_1_1) { e_1 = { error: e_1_1 }; }
1874
- finally {
1875
- try {
1876
- if (text_1_1 && !text_1_1.done && (_a = text_1.return)) _a.call(text_1);
1877
- }
1878
- finally { if (e_1) throw e_1.error; }
1879
- }
1880
- normalizedName = normalizedName.split(/-+/g).join('-');
1881
- normalizedName = normalizedName.split(/-?\/-?/g).join('/');
1882
- normalizedName = normalizedName.replace(/^-/, '');
1883
- normalizedName = normalizedName.replace(/-$/, '');
1884
- return normalizedName;
1885
- }
1886
- /**
1887
- * Note: [💞] Ignore a discrepancy between file name and entity name
1888
- */
1889
-
1890
- /**
1891
- * Removes emojis from a string and fix whitespaces
1892
- *
1893
- * @param text with emojis
1894
- * @returns text without emojis
1895
- * @public exported from `@promptbook/utils`
1896
- */
1897
- function removeEmojis(text) {
1898
- // Replace emojis (and also ZWJ sequence) with hyphens
1899
- text = text.replace(/(\p{Extended_Pictographic})\p{Modifier_Symbol}/gu, '$1');
1900
- text = text.replace(/(\p{Extended_Pictographic})[\u{FE00}-\u{FE0F}]/gu, '$1');
1901
- text = text.replace(/(\p{Extended_Pictographic})(\u{200D}\p{Extended_Pictographic})*/gu, '$1');
1902
- text = text.replace(/\p{Extended_Pictographic}/gu, '');
1903
- return text;
1904
- }
1905
-
1906
- /**
1907
- * Tests if given string is valid URL.
1908
- *
1909
- * Note: This does not check if the file exists only if the path is valid
1910
- * @public exported from `@promptbook/utils`
1911
- */
1912
- function isValidFilePath(filename) {
1913
- if (typeof filename !== 'string') {
1914
- return false;
1915
- }
1916
- var filenameSlashes = filename.split('\\').join('/');
1917
- // Absolute Unix path: /hello.txt
1918
- if (/^(\/)/i.test(filenameSlashes)) {
1919
- return true;
1920
- }
1921
- // Absolute Windows path: /hello.txt
1922
- if (/^([A-Z]{1,2}:\/?)\//i.test(filenameSlashes)) {
1923
- return true;
1924
- }
1925
- // Relative path: ./hello.txt
1926
- if (/^(\.\.?\/)+/i.test(filenameSlashes)) {
1927
- return true;
1928
- }
1929
- return false;
1930
- }
1931
-
1932
- /**
1933
- * @@@
1934
- *
1935
- * @param value @@@
1936
- * @returns @@@
1937
- * @example @@@
1938
- * @public exported from `@promptbook/utils`
1939
- */
1940
- function titleToName(value) {
1941
- if (isValidUrl(value)) {
1942
- value = value.replace(/^https?:\/\//, '');
1943
- value = value.replace(/\.html$/, '');
1944
- }
1945
- else if (isValidFilePath(value)) {
1946
- value = path.basename(value);
1947
- // Note: Keeping extension in the name
1948
- }
1949
- value = value.split('/').join('-');
1950
- value = removeEmojis(value);
1951
- value = normalizeToKebabCase(value);
1952
- // TODO: [🧠] Maybe warn or add some padding to short name which are not good identifiers
1953
- return value;
1954
- }
1955
-
1956
- /**
1957
- * This error type indicates that some tools are missing for pipeline execution or preparation
1958
- *
1959
- * @public exported from `@promptbook/core`
1960
- */
1961
- var MissingToolsError = /** @class */ (function (_super) {
1962
- __extends(MissingToolsError, _super);
1963
- function MissingToolsError(message) {
1964
- var _this = _super.call(this, spaceTrim.spaceTrim(function (block) { return "\n ".concat(block(message), "\n\n Note: You have probbably forgot to provide some tools for pipeline execution or preparation\n\n "); })) || this;
1965
- _this.name = 'MissingToolsError';
1966
- Object.setPrototypeOf(_this, MissingToolsError.prototype);
1967
- return _this;
1968
- }
1969
- return MissingToolsError;
1970
- }(Error));
1971
-
1972
- /**
1973
- * This error indicates errors during the execution of the pipeline
1974
- *
1975
- * @public exported from `@promptbook/core`
1586
+ * @public exported from `@promptbook/core`
1976
1587
  */
1977
1588
  var PipelineExecutionError = /** @class */ (function (_super) {
1978
1589
  __extends(PipelineExecutionError, _super);
@@ -3102,148 +2713,471 @@
3102
2713
  return $Register;
3103
2714
  }());
3104
2715
 
3105
- /**
3106
- * @@@
3107
- *
3108
- * Note: `$` is used to indicate that this interacts with the global scope
3109
- * @singleton Only one instance of each register is created per build, but thare can be more @@@
3110
- * @public exported from `@promptbook/core`
3111
- */
3112
- var $scrapersMetadataRegister = new $Register('scrapers_metadata');
3113
- /**
3114
- * TODO: [®] DRY Register logic
3115
- */
2716
+ /**
2717
+ * @@@
2718
+ *
2719
+ * Note: `$` is used to indicate that this interacts with the global scope
2720
+ * @singleton Only one instance of each register is created per build, but thare can be more @@@
2721
+ * @public exported from `@promptbook/core`
2722
+ */
2723
+ var $scrapersMetadataRegister = new $Register('scrapers_metadata');
2724
+ /**
2725
+ * TODO: [®] DRY Register logic
2726
+ */
2727
+
2728
+ /**
2729
+ * @@@
2730
+ *
2731
+ * Note: `$` is used to indicate that this interacts with the global scope
2732
+ * @singleton Only one instance of each register is created per build, but thare can be more @@@
2733
+ * @public exported from `@promptbook/core`
2734
+ */
2735
+ var $scrapersRegister = new $Register('scraper_constructors');
2736
+ /**
2737
+ * TODO: [®] DRY Register logic
2738
+ */
2739
+
2740
+ /**
2741
+ * Creates a message with all registered scrapers
2742
+ *
2743
+ * Note: This function is used to create a (error) message when there is no scraper for particular mime type
2744
+ *
2745
+ * @private internal function of `createScrapersFromConfiguration` and `createScrapersFromEnv`
2746
+ */
2747
+ function $registeredScrapersMessage(availableScrapers) {
2748
+ var e_1, _a, e_2, _b, e_3, _c;
2749
+ /**
2750
+ * Mixes registered scrapers from $scrapersMetadataRegister and $scrapersRegister
2751
+ */
2752
+ var all = [];
2753
+ var _loop_1 = function (packageName, className, mimeTypes, documentationUrl, isAvilableInBrowser) {
2754
+ if (all.some(function (item) { return item.packageName === packageName && item.className === className; })) {
2755
+ return "continue";
2756
+ }
2757
+ all.push({ packageName: packageName, className: className, mimeTypes: mimeTypes, documentationUrl: documentationUrl, isAvilableInBrowser: isAvilableInBrowser });
2758
+ };
2759
+ try {
2760
+ for (var _d = __values($scrapersMetadataRegister.list()), _e = _d.next(); !_e.done; _e = _d.next()) {
2761
+ var _f = _e.value, packageName = _f.packageName, className = _f.className, mimeTypes = _f.mimeTypes, documentationUrl = _f.documentationUrl, isAvilableInBrowser = _f.isAvilableInBrowser;
2762
+ _loop_1(packageName, className, mimeTypes, documentationUrl, isAvilableInBrowser);
2763
+ }
2764
+ }
2765
+ catch (e_1_1) { e_1 = { error: e_1_1 }; }
2766
+ finally {
2767
+ try {
2768
+ if (_e && !_e.done && (_a = _d.return)) _a.call(_d);
2769
+ }
2770
+ finally { if (e_1) throw e_1.error; }
2771
+ }
2772
+ var _loop_2 = function (packageName, className, mimeTypes, documentationUrl, isAvilableInBrowser) {
2773
+ if (all.some(function (item) { return item.packageName === packageName && item.className === className; })) {
2774
+ return "continue";
2775
+ }
2776
+ all.push({ packageName: packageName, className: className, mimeTypes: mimeTypes, documentationUrl: documentationUrl, isAvilableInBrowser: isAvilableInBrowser });
2777
+ };
2778
+ try {
2779
+ for (var _g = __values($scrapersRegister.list()), _h = _g.next(); !_h.done; _h = _g.next()) {
2780
+ var _j = _h.value, packageName = _j.packageName, className = _j.className, mimeTypes = _j.mimeTypes, documentationUrl = _j.documentationUrl, isAvilableInBrowser = _j.isAvilableInBrowser;
2781
+ _loop_2(packageName, className, mimeTypes, documentationUrl, isAvilableInBrowser);
2782
+ }
2783
+ }
2784
+ catch (e_2_1) { e_2 = { error: e_2_1 }; }
2785
+ finally {
2786
+ try {
2787
+ if (_h && !_h.done && (_b = _g.return)) _b.call(_g);
2788
+ }
2789
+ finally { if (e_2) throw e_2.error; }
2790
+ }
2791
+ try {
2792
+ for (var availableScrapers_1 = __values(availableScrapers), availableScrapers_1_1 = availableScrapers_1.next(); !availableScrapers_1_1.done; availableScrapers_1_1 = availableScrapers_1.next()) {
2793
+ var metadata_1 = availableScrapers_1_1.value.metadata;
2794
+ all.push(metadata_1);
2795
+ }
2796
+ }
2797
+ catch (e_3_1) { e_3 = { error: e_3_1 }; }
2798
+ finally {
2799
+ try {
2800
+ if (availableScrapers_1_1 && !availableScrapers_1_1.done && (_c = availableScrapers_1.return)) _c.call(availableScrapers_1);
2801
+ }
2802
+ finally { if (e_3) throw e_3.error; }
2803
+ }
2804
+ var metadata = all.map(function (metadata) {
2805
+ var isMetadataAviailable = $scrapersMetadataRegister
2806
+ .list()
2807
+ .find(function (_a) {
2808
+ var packageName = _a.packageName, className = _a.className;
2809
+ return metadata.packageName === packageName && metadata.className === className;
2810
+ });
2811
+ var isInstalled = $scrapersRegister
2812
+ .list()
2813
+ .find(function (_a) {
2814
+ var packageName = _a.packageName, className = _a.className;
2815
+ return metadata.packageName === packageName && metadata.className === className;
2816
+ });
2817
+ var isAvilableInTools = availableScrapers.some(function (_a) {
2818
+ var _b = _a.metadata, packageName = _b.packageName, className = _b.className;
2819
+ return metadata.packageName === packageName && metadata.className === className;
2820
+ });
2821
+ return __assign(__assign({}, metadata), { isMetadataAviailable: isMetadataAviailable, isInstalled: isInstalled, isAvilableInTools: isAvilableInTools });
2822
+ });
2823
+ if (metadata.length === 0) {
2824
+ return spaceTrim__default["default"]("\n **No scrapers are available**\n\n This is a unexpected behavior, you are probably using some broken version of Promptbook\n At least there should be available the metadata of the scrapers\n ");
2825
+ }
2826
+ return spaceTrim__default["default"](function (block) { return "\n Available scrapers are:\n ".concat(block(metadata
2827
+ .map(function (_a, i) {
2828
+ var packageName = _a.packageName, className = _a.className, isMetadataAviailable = _a.isMetadataAviailable, isInstalled = _a.isInstalled, mimeTypes = _a.mimeTypes, isAvilableInBrowser = _a.isAvilableInBrowser, isAvilableInTools = _a.isAvilableInTools;
2829
+ var more = [];
2830
+ // TODO: [🧠] Maybe use `documentationUrl`
2831
+ if (isMetadataAviailable) {
2832
+ more.push("\u2B1C Metadata registered");
2833
+ } // not else
2834
+ if (isInstalled) {
2835
+ more.push("\uD83D\uDFE9 Installed");
2836
+ } // not else
2837
+ if (isAvilableInTools) {
2838
+ more.push("\uD83D\uDFE6 Available in tools");
2839
+ } // not else
2840
+ if (!isMetadataAviailable && isInstalled) {
2841
+ more.push("When no metadata registered but scraper is installed, it is an unexpected behavior");
2842
+ } // not else
2843
+ if (!isInstalled && isAvilableInTools) {
2844
+ more.push("When the scraper is not installed but available in tools, it is an unexpected compatibility behavior");
2845
+ } // not else
2846
+ if (!isAvilableInBrowser) {
2847
+ more.push("Not usable in browser");
2848
+ }
2849
+ var moreText = more.length === 0 ? '' : " *(".concat(more.join('; '), ")*");
2850
+ return "".concat(i + 1, ") `").concat(className, "` from `").concat(packageName, "` compatible to scrape ").concat(mimeTypes
2851
+ .map(function (mimeType) { return "\"".concat(mimeType, "\""); })
2852
+ .join(', ')).concat(moreText);
2853
+ })
2854
+ .join('\n')), "\n\n Legend:\n - \u2B1C **Metadata registered** means that Promptbook knows about the scraper, it is similar to registration in some registry\n - \uD83D\uDFE9 **Installed** means that you have imported package with particular scraper\n - \uD83D\uDFE6 **Available in tools** means that you have passed scraper as dependency into prepare or execution process\n\n "); });
2855
+ }
2856
+ /**
2857
+ * TODO: [®] DRY Register logic
2858
+ */
2859
+
2860
+ var defaultDiacriticsRemovalMap = [
2861
+ {
2862
+ base: 'A',
2863
+ letters: '\u0041\u24B6\uFF21\u00C0\u00C1\u00C2\u1EA6\u1EA4\u1EAA\u1EA8\u00C3\u0100\u0102\u1EB0\u1EAE\u1EB4\u1EB2\u0226\u01E0\u00C4\u01DE\u1EA2\u00C5\u01FA\u01CD\u0200\u0202\u1EA0\u1EAC\u1EB6\u1E00\u0104\u023A\u2C6F',
2864
+ },
2865
+ { base: 'AA', letters: '\uA732' },
2866
+ { base: 'AE', letters: '\u00C6\u01FC\u01E2' },
2867
+ { base: 'AO', letters: '\uA734' },
2868
+ { base: 'AU', letters: '\uA736' },
2869
+ { base: 'AV', letters: '\uA738\uA73A' },
2870
+ { base: 'AY', letters: '\uA73C' },
2871
+ {
2872
+ base: 'B',
2873
+ letters: '\u0042\u24B7\uFF22\u1E02\u1E04\u1E06\u0243\u0182\u0181',
2874
+ },
2875
+ {
2876
+ base: 'C',
2877
+ letters: '\u0043\u24B8\uFF23\u0106\u0108\u010A\u010C\u00C7\u1E08\u0187\u023B\uA73E',
2878
+ },
2879
+ {
2880
+ base: 'D',
2881
+ letters: '\u0044\u24B9\uFF24\u1E0A\u010E\u1E0C\u1E10\u1E12\u1E0E\u0110\u018B\u018A\u0189\uA779\u00D0',
2882
+ },
2883
+ { base: 'DZ', letters: '\u01F1\u01C4' },
2884
+ { base: 'Dz', letters: '\u01F2\u01C5' },
2885
+ {
2886
+ base: 'E',
2887
+ letters: '\u0045\u24BA\uFF25\u00C8\u00C9\u00CA\u1EC0\u1EBE\u1EC4\u1EC2\u1EBC\u0112\u1E14\u1E16\u0114\u0116\u00CB\u1EBA\u011A\u0204\u0206\u1EB8\u1EC6\u0228\u1E1C\u0118\u1E18\u1E1A\u0190\u018E',
2888
+ },
2889
+ { base: 'F', letters: '\u0046\u24BB\uFF26\u1E1E\u0191\uA77B' },
2890
+ {
2891
+ base: 'G',
2892
+ letters: '\u0047\u24BC\uFF27\u01F4\u011C\u1E20\u011E\u0120\u01E6\u0122\u01E4\u0193\uA7A0\uA77D\uA77E',
2893
+ },
2894
+ {
2895
+ base: 'H',
2896
+ letters: '\u0048\u24BD\uFF28\u0124\u1E22\u1E26\u021E\u1E24\u1E28\u1E2A\u0126\u2C67\u2C75\uA78D',
2897
+ },
2898
+ {
2899
+ base: 'I',
2900
+ letters: '\u0049\u24BE\uFF29\u00CC\u00CD\u00CE\u0128\u012A\u012C\u0130\u00CF\u1E2E\u1EC8\u01CF\u0208\u020A\u1ECA\u012E\u1E2C\u0197',
2901
+ },
2902
+ { base: 'J', letters: '\u004A\u24BF\uFF2A\u0134\u0248' },
2903
+ {
2904
+ base: 'K',
2905
+ letters: '\u004B\u24C0\uFF2B\u1E30\u01E8\u1E32\u0136\u1E34\u0198\u2C69\uA740\uA742\uA744\uA7A2',
2906
+ },
2907
+ {
2908
+ base: 'L',
2909
+ letters: '\u004C\u24C1\uFF2C\u013F\u0139\u013D\u1E36\u1E38\u013B\u1E3C\u1E3A\u0141\u023D\u2C62\u2C60\uA748\uA746\uA780',
2910
+ },
2911
+ { base: 'LJ', letters: '\u01C7' },
2912
+ { base: 'Lj', letters: '\u01C8' },
2913
+ { base: 'M', letters: '\u004D\u24C2\uFF2D\u1E3E\u1E40\u1E42\u2C6E\u019C' },
2914
+ {
2915
+ base: 'N',
2916
+ letters: '\u004E\u24C3\uFF2E\u01F8\u0143\u00D1\u1E44\u0147\u1E46\u0145\u1E4A\u1E48\u0220\u019D\uA790\uA7A4',
2917
+ },
2918
+ { base: 'NJ', letters: '\u01CA' },
2919
+ { base: 'Nj', letters: '\u01CB' },
2920
+ {
2921
+ base: 'O',
2922
+ letters: '\u004F\u24C4\uFF2F\u00D2\u00D3\u00D4\u1ED2\u1ED0\u1ED6\u1ED4\u00D5\u1E4C\u022C\u1E4E\u014C\u1E50\u1E52\u014E\u022E\u0230\u00D6\u022A\u1ECE\u0150\u01D1\u020C\u020E\u01A0\u1EDC\u1EDA\u1EE0\u1EDE\u1EE2\u1ECC\u1ED8\u01EA\u01EC\u00D8\u01FE\u0186\u019F\uA74A\uA74C',
2923
+ },
2924
+ { base: 'OI', letters: '\u01A2' },
2925
+ { base: 'OO', letters: '\uA74E' },
2926
+ { base: 'OU', letters: '\u0222' },
2927
+ { base: 'OE', letters: '\u008C\u0152' },
2928
+ { base: 'oe', letters: '\u009C\u0153' },
2929
+ {
2930
+ base: 'P',
2931
+ letters: '\u0050\u24C5\uFF30\u1E54\u1E56\u01A4\u2C63\uA750\uA752\uA754',
2932
+ },
2933
+ { base: 'Q', letters: '\u0051\u24C6\uFF31\uA756\uA758\u024A' },
2934
+ {
2935
+ base: 'R',
2936
+ letters: '\u0052\u24C7\uFF32\u0154\u1E58\u0158\u0210\u0212\u1E5A\u1E5C\u0156\u1E5E\u024C\u2C64\uA75A\uA7A6\uA782',
2937
+ },
2938
+ {
2939
+ base: 'S',
2940
+ letters: '\u0053\u24C8\uFF33\u1E9E\u015A\u1E64\u015C\u1E60\u0160\u1E66\u1E62\u1E68\u0218\u015E\u2C7E\uA7A8\uA784',
2941
+ },
2942
+ {
2943
+ base: 'T',
2944
+ letters: '\u0054\u24C9\uFF34\u1E6A\u0164\u1E6C\u021A\u0162\u1E70\u1E6E\u0166\u01AC\u01AE\u023E\uA786',
2945
+ },
2946
+ { base: 'TZ', letters: '\uA728' },
2947
+ {
2948
+ base: 'U',
2949
+ letters: '\u0055\u24CA\uFF35\u00D9\u00DA\u00DB\u0168\u1E78\u016A\u1E7A\u016C\u00DC\u01DB\u01D7\u01D5\u01D9\u1EE6\u016E\u0170\u01D3\u0214\u0216\u01AF\u1EEA\u1EE8\u1EEE\u1EEC\u1EF0\u1EE4\u1E72\u0172\u1E76\u1E74\u0244',
2950
+ },
2951
+ { base: 'V', letters: '\u0056\u24CB\uFF36\u1E7C\u1E7E\u01B2\uA75E\u0245' },
2952
+ { base: 'VY', letters: '\uA760' },
2953
+ {
2954
+ base: 'W',
2955
+ letters: '\u0057\u24CC\uFF37\u1E80\u1E82\u0174\u1E86\u1E84\u1E88\u2C72',
2956
+ },
2957
+ { base: 'X', letters: '\u0058\u24CD\uFF38\u1E8A\u1E8C' },
2958
+ {
2959
+ base: 'Y',
2960
+ letters: '\u0059\u24CE\uFF39\u1EF2\u00DD\u0176\u1EF8\u0232\u1E8E\u0178\u1EF6\u1EF4\u01B3\u024E\u1EFE',
2961
+ },
2962
+ {
2963
+ base: 'Z',
2964
+ letters: '\u005A\u24CF\uFF3A\u0179\u1E90\u017B\u017D\u1E92\u1E94\u01B5\u0224\u2C7F\u2C6B\uA762',
2965
+ },
2966
+ {
2967
+ base: 'a',
2968
+ letters: '\u0061\u24D0\uFF41\u1E9A\u00E0\u00E1\u00E2\u1EA7\u1EA5\u1EAB\u1EA9\u00E3\u0101\u0103\u1EB1\u1EAF\u1EB5\u1EB3\u0227\u01E1\u00E4\u01DF\u1EA3\u00E5\u01FB\u01CE\u0201\u0203\u1EA1\u1EAD\u1EB7\u1E01\u0105\u2C65\u0250',
2969
+ },
2970
+ { base: 'aa', letters: '\uA733' },
2971
+ { base: 'ae', letters: '\u00E6\u01FD\u01E3' },
2972
+ { base: 'ao', letters: '\uA735' },
2973
+ { base: 'au', letters: '\uA737' },
2974
+ { base: 'av', letters: '\uA739\uA73B' },
2975
+ { base: 'ay', letters: '\uA73D' },
2976
+ {
2977
+ base: 'b',
2978
+ letters: '\u0062\u24D1\uFF42\u1E03\u1E05\u1E07\u0180\u0183\u0253',
2979
+ },
2980
+ {
2981
+ base: 'c',
2982
+ letters: '\u0063\u24D2\uFF43\u0107\u0109\u010B\u010D\u00E7\u1E09\u0188\u023C\uA73F\u2184',
2983
+ },
2984
+ {
2985
+ base: 'd',
2986
+ letters: '\u0064\u24D3\uFF44\u1E0B\u010F\u1E0D\u1E11\u1E13\u1E0F\u0111\u018C\u0256\u0257\uA77A',
2987
+ },
2988
+ { base: 'dz', letters: '\u01F3\u01C6' },
2989
+ {
2990
+ base: 'e',
2991
+ letters: '\u0065\u24D4\uFF45\u00E8\u00E9\u00EA\u1EC1\u1EBF\u1EC5\u1EC3\u1EBD\u0113\u1E15\u1E17\u0115\u0117\u00EB\u1EBB\u011B\u0205\u0207\u1EB9\u1EC7\u0229\u1E1D\u0119\u1E19\u1E1B\u0247\u025B\u01DD',
2992
+ },
2993
+ { base: 'f', letters: '\u0066\u24D5\uFF46\u1E1F\u0192\uA77C' },
2994
+ {
2995
+ base: 'g',
2996
+ letters: '\u0067\u24D6\uFF47\u01F5\u011D\u1E21\u011F\u0121\u01E7\u0123\u01E5\u0260\uA7A1\u1D79\uA77F',
2997
+ },
2998
+ {
2999
+ base: 'h',
3000
+ letters: '\u0068\u24D7\uFF48\u0125\u1E23\u1E27\u021F\u1E25\u1E29\u1E2B\u1E96\u0127\u2C68\u2C76\u0265',
3001
+ },
3002
+ { base: 'hv', letters: '\u0195' },
3003
+ {
3004
+ base: 'i',
3005
+ letters: '\u0069\u24D8\uFF49\u00EC\u00ED\u00EE\u0129\u012B\u012D\u00EF\u1E2F\u1EC9\u01D0\u0209\u020B\u1ECB\u012F\u1E2D\u0268\u0131',
3006
+ },
3007
+ { base: 'j', letters: '\u006A\u24D9\uFF4A\u0135\u01F0\u0249' },
3008
+ {
3009
+ base: 'k',
3010
+ letters: '\u006B\u24DA\uFF4B\u1E31\u01E9\u1E33\u0137\u1E35\u0199\u2C6A\uA741\uA743\uA745\uA7A3',
3011
+ },
3012
+ {
3013
+ base: 'l',
3014
+ letters: '\u006C\u24DB\uFF4C\u0140\u013A\u013E\u1E37\u1E39\u013C\u1E3D\u1E3B\u017F\u0142\u019A\u026B\u2C61\uA749\uA781\uA747',
3015
+ },
3016
+ { base: 'lj', letters: '\u01C9' },
3017
+ { base: 'm', letters: '\u006D\u24DC\uFF4D\u1E3F\u1E41\u1E43\u0271\u026F' },
3018
+ {
3019
+ base: 'n',
3020
+ letters: '\u006E\u24DD\uFF4E\u01F9\u0144\u00F1\u1E45\u0148\u1E47\u0146\u1E4B\u1E49\u019E\u0272\u0149\uA791\uA7A5',
3021
+ },
3022
+ { base: 'nj', letters: '\u01CC' },
3023
+ {
3024
+ base: 'o',
3025
+ letters: '\u006F\u24DE\uFF4F\u00F2\u00F3\u00F4\u1ED3\u1ED1\u1ED7\u1ED5\u00F5\u1E4D\u022D\u1E4F\u014D\u1E51\u1E53\u014F\u022F\u0231\u00F6\u022B\u1ECF\u0151\u01D2\u020D\u020F\u01A1\u1EDD\u1EDB\u1EE1\u1EDF\u1EE3\u1ECD\u1ED9\u01EB\u01ED\u00F8\u01FF\u0254\uA74B\uA74D\u0275',
3026
+ },
3027
+ { base: 'oi', letters: '\u01A3' },
3028
+ { base: 'ou', letters: '\u0223' },
3029
+ { base: 'oo', letters: '\uA74F' },
3030
+ {
3031
+ base: 'p',
3032
+ letters: '\u0070\u24DF\uFF50\u1E55\u1E57\u01A5\u1D7D\uA751\uA753\uA755',
3033
+ },
3034
+ { base: 'q', letters: '\u0071\u24E0\uFF51\u024B\uA757\uA759' },
3035
+ {
3036
+ base: 'r',
3037
+ letters: '\u0072\u24E1\uFF52\u0155\u1E59\u0159\u0211\u0213\u1E5B\u1E5D\u0157\u1E5F\u024D\u027D\uA75B\uA7A7\uA783',
3038
+ },
3039
+ {
3040
+ base: 's',
3041
+ letters: '\u0073\u24E2\uFF53\u00DF\u015B\u1E65\u015D\u1E61\u0161\u1E67\u1E63\u1E69\u0219\u015F\u023F\uA7A9\uA785\u1E9B',
3042
+ },
3043
+ {
3044
+ base: 't',
3045
+ letters: '\u0074\u24E3\uFF54\u1E6B\u1E97\u0165\u1E6D\u021B\u0163\u1E71\u1E6F\u0167\u01AD\u0288\u2C66\uA787',
3046
+ },
3047
+ { base: 'tz', letters: '\uA729' },
3048
+ {
3049
+ base: 'u',
3050
+ letters: '\u0075\u24E4\uFF55\u00F9\u00FA\u00FB\u0169\u1E79\u016B\u1E7B\u016D\u00FC\u01DC\u01D8\u01D6\u01DA\u1EE7\u016F\u0171\u01D4\u0215\u0217\u01B0\u1EEB\u1EE9\u1EEF\u1EED\u1EF1\u1EE5\u1E73\u0173\u1E77\u1E75\u0289',
3051
+ },
3052
+ { base: 'v', letters: '\u0076\u24E5\uFF56\u1E7D\u1E7F\u028B\uA75F\u028C' },
3053
+ { base: 'vy', letters: '\uA761' },
3054
+ {
3055
+ base: 'w',
3056
+ letters: '\u0077\u24E6\uFF57\u1E81\u1E83\u0175\u1E87\u1E85\u1E98\u1E89\u2C73',
3057
+ },
3058
+ { base: 'x', letters: '\u0078\u24E7\uFF58\u1E8B\u1E8D' },
3059
+ {
3060
+ base: 'y',
3061
+ letters: '\u0079\u24E8\uFF59\u1EF3\u00FD\u0177\u1EF9\u0233\u1E8F\u00FF\u1EF7\u1E99\u1EF5\u01B4\u024F\u1EFF',
3062
+ },
3063
+ {
3064
+ base: 'z',
3065
+ letters: '\u007A\u24E9\uFF5A\u017A\u1E91\u017C\u017E\u1E93\u1E95\u01B6\u0225\u0240\u2C6C\uA763',
3066
+ },
3067
+ ];
3068
+ /**
3069
+ * Map of letters from diacritic variant to diacritless variant
3070
+ * Contains lowercase and uppercase separatelly
3071
+ *
3072
+ * > "á" => "a"
3073
+ * > "ě" => "e"
3074
+ * > "Ă" => "A"
3075
+ * > ...
3076
+ *
3077
+ * @public exported from `@promptbook/utils`
3078
+ */
3079
+ var DIACRITIC_VARIANTS_LETTERS = {};
3080
+ // tslint:disable-next-line: prefer-for-of
3081
+ for (var i = 0; i < defaultDiacriticsRemovalMap.length; i++) {
3082
+ var letters = defaultDiacriticsRemovalMap[i].letters;
3083
+ // tslint:disable-next-line: prefer-for-of
3084
+ for (var j = 0; j < letters.length; j++) {
3085
+ DIACRITIC_VARIANTS_LETTERS[letters[j]] = defaultDiacriticsRemovalMap[i].base;
3086
+ }
3087
+ }
3088
+ // <- TODO: [🍓] Put to maker function to save execution time if not needed
3089
+ /*
3090
+ @see https://stackoverflow.com/questions/990904/remove-accents-diacritics-in-a-string-in-javascript
3091
+ Licensed under the Apache License, Version 2.0 (the "License");
3092
+ you may not use this file except in compliance with the License.
3093
+ You may obtain a copy of the License at
3094
+
3095
+ http://www.apache.org/licenses/LICENSE-2.0
3096
+
3097
+ Unless required by applicable law or agreed to in writing, software
3098
+ distributed under the License is distributed on an "AS IS" BASIS,
3099
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
3100
+ See the License for the specific language governing permissions and
3101
+ limitations under the License.
3102
+ */
3116
3103
 
3117
3104
  /**
3118
3105
  * @@@
3119
3106
  *
3120
- * Note: `$` is used to indicate that this interacts with the global scope
3121
- * @singleton Only one instance of each register is created per build, but thare can be more @@@
3122
- * @public exported from `@promptbook/core`
3107
+ * @param input @@@
3108
+ * @returns @@@
3109
+ * @public exported from `@promptbook/utils`
3123
3110
  */
3124
- var $scrapersRegister = new $Register('scraper_constructors');
3111
+ function removeDiacritics(input) {
3112
+ /*eslint no-control-regex: "off"*/
3113
+ return input.replace(/[^\u0000-\u007E]/g, function (a) {
3114
+ return DIACRITIC_VARIANTS_LETTERS[a] || a;
3115
+ });
3116
+ }
3125
3117
  /**
3126
- * TODO: [®] DRY Register logic
3118
+ * TODO: [Ж] Variant for cyrillic (and in general non-latin) letters
3127
3119
  */
3128
3120
 
3129
3121
  /**
3130
- * Creates a message with all registered scrapers
3131
- *
3132
- * Note: This function is used to create a (error) message when there is no scraper for particular mime type
3122
+ * @@@
3133
3123
  *
3134
- * @private internal function of `createScrapersFromConfiguration` and `createScrapersFromEnv`
3124
+ * @param text @@@
3125
+ * @returns @@@
3126
+ * @example 'hello-world'
3127
+ * @example 'i-love-promptbook'
3128
+ * @public exported from `@promptbook/utils`
3135
3129
  */
3136
- function $registeredScrapersMessage(availableScrapers) {
3137
- var e_1, _a, e_2, _b, e_3, _c;
3138
- /**
3139
- * Mixes registered scrapers from $scrapersMetadataRegister and $scrapersRegister
3140
- */
3141
- var all = [];
3142
- var _loop_1 = function (packageName, className, mimeTypes, documentationUrl, isAvilableInBrowser) {
3143
- if (all.some(function (item) { return item.packageName === packageName && item.className === className; })) {
3144
- return "continue";
3145
- }
3146
- all.push({ packageName: packageName, className: className, mimeTypes: mimeTypes, documentationUrl: documentationUrl, isAvilableInBrowser: isAvilableInBrowser });
3147
- };
3130
+ function normalizeToKebabCase(text) {
3131
+ var e_1, _a;
3132
+ text = removeDiacritics(text);
3133
+ var charType;
3134
+ var lastCharType = 'OTHER';
3135
+ var normalizedName = '';
3148
3136
  try {
3149
- for (var _d = __values($scrapersMetadataRegister.list()), _e = _d.next(); !_e.done; _e = _d.next()) {
3150
- var _f = _e.value, packageName = _f.packageName, className = _f.className, mimeTypes = _f.mimeTypes, documentationUrl = _f.documentationUrl, isAvilableInBrowser = _f.isAvilableInBrowser;
3151
- _loop_1(packageName, className, mimeTypes, documentationUrl, isAvilableInBrowser);
3137
+ for (var text_1 = __values(text), text_1_1 = text_1.next(); !text_1_1.done; text_1_1 = text_1.next()) {
3138
+ var char = text_1_1.value;
3139
+ var normalizedChar = void 0;
3140
+ if (/^[a-z]$/.test(char)) {
3141
+ charType = 'LOWERCASE';
3142
+ normalizedChar = char;
3143
+ }
3144
+ else if (/^[A-Z]$/.test(char)) {
3145
+ charType = 'UPPERCASE';
3146
+ normalizedChar = char.toLowerCase();
3147
+ }
3148
+ else if (/^[0-9]$/.test(char)) {
3149
+ charType = 'NUMBER';
3150
+ normalizedChar = char;
3151
+ }
3152
+ else {
3153
+ charType = 'OTHER';
3154
+ normalizedChar = '-';
3155
+ }
3156
+ if (charType !== lastCharType &&
3157
+ !(lastCharType === 'UPPERCASE' && charType === 'LOWERCASE') &&
3158
+ !(lastCharType === 'NUMBER') &&
3159
+ !(charType === 'NUMBER')) {
3160
+ normalizedName += '-';
3161
+ }
3162
+ normalizedName += normalizedChar;
3163
+ lastCharType = charType;
3152
3164
  }
3153
3165
  }
3154
3166
  catch (e_1_1) { e_1 = { error: e_1_1 }; }
3155
3167
  finally {
3156
3168
  try {
3157
- if (_e && !_e.done && (_a = _d.return)) _a.call(_d);
3169
+ if (text_1_1 && !text_1_1.done && (_a = text_1.return)) _a.call(text_1);
3158
3170
  }
3159
3171
  finally { if (e_1) throw e_1.error; }
3160
3172
  }
3161
- var _loop_2 = function (packageName, className, mimeTypes, documentationUrl, isAvilableInBrowser) {
3162
- if (all.some(function (item) { return item.packageName === packageName && item.className === className; })) {
3163
- return "continue";
3164
- }
3165
- all.push({ packageName: packageName, className: className, mimeTypes: mimeTypes, documentationUrl: documentationUrl, isAvilableInBrowser: isAvilableInBrowser });
3166
- };
3167
- try {
3168
- for (var _g = __values($scrapersRegister.list()), _h = _g.next(); !_h.done; _h = _g.next()) {
3169
- var _j = _h.value, packageName = _j.packageName, className = _j.className, mimeTypes = _j.mimeTypes, documentationUrl = _j.documentationUrl, isAvilableInBrowser = _j.isAvilableInBrowser;
3170
- _loop_2(packageName, className, mimeTypes, documentationUrl, isAvilableInBrowser);
3171
- }
3172
- }
3173
- catch (e_2_1) { e_2 = { error: e_2_1 }; }
3174
- finally {
3175
- try {
3176
- if (_h && !_h.done && (_b = _g.return)) _b.call(_g);
3177
- }
3178
- finally { if (e_2) throw e_2.error; }
3179
- }
3180
- try {
3181
- for (var availableScrapers_1 = __values(availableScrapers), availableScrapers_1_1 = availableScrapers_1.next(); !availableScrapers_1_1.done; availableScrapers_1_1 = availableScrapers_1.next()) {
3182
- var metadata_1 = availableScrapers_1_1.value.metadata;
3183
- all.push(metadata_1);
3184
- }
3185
- }
3186
- catch (e_3_1) { e_3 = { error: e_3_1 }; }
3187
- finally {
3188
- try {
3189
- if (availableScrapers_1_1 && !availableScrapers_1_1.done && (_c = availableScrapers_1.return)) _c.call(availableScrapers_1);
3190
- }
3191
- finally { if (e_3) throw e_3.error; }
3192
- }
3193
- var metadata = all.map(function (metadata) {
3194
- var isMetadataAviailable = $scrapersMetadataRegister
3195
- .list()
3196
- .find(function (_a) {
3197
- var packageName = _a.packageName, className = _a.className;
3198
- return metadata.packageName === packageName && metadata.className === className;
3199
- });
3200
- var isInstalled = $scrapersRegister
3201
- .list()
3202
- .find(function (_a) {
3203
- var packageName = _a.packageName, className = _a.className;
3204
- return metadata.packageName === packageName && metadata.className === className;
3205
- });
3206
- var isAvilableInTools = availableScrapers.some(function (_a) {
3207
- var _b = _a.metadata, packageName = _b.packageName, className = _b.className;
3208
- return metadata.packageName === packageName && metadata.className === className;
3209
- });
3210
- return __assign(__assign({}, metadata), { isMetadataAviailable: isMetadataAviailable, isInstalled: isInstalled, isAvilableInTools: isAvilableInTools });
3211
- });
3212
- if (metadata.length === 0) {
3213
- return spaceTrim__default["default"]("\n **No scrapers are available**\n\n This is a unexpected behavior, you are probably using some broken version of Promptbook\n At least there should be available the metadata of the scrapers\n ");
3214
- }
3215
- return spaceTrim__default["default"](function (block) { return "\n Available scrapers are:\n ".concat(block(metadata
3216
- .map(function (_a, i) {
3217
- var packageName = _a.packageName, className = _a.className, isMetadataAviailable = _a.isMetadataAviailable, isInstalled = _a.isInstalled, mimeTypes = _a.mimeTypes, isAvilableInBrowser = _a.isAvilableInBrowser, isAvilableInTools = _a.isAvilableInTools;
3218
- var more = [];
3219
- // TODO: [🧠] Maybe use `documentationUrl`
3220
- if (isMetadataAviailable) {
3221
- more.push("\u2B1C Metadata registered");
3222
- } // not else
3223
- if (isInstalled) {
3224
- more.push("\uD83D\uDFE9 Installed");
3225
- } // not else
3226
- if (isAvilableInTools) {
3227
- more.push("\uD83D\uDFE6 Available in tools");
3228
- } // not else
3229
- if (!isMetadataAviailable && isInstalled) {
3230
- more.push("When no metadata registered but scraper is installed, it is an unexpected behavior");
3231
- } // not else
3232
- if (!isInstalled && isAvilableInTools) {
3233
- more.push("When the scraper is not installed but available in tools, it is an unexpected compatibility behavior");
3234
- } // not else
3235
- if (!isAvilableInBrowser) {
3236
- more.push("Not usable in browser");
3237
- }
3238
- var moreText = more.length === 0 ? '' : " *(".concat(more.join('; '), ")*");
3239
- return "".concat(i + 1, ") `").concat(className, "` from `").concat(packageName, "` compatible to scrape ").concat(mimeTypes
3240
- .map(function (mimeType) { return "\"".concat(mimeType, "\""); })
3241
- .join(', ')).concat(moreText);
3242
- })
3243
- .join('\n')), "\n\n Legend:\n - \u2B1C **Metadata registered** means that Promptbook knows about the scraper, it is similar to registration in some registry\n - \uD83D\uDFE9 **Installed** means that you have imported package with particular scraper\n - \uD83D\uDFE6 **Available in tools** means that you have passed scraper as dependency into prepare or execution process\n\n "); });
3173
+ normalizedName = normalizedName.split(/-+/g).join('-');
3174
+ normalizedName = normalizedName.split(/-?\/-?/g).join('/');
3175
+ normalizedName = normalizedName.replace(/^-/, '');
3176
+ normalizedName = normalizedName.replace(/-$/, '');
3177
+ return normalizedName;
3244
3178
  }
3245
3179
  /**
3246
- * TODO: [®] DRY Register logic
3180
+ * Note: [💞] Ignore a discrepancy between file name and entity name
3247
3181
  */
3248
3182
 
3249
3183
  /**
@@ -3322,6 +3256,32 @@
3322
3256
  * TODO: [🖇] What about symlinks?
3323
3257
  */
3324
3258
 
3259
+ /**
3260
+ * Tests if given string is valid URL.
3261
+ *
3262
+ * Note: This does not check if the file exists only if the path is valid
3263
+ * @public exported from `@promptbook/utils`
3264
+ */
3265
+ function isValidFilePath(filename) {
3266
+ if (typeof filename !== 'string') {
3267
+ return false;
3268
+ }
3269
+ var filenameSlashes = filename.split('\\').join('/');
3270
+ // Absolute Unix path: /hello.txt
3271
+ if (/^(\/)/i.test(filenameSlashes)) {
3272
+ return true;
3273
+ }
3274
+ // Absolute Windows path: /hello.txt
3275
+ if (/^([A-Z]{1,2}:\/?)\//i.test(filenameSlashes)) {
3276
+ return true;
3277
+ }
3278
+ // Relative path: ./hello.txt
3279
+ if (/^(\.\.?\/)+/i.test(filenameSlashes)) {
3280
+ return true;
3281
+ }
3282
+ return false;
3283
+ }
3284
+
3325
3285
  /**
3326
3286
  * @@@
3327
3287
  *
@@ -3783,9 +3743,9 @@
3783
3743
  * @param script from which to extract the variables
3784
3744
  * @returns the list of variable names
3785
3745
  * @throws {ParseError} if the script is invalid
3786
- * @public exported from `@promptbook/utils`
3746
+ * @public exported from `@promptbook/utils` <- Note: [👖] This is usable elsewhere than in Promptbook, so keeping in utils
3787
3747
  */
3788
- function extractVariables(script) {
3748
+ function extractVariablesFromScript(script) {
3789
3749
  var variables = new Set();
3790
3750
  script = "(()=>{".concat(script, "})()");
3791
3751
  try {
@@ -3832,7 +3792,7 @@
3832
3792
  * @param task the task with used parameters
3833
3793
  * @returns the set of parameter names
3834
3794
  * @throws {ParseError} if the script is invalid
3835
- * @public exported from `@promptbook/utils`
3795
+ * @public exported from `@promptbook/core` <- Note: [👖] This utility is so tightly interconnected with the Promptbook that it is not exported as util but in core
3836
3796
  */
3837
3797
  function extractParameterNamesFromTask(task) {
3838
3798
  var e_1, _a, e_2, _b, e_3, _c, e_4, _d;
@@ -3853,7 +3813,7 @@
3853
3813
  }
3854
3814
  if (taskType === 'SCRIPT_TASK') {
3855
3815
  try {
3856
- for (var _g = __values(extractVariables(content)), _h = _g.next(); !_h.done; _h = _g.next()) {
3816
+ for (var _g = __values(extractVariablesFromScript(content)), _h = _g.next(); !_h.done; _h = _g.next()) {
3857
3817
  var parameterName = _h.value;
3858
3818
  parameterNames.add(parameterName);
3859
3819
  }
@@ -5625,6 +5585,46 @@
5625
5585
  * TODO: [🐚] Change onProgress to object that represents the running execution, can be subscribed via RxJS to and also awaited
5626
5586
  */
5627
5587
 
5588
+ /**
5589
+ * Removes emojis from a string and fix whitespaces
5590
+ *
5591
+ * @param text with emojis
5592
+ * @returns text without emojis
5593
+ * @public exported from `@promptbook/utils`
5594
+ */
5595
+ function removeEmojis(text) {
5596
+ // Replace emojis (and also ZWJ sequence) with hyphens
5597
+ text = text.replace(/(\p{Extended_Pictographic})\p{Modifier_Symbol}/gu, '$1');
5598
+ text = text.replace(/(\p{Extended_Pictographic})[\u{FE00}-\u{FE0F}]/gu, '$1');
5599
+ text = text.replace(/(\p{Extended_Pictographic})(\u{200D}\p{Extended_Pictographic})*/gu, '$1');
5600
+ text = text.replace(/\p{Extended_Pictographic}/gu, '');
5601
+ return text;
5602
+ }
5603
+
5604
+ /**
5605
+ * @@@
5606
+ *
5607
+ * @param value @@@
5608
+ * @returns @@@
5609
+ * @example @@@
5610
+ * @public exported from `@promptbook/utils`
5611
+ */
5612
+ function titleToName(value) {
5613
+ if (isValidUrl(value)) {
5614
+ value = value.replace(/^https?:\/\//, '');
5615
+ value = value.replace(/\.html$/, '');
5616
+ }
5617
+ else if (isValidFilePath(value)) {
5618
+ value = path.basename(value);
5619
+ // Note: Keeping extension in the name
5620
+ }
5621
+ value = value.split('/').join('-');
5622
+ value = removeEmojis(value);
5623
+ value = normalizeToKebabCase(value);
5624
+ // TODO: [🧠] Maybe warn or add some padding to short name which are not good identifiers
5625
+ return value;
5626
+ }
5627
+
5628
5628
  /**
5629
5629
  * Metadata of the scraper
5630
5630
  *
@@ -6062,6 +6062,7 @@
6062
6062
  var e_1, _a;
6063
6063
  var lines = markdown.split('\n');
6064
6064
  var sections = [];
6065
+ // TODO: [🧽] DRY
6065
6066
  var currentType = 'MARKDOWN';
6066
6067
  var buffer = [];
6067
6068
  var finishSection = function () {