/testing/selenium-core/xpath/xmltoken.js
JavaScript | 149 lines | 126 code | 9 blank | 14 comment | 3 complexity | dbdd171a5232832765efbe32ec34dbd7 MD5 | raw file
Possible License(s): Apache-2.0
1// Copyright 2006 Google Inc. 2// All Rights Reserved 3// 4// Defines regular expression patterns to extract XML tokens from string. 5// See <http://www.w3.org/TR/REC-xml/#sec-common-syn>, 6// <http://www.w3.org/TR/xml11/#sec-common-syn> and 7// <http://www.w3.org/TR/REC-xml-names/#NT-NCName> for the specifications. 8// 9// Author: Junji Takagi <jtakagi@google.com> 10 11// Detect whether RegExp supports Unicode characters or not. 12 13var REGEXP_UNICODE = function() { 14 var tests = [' ', '\u0120', -1, // Konquerer 3.4.0 fails here. 15 '!', '\u0120', -1, 16 '\u0120', '\u0120', 0, 17 '\u0121', '\u0120', -1, 18 '\u0121', '\u0120|\u0121', 0, 19 '\u0122', '\u0120|\u0121', -1, 20 '\u0120', '[\u0120]', 0, // Safari 2.0.3 fails here. 21 '\u0121', '[\u0120]', -1, 22 '\u0121', '[\u0120\u0121]', 0, // Safari 2.0.3 fails here. 23 '\u0122', '[\u0120\u0121]', -1, 24 '\u0121', '[\u0120-\u0121]', 0, // Safari 2.0.3 fails here. 25 '\u0122', '[\u0120-\u0121]', -1]; 26 for (var i = 0; i < tests.length; i += 3) { 27 if (tests[i].search(new RegExp(tests[i + 1])) != tests[i + 2]) { 28 return false; 29 } 30 } 31 return true; 32}(); 33 34// Common tokens in XML 1.0 and XML 1.1. 35 36var XML_S = '[ \t\r\n]+'; 37var XML_EQ = '(' + XML_S + ')?=(' + XML_S + ')?'; 38var XML_CHAR_REF = '&#[0-9]+;|&#x[0-9a-fA-F]+;'; 39 40// XML 1.0 tokens. 41 42var XML10_VERSION_INFO = XML_S + 'version' + XML_EQ + '("1\\.0"|' + "'1\\.0')"; 43var XML10_BASE_CHAR = (REGEXP_UNICODE) ? 44 '\u0041-\u005a\u0061-\u007a\u00c0-\u00d6\u00d8-\u00f6\u00f8-\u00ff' + 45 '\u0100-\u0131\u0134-\u013e\u0141-\u0148\u014a-\u017e\u0180-\u01c3' + 46 '\u01cd-\u01f0\u01f4-\u01f5\u01fa-\u0217\u0250-\u02a8\u02bb-\u02c1\u0386' + 47 '\u0388-\u038a\u038c\u038e-\u03a1\u03a3-\u03ce\u03d0-\u03d6\u03da\u03dc' + 48 '\u03de\u03e0\u03e2-\u03f3\u0401-\u040c\u040e-\u044f\u0451-\u045c' + 49 '\u045e-\u0481\u0490-\u04c4\u04c7-\u04c8\u04cb-\u04cc\u04d0-\u04eb' + 50 '\u04ee-\u04f5\u04f8-\u04f9\u0531-\u0556\u0559\u0561-\u0586\u05d0-\u05ea' + 51 '\u05f0-\u05f2\u0621-\u063a\u0641-\u064a\u0671-\u06b7\u06ba-\u06be' + 52 '\u06c0-\u06ce\u06d0-\u06d3\u06d5\u06e5-\u06e6\u0905-\u0939\u093d' + 53 '\u0958-\u0961\u0985-\u098c\u098f-\u0990\u0993-\u09a8\u09aa-\u09b0\u09b2' + 54 '\u09b6-\u09b9\u09dc-\u09dd\u09df-\u09e1\u09f0-\u09f1\u0a05-\u0a0a' + 55 '\u0a0f-\u0a10\u0a13-\u0a28\u0a2a-\u0a30\u0a32-\u0a33\u0a35-\u0a36' + 56 '\u0a38-\u0a39\u0a59-\u0a5c\u0a5e\u0a72-\u0a74\u0a85-\u0a8b\u0a8d' + 57 '\u0a8f-\u0a91\u0a93-\u0aa8\u0aaa-\u0ab0\u0ab2-\u0ab3\u0ab5-\u0ab9' + 58 '\u0abd\u0ae0\u0b05-\u0b0c\u0b0f-\u0b10\u0b13-\u0b28\u0b2a-\u0b30' + 59 '\u0b32-\u0b33\u0b36-\u0b39\u0b3d\u0b5c-\u0b5d\u0b5f-\u0b61\u0b85-\u0b8a' + 60 '\u0b8e-\u0b90\u0b92-\u0b95\u0b99-\u0b9a\u0b9c\u0b9e-\u0b9f\u0ba3-\u0ba4' + 61 '\u0ba8-\u0baa\u0bae-\u0bb5\u0bb7-\u0bb9\u0c05-\u0c0c\u0c0e-\u0c10' + 62 '\u0c12-\u0c28\u0c2a-\u0c33\u0c35-\u0c39\u0c60-\u0c61\u0c85-\u0c8c' + 63 '\u0c8e-\u0c90\u0c92-\u0ca8\u0caa-\u0cb3\u0cb5-\u0cb9\u0cde\u0ce0-\u0ce1' + 64 '\u0d05-\u0d0c\u0d0e-\u0d10\u0d12-\u0d28\u0d2a-\u0d39\u0d60-\u0d61' + 65 '\u0e01-\u0e2e\u0e30\u0e32-\u0e33\u0e40-\u0e45\u0e81-\u0e82\u0e84' + 66 '\u0e87-\u0e88\u0e8a\u0e8d\u0e94-\u0e97\u0e99-\u0e9f\u0ea1-\u0ea3\u0ea5' + 67 '\u0ea7\u0eaa-\u0eab\u0ead-\u0eae\u0eb0\u0eb2-\u0eb3\u0ebd\u0ec0-\u0ec4' + 68 '\u0f40-\u0f47\u0f49-\u0f69\u10a0-\u10c5\u10d0-\u10f6\u1100\u1102-\u1103' + 69 '\u1105-\u1107\u1109\u110b-\u110c\u110e-\u1112\u113c\u113e\u1140\u114c' + 70 '\u114e\u1150\u1154-\u1155\u1159\u115f-\u1161\u1163\u1165\u1167\u1169' + 71 '\u116d-\u116e\u1172-\u1173\u1175\u119e\u11a8\u11ab\u11ae-\u11af' + 72 '\u11b7-\u11b8\u11ba\u11bc-\u11c2\u11eb\u11f0\u11f9\u1e00-\u1e9b' + 73 '\u1ea0-\u1ef9\u1f00-\u1f15\u1f18-\u1f1d\u1f20-\u1f45\u1f48-\u1f4d' + 74 '\u1f50-\u1f57\u1f59\u1f5b\u1f5d\u1f5f-\u1f7d\u1f80-\u1fb4\u1fb6-\u1fbc' + 75 '\u1fbe\u1fc2-\u1fc4\u1fc6-\u1fcc\u1fd0-\u1fd3\u1fd6-\u1fdb\u1fe0-\u1fec' + 76 '\u1ff2-\u1ff4\u1ff6-\u1ffc\u2126\u212a-\u212b\u212e\u2180-\u2182' + 77 '\u3041-\u3094\u30a1-\u30fa\u3105-\u312c\uac00-\ud7a3' : 78 'A-Za-z'; 79var XML10_IDEOGRAPHIC = (REGEXP_UNICODE) ? 80 '\u4e00-\u9fa5\u3007\u3021-\u3029' : 81 ''; 82var XML10_COMBINING_CHAR = (REGEXP_UNICODE) ? 83 '\u0300-\u0345\u0360-\u0361\u0483-\u0486\u0591-\u05a1\u05a3-\u05b9' + 84 '\u05bb-\u05bd\u05bf\u05c1-\u05c2\u05c4\u064b-\u0652\u0670\u06d6-\u06dc' + 85 '\u06dd-\u06df\u06e0-\u06e4\u06e7-\u06e8\u06ea-\u06ed\u0901-\u0903\u093c' + 86 '\u093e-\u094c\u094d\u0951-\u0954\u0962-\u0963\u0981-\u0983\u09bc\u09be' + 87 '\u09bf\u09c0-\u09c4\u09c7-\u09c8\u09cb-\u09cd\u09d7\u09e2-\u09e3\u0a02' + 88 '\u0a3c\u0a3e\u0a3f\u0a40-\u0a42\u0a47-\u0a48\u0a4b-\u0a4d\u0a70-\u0a71' + 89 '\u0a81-\u0a83\u0abc\u0abe-\u0ac5\u0ac7-\u0ac9\u0acb-\u0acd\u0b01-\u0b03' + 90 '\u0b3c\u0b3e-\u0b43\u0b47-\u0b48\u0b4b-\u0b4d\u0b56-\u0b57\u0b82-\u0b83' + 91 '\u0bbe-\u0bc2\u0bc6-\u0bc8\u0bca-\u0bcd\u0bd7\u0c01-\u0c03\u0c3e-\u0c44' + 92 '\u0c46-\u0c48\u0c4a-\u0c4d\u0c55-\u0c56\u0c82-\u0c83\u0cbe-\u0cc4' + 93 '\u0cc6-\u0cc8\u0cca-\u0ccd\u0cd5-\u0cd6\u0d02-\u0d03\u0d3e-\u0d43' + 94 '\u0d46-\u0d48\u0d4a-\u0d4d\u0d57\u0e31\u0e34-\u0e3a\u0e47-\u0e4e\u0eb1' + 95 '\u0eb4-\u0eb9\u0ebb-\u0ebc\u0ec8-\u0ecd\u0f18-\u0f19\u0f35\u0f37\u0f39' + 96 '\u0f3e\u0f3f\u0f71-\u0f84\u0f86-\u0f8b\u0f90-\u0f95\u0f97\u0f99-\u0fad' + 97 '\u0fb1-\u0fb7\u0fb9\u20d0-\u20dc\u20e1\u302a-\u302f\u3099\u309a' : 98 ''; 99var XML10_DIGIT = (REGEXP_UNICODE) ? 100 '\u0030-\u0039\u0660-\u0669\u06f0-\u06f9\u0966-\u096f\u09e6-\u09ef' + 101 '\u0a66-\u0a6f\u0ae6-\u0aef\u0b66-\u0b6f\u0be7-\u0bef\u0c66-\u0c6f' + 102 '\u0ce6-\u0cef\u0d66-\u0d6f\u0e50-\u0e59\u0ed0-\u0ed9\u0f20-\u0f29' : 103 '0-9'; 104var XML10_EXTENDER = (REGEXP_UNICODE) ? 105 '\u00b7\u02d0\u02d1\u0387\u0640\u0e46\u0ec6\u3005\u3031-\u3035' + 106 '\u309d-\u309e\u30fc-\u30fe' : 107 ''; 108var XML10_LETTER = XML10_BASE_CHAR + XML10_IDEOGRAPHIC; 109var XML10_NAME_CHAR = XML10_LETTER + XML10_DIGIT + '\\._:' + 110 XML10_COMBINING_CHAR + XML10_EXTENDER + '-'; 111var XML10_NAME = '[' + XML10_LETTER + '_:][' + XML10_NAME_CHAR + ']*'; 112 113var XML10_ENTITY_REF = '&' + XML10_NAME + ';'; 114var XML10_REFERENCE = XML10_ENTITY_REF + '|' + XML_CHAR_REF; 115var XML10_ATT_VALUE = '"(([^<&"]|' + XML10_REFERENCE + ')*)"|' + 116 "'(([^<&']|" + XML10_REFERENCE + ")*)'"; 117var XML10_ATTRIBUTE = 118 '(' + XML10_NAME + ')' + XML_EQ + '(' + XML10_ATT_VALUE + ')'; 119 120// XML 1.1 tokens. 121// TODO(jtakagi): NameStartChar also includes \u10000-\ueffff. 122// ECMAScript Language Specifiction defines UnicodeEscapeSequence as 123// "\u HexDigit HexDigit HexDigit HexDigit" and we may need to use 124// surrogate pairs, but any browser doesn't support surrogate paris in 125// character classes of regular expression, so avoid including them for now. 126 127var XML11_VERSION_INFO = XML_S + 'version' + XML_EQ + '("1\\.1"|' + "'1\\.1')"; 128var XML11_NAME_START_CHAR = (REGEXP_UNICODE) ? 129 ':A-Z_a-z\u00c0-\u00d6\u00d8-\u00f6\u00f8-\u02ff\u0370-\u037d' + 130 '\u037f-\u1fff\u200c-\u200d\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff' + 131 '\uf900-\ufdcf\ufdf0-\ufffd' : 132 ':A-Z_a-z'; 133var XML11_NAME_CHAR = XML11_NAME_START_CHAR + 134 ((REGEXP_UNICODE) ? '\\.0-9\u00b7\u0300-\u036f\u203f-\u2040-' : '\\.0-9-'); 135var XML11_NAME = '[' + XML11_NAME_START_CHAR + '][' + XML11_NAME_CHAR + ']*'; 136 137var XML11_ENTITY_REF = '&' + XML11_NAME + ';'; 138var XML11_REFERENCE = XML11_ENTITY_REF + '|' + XML_CHAR_REF; 139var XML11_ATT_VALUE = '"(([^<&"]|' + XML11_REFERENCE + ')*)"|' + 140 "'(([^<&']|" + XML11_REFERENCE + ")*)'"; 141var XML11_ATTRIBUTE = 142 '(' + XML11_NAME + ')' + XML_EQ + '(' + XML11_ATT_VALUE + ')'; 143 144// XML Namespace tokens. 145// Used in XML parser and XPath parser. 146 147var XML_NC_NAME_CHAR = XML10_LETTER + XML10_DIGIT + '\\._' + 148 XML10_COMBINING_CHAR + XML10_EXTENDER + '-'; 149var XML_NC_NAME = '[' + XML10_LETTER + '_][' + XML_NC_NAME_CHAR + ']*';