Created
November 14, 2012 11:58
-
-
Save devote/4071739 to your computer and use it in GitHub Desktop.
PCRE Converter
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
<?php | |
/** | |
* Конвертер регулярного выражения | |
* | |
* @author Pakhtinov Dmitriy | |
* @copyright 2012 Pakhtinov Dmitriy | |
* @data 14.11.2012 | |
*/ | |
class PCREConverter { | |
private static $pcre_tag_type = 0; | |
private static $handlers = array( | |
'\\X' => 'upper_X_Handler', | |
'\\P' => 'upper_lower_P_Handler', | |
'\\p' => 'upper_lower_P_Handler', | |
); | |
public static $unicode_blocks = array( | |
'InBasic_Latin' => '\u0000-\u007F', | |
'InLatin-1_Supplement' => '\u0080-\u00FF', | |
'InLatin_Extended-A' => '\u0100-\u017F', | |
'InLatin_Extended-B' => '\u0180-\u024F', | |
'InIPA_Extensions' => '\u0250-\u02AF', | |
'InSpacing_Modifier_Letters' => '\u02B0-\u02FF', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036F', | |
'InGreek_and_Coptic' => '\u0370-\u03FF', | |
'InCyrillic' => '\u0400-\u04FF', | |
'InCyrillic_Supplementary' => '\u0500-\u052F', | |
'InArmenian' => '\u0530-\u058F', | |
'InHebrew' => '\u0590-\u05FF', | |
'InArabic' => '\u0600-\u06FF', | |
'InSyriac' => '\u0700-\u074F', | |
'InThaana' => '\u0780-\u07BF', | |
'InDevanagari' => '\u0900-\u097F', | |
'InBengali' => '\u0980-\u09FF', | |
'InGurmukhi' => '\u0A00-\u0A7F', | |
'InGujarati' => '\u0A80-\u0AFF', | |
'InOriya' => '\u0B00-\u0B7F', | |
'InTamil' => '\u0B80-\u0BFF', | |
'InTelugu' => '\u0C00-\u0C7F', | |
'InKannada' => '\u0C80-\u0CFF', | |
'InMalayalam' => '\u0D00-\u0D7F', | |
'InSinhala' => '\u0D80-\u0DFF', | |
'InThai' => '\u0E00-\u0E7F', | |
'InLao' => '\u0E80-\u0EFF', | |
'InTibetan' => '\u0F00-\u0FFF', | |
'InMyanmar' => '\u1000-\u109F', | |
'InGeorgian' => '\u10A0-\u10FF', | |
'InHangul_Jamo' => '\u1100-\u11FF', | |
'InEthiopic' => '\u1200-\u137F', | |
'InCherokee' => '\u13A0-\u13FF', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167F', | |
'InOgham' => '\u1680-\u169F', | |
'InRunic' => '\u16A0-\u16FF', | |
'InTagalog' => '\u1700-\u171F', | |
'InHanunoo' => '\u1720-\u173F', | |
'InBuhid' => '\u1740-\u175F', | |
'InTagbanwa' => '\u1760-\u177F', | |
'InKhmer' => '\u1780-\u17FF', | |
'InMongolian' => '\u1800-\u18AF', | |
'InLimbu' => '\u1900-\u194F', | |
'InTai_Le' => '\u1950-\u197F', | |
'InKhmer_Symbols' => '\u19E0-\u19FF', | |
'InPhonetic_Extensions' => '\u1D00-\u1D7F', | |
'InLatin_Extended_Additional' => '\u1E00-\u1EFF', | |
'InGreek_Extended' => '\u1F00-\u1FFF', | |
'InGeneral_Punctuation' => '\u2000-\u206F', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209F', | |
'InCurrency_Symbols' => '\u20A0-\u20CF', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20D0-\u20FF', | |
'InLetterlike_Symbols' => '\u2100-\u214F', | |
'InNumber_Forms' => '\u2150-\u218F', | |
'InArrows' => '\u2190-\u21FF', | |
'InMathematical_Operators' => '\u2200-\u22FF', | |
'InMiscellaneous_Technical' => '\u2300-\u23FF', | |
'InControl_Pictures' => '\u2400-\u243F', | |
'InOptical_Character_Recognition' => '\u2440-\u245F', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24FF', | |
'InBox_Drawing' => '\u2500-\u257F', | |
'InBlock_Elements' => '\u2580-\u259F', | |
'InGeometric_Shapes' => '\u25A0-\u25FF', | |
'InMiscellaneous_Symbols' => '\u2600-\u26FF', | |
'InDingbats' => '\u2700-\u27BF', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27C0-\u27EF', | |
'InSupplemental_Arrows-A' => '\u27F0-\u27FF', | |
'InBraille_Patterns' => '\u2800-\u28FF', | |
'InSupplemental_Arrows-B' => '\u2900-\u297F', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29FF', | |
'InSupplemental_Mathematical_Operators' => '\u2A00-\u2AFF', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2B00-\u2BFF', | |
'InCJK_Radicals_Supplement' => '\u2E80-\u2EFF', | |
'InKangxi_Radicals' => '\u2F00-\u2FDF', | |
'InIdeographic_Description_Characters' => '\u2FF0-\u2FFF', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303F', | |
'InHiragana' => '\u3040-\u309F', | |
'InKatakana' => '\u30A0-\u30FF', | |
'InBopomofo' => '\u3100-\u312F', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318F', | |
'InKanbun' => '\u3190-\u319F', | |
'InBopomofo_Extended' => '\u31A0-\u31BF', | |
'InKatakana_Phonetic_Extensions' => '\u31F0-\u31FF', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32FF', | |
'InCJK_Compatibility' => '\u3300-\u33FF', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4DBF', | |
'InYijing_Hexagram_Symbols' => '\u4DC0-\u4DFF', | |
'InCJK_Unified_Ideographs' => '\u4E00-\u9FFF', | |
'InYi_Syllables' => '\uA000-\uA48F', | |
'InYi_Radicals' => '\uA490-\uA4CF', | |
'InHangul_Syllables' => '\uAC00-\uD7AF', | |
'InHigh_Surrogates' => '\uD800-\uDB7F', | |
'InHigh_Private_Use_Surrogates' => '\uDB80-\uDBFF', | |
'InLow_Surrogates' => '\uDC00-\uDFFF', | |
'InPrivate_Use_Area' => '\uE000-\uF8FF', | |
'InCJK_Compatibility_Ideographs' => '\uF900-\uFAFF', | |
'InAlphabetic_Presentation_Forms' => '\uFB00-\uFB4F', | |
'InArabic_Presentation_Forms-A' => '\uFB50-\uFDFF', | |
'InVariation_Selectors' => '\uFE00-\uFE0F', | |
'InCombining_Half_Marks' => '\uFE20-\uFE2F', | |
'InCJK_Compatibility_Forms' => '\uFE30-\uFE4F', | |
'InSmall_Form_Variants' => '\uFE50-\uFE6F', | |
'InArabic_Presentation_Forms-B' => '\uFE70-\uFEFF', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uFF00-\uFFEF', | |
'InSpecials' => '\uFFF0-\uFFFF', | |
); | |
public static $equalRules = array( | |
'C' => array( | |
'InBasic_Latin' => '\u0000-\u001f\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u009f\u00ad', | |
'InGreek_and_Coptic' => '\u0378\u0379\u037f-\u0383\u038b\u038d\u03a2', | |
'InCyrillic_Supplementary' => '\u0526-\u052f', | |
'InArmenian' => '\u0530\u0557\u0558\u0560\u0588\u058b-\u058f', | |
'InHebrew' => '\u0590\u05c8-\u05cf\u05eb-\u05ef\u05f5-\u05ff', | |
'InArabic' => '\u0600-\u0605\u061c\u061d\u0620\u065f\u06dd', | |
'InSyriac' => '\u070e\u070f\u074b\u074c', | |
'InThaana' => '\u07b2-\u07bf', | |
'InDevanagari' => '\u093a\u093b\u094f\u0956\u0957\u0973-\u0978', | |
'InBengali' => '\u0980\u0984\u098d\u098e\u0991\u0992\u09a9\u09b1\u09b3-\u09b5\u09ba\u09bb\u09c5\u09c6\u09c9\u09ca\u09cf-\u09d6\u09d8-\u09db\u09de\u09e4\u09e5\u09fc-\u09ff', | |
'InGurmukhi' => '\u0a00\u0a04\u0a0b-\u0a0e\u0a11\u0a12\u0a29\u0a31\u0a34\u0a37\u0a3a\u0a3b\u0a3d\u0a43-\u0a46\u0a49\u0a4a\u0a4e-\u0a50\u0a52-\u0a58\u0a5d\u0a5f-\u0a65\u0a76-\u0a7f', | |
'InGujarati' => '\u0a80\u0a84\u0a8e\u0a92\u0aa9\u0ab1\u0ab4\u0aba\u0abb\u0ac6\u0aca\u0ace\u0acf\u0ad1-\u0adf\u0ae4\u0ae5\u0af0\u0af2-\u0aff', | |
'InOriya' => '\u0b00\u0b04\u0b0d\u0b0e\u0b11\u0b12\u0b29\u0b31\u0b34\u0b3a\u0b3b\u0b45\u0b46\u0b49\u0b4a\u0b4e-\u0b55\u0b58-\u0b5b\u0b5e\u0b64\u0b65\u0b72-\u0b7f', | |
'InTamil' => '\u0b80\u0b81\u0b84\u0b8b-\u0b8d\u0b91\u0b96-\u0b98\u0b9b\u0b9d\u0ba0-\u0ba2\u0ba5-\u0ba7\u0bab-\u0bad\u0bba-\u0bbd\u0bc3-\u0bc5\u0bc9\u0bce\u0bcf\u0bd1-\u0bd6\u0bd8-\u0be5\u0bfb-\u0bff', | |
'InTelugu' => '\u0c00\u0c04\u0c0d\u0c11\u0c29\u0c34\u0c3a-\u0c3c\u0c45\u0c49\u0c4e-\u0c54\u0c57\u0c5a-\u0c5f\u0c64\u0c65\u0c70-\u0c77', | |
'InKannada' => '\u0c80\u0c81\u0c84\u0c8d\u0c91\u0ca9\u0cb4\u0cba\u0cbb\u0cc5\u0cc9\u0cce-\u0cd4\u0cd7-\u0cdd\u0cdf\u0ce4\u0ce5\u0cf0\u0cf3-\u0cff', | |
'InMalayalam' => '\u0d00\u0d01\u0d04\u0d0d\u0d11\u0d29\u0d3a-\u0d3c\u0d45\u0d49\u0d4e-\u0d56\u0d58-\u0d5f\u0d64\u0d65\u0d76-\u0d78', | |
'InSinhala' => '\u0d80\u0d81\u0d84\u0d97-\u0d99\u0db2\u0dbc\u0dbe\u0dbf\u0dc7-\u0dc9\u0dcb-\u0dce\u0dd5\u0dd7\u0de0-\u0df1\u0df5-\u0dff', | |
'InThai' => '\u0e00\u0e3b-\u0e3e\u0e5c-\u0e7f', | |
'InLao' => '\u0e80\u0e83\u0e85\u0e86\u0e89\u0e8b\u0e8c\u0e8e-\u0e93\u0e98\u0ea0\u0ea4\u0ea6\u0ea8\u0ea9\u0eac\u0eba\u0ebe\u0ebf\u0ec5\u0ec7\u0ece\u0ecf\u0eda\u0edb\u0ede-\u0eff', | |
'InTibetan' => '\u0f48\u0f6d-\u0f70\u0f8c-\u0f8f\u0f98\u0fbd\u0fcd\u0fd9-\u0fff', | |
'InGeorgian' => '\u10c6-\u10cf\u10fd-\u10ff', | |
'InEthiopic' => '\u1249\u124e\u124f\u1257\u1259\u125e\u125f\u1289\u128e\u128f\u12b1\u12b6\u12b7\u12bf\u12c1\u12c6\u12c7\u12d7\u1311\u1316\u1317\u135b-\u135e\u137d-\u137f', | |
'InCherokee' => '\u13f5-\u13ff', | |
'InOgham' => '\u169d-\u169f', | |
'InRunic' => '\u16f1-\u16ff', | |
'InTagalog' => '\u170d\u1715-\u171f', | |
'InHanunoo' => '\u1737-\u173f', | |
'InBuhid' => '\u1754-\u175f', | |
'InTagbanwa' => '\u176d\u1771\u1774-\u177f', | |
'InKhmer' => '\u17b4\u17b5\u17de\u17df\u17ea-\u17ef\u17fa-\u17ff', | |
'InMongolian' => '\u180f\u181a-\u181f\u1878-\u187f\u18ab-\u18af', | |
'InLimbu' => '\u191d-\u191f\u192c-\u192f\u193c-\u193f\u1941-\u1943', | |
'InTai_Le' => '\u196e\u196f\u1975-\u197f', | |
'InGreek_Extended' => '\u1f16\u1f17\u1f1e\u1f1f\u1f46\u1f47\u1f4e\u1f4f\u1f58\u1f5a\u1f5c\u1f5e\u1f7e\u1f7f\u1fb5\u1fc5\u1fd4\u1fd5\u1fdc\u1ff0\u1ff1\u1ff5\u1fff', | |
'InGeneral_Punctuation' => '\u200b-\u200f\u202a-\u202e\u2060-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2072\u2073\u208f\u2095-\u209f', | |
'InCurrency_Symbols' => '\u20b9-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20f1-\u20ff', | |
'InNumber_Forms' => '\u218a-\u218f', | |
'InMiscellaneous_Technical' => '\u23e9-\u23ff', | |
'InControl_Pictures' => '\u2427-\u243f', | |
'InOptical_Character_Recognition' => '\u244b-\u245f', | |
'InMiscellaneous_Symbols' => '\u26ce\u26e2\u26e4-\u26e7', | |
'InDingbats' => '\u2700\u2705\u270a\u270b\u2728\u274c\u274e\u2753-\u2755\u275f\u2760\u2795-\u2797\u27b0\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27cb\u27cd-\u27cf', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b4d-\u2b4f\u2b5a-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e9a\u2ef4-\u2eff', | |
'InKangxi_Radicals' => '\u2fd6-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ffc-\u2fff', | |
'InHiragana' => '\u3040\u3097\u3098', | |
'InBopomofo' => '\u3100-\u3104\u312e\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130\u318f', | |
'InBopomofo_Extended' => '\u31b8-\u31bf', | |
'InEnclosed_CJK_Letters_and_Months' => '\u321f\u32ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u4db6-\u4dbf', | |
'InCJK_Unified_Ideographs' => '\u9fcc-\u9fff', | |
'InYi_Syllables' => '\ua48d-\ua48f', | |
'InYi_Radicals' => '\ua4c7-\ua4cf', | |
'InHangul_Syllables' => '\ud7a4-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\ufa2e\ufa2f\ufa6e\ufa6f\ufada-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb07-\ufb12\ufb18-\ufb1c\ufb37\ufb3d\ufb3f\ufb42\ufb45', | |
'InArabic_Presentation_Forms-A' => '\ufbb2-\ufbd2\ufd40-\ufd4f\ufd90\ufd91\ufdc8-\ufdef\ufdfe\ufdff', | |
'InCombining_Half_Marks' => '\ufe27-\ufe2f', | |
'InSmall_Form_Variants' => '\ufe53\ufe67\ufe6c-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe75\ufefd-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00\uffbf-\uffc1\uffc8\uffc9\uffd0\uffd1\uffd8\uffd9\uffdd-\uffdf\uffe7\uffef', | |
'InSpecials' => '\ufff0-\ufffb\ufffe\uffff', | |
), | |
'Cc' => array( | |
'InBasic_Latin' => '\u0000-\u001f\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u009f', | |
), | |
'Cf' => array( | |
'InLatin-1_Supplement' => '\u00ad', | |
'InArabic' => '\u0600-\u0603\u06dd', | |
'InSyriac' => '\u070f', | |
'InKhmer' => '\u17b4\u17b5', | |
'InGeneral_Punctuation' => '\u200b-\u200f\u202a-\u202e\u2060-\u2064\u206a-\u206f', | |
'InArabic_Presentation_Forms-B' => '\ufeff', | |
'InSpecials' => '\ufff9-\ufffb', | |
), | |
'Cn' => array( | |
'InGreek_and_Coptic' => '\u0378\u0379\u037f-\u0383\u038b\u038d\u03a2', | |
'InCyrillic_Supplementary' => '\u0526-\u052f', | |
'InArmenian' => '\u0530\u0557\u0558\u0560\u0588\u058b-\u058f', | |
'InHebrew' => '\u0590\u05c8-\u05cf\u05eb-\u05ef\u05f5-\u05ff', | |
'InArabic' => '\u0604\u0605\u061c\u061d\u0620\u065f', | |
'InSyriac' => '\u070e\u074b\u074c', | |
'InThaana' => '\u07b2-\u07bf', | |
'InDevanagari' => '\u093a\u093b\u094f\u0956\u0957\u0973-\u0978', | |
'InBengali' => '\u0980\u0984\u098d\u098e\u0991\u0992\u09a9\u09b1\u09b3-\u09b5\u09ba\u09bb\u09c5\u09c6\u09c9\u09ca\u09cf-\u09d6\u09d8-\u09db\u09de\u09e4\u09e5\u09fc-\u09ff', | |
'InGurmukhi' => '\u0a00\u0a04\u0a0b-\u0a0e\u0a11\u0a12\u0a29\u0a31\u0a34\u0a37\u0a3a\u0a3b\u0a3d\u0a43-\u0a46\u0a49\u0a4a\u0a4e-\u0a50\u0a52-\u0a58\u0a5d\u0a5f-\u0a65\u0a76-\u0a7f', | |
'InGujarati' => '\u0a80\u0a84\u0a8e\u0a92\u0aa9\u0ab1\u0ab4\u0aba\u0abb\u0ac6\u0aca\u0ace\u0acf\u0ad1-\u0adf\u0ae4\u0ae5\u0af0\u0af2-\u0aff', | |
'InOriya' => '\u0b00\u0b04\u0b0d\u0b0e\u0b11\u0b12\u0b29\u0b31\u0b34\u0b3a\u0b3b\u0b45\u0b46\u0b49\u0b4a\u0b4e-\u0b55\u0b58-\u0b5b\u0b5e\u0b64\u0b65\u0b72-\u0b7f', | |
'InTamil' => '\u0b80\u0b81\u0b84\u0b8b-\u0b8d\u0b91\u0b96-\u0b98\u0b9b\u0b9d\u0ba0-\u0ba2\u0ba5-\u0ba7\u0bab-\u0bad\u0bba-\u0bbd\u0bc3-\u0bc5\u0bc9\u0bce\u0bcf\u0bd1-\u0bd6\u0bd8-\u0be5\u0bfb-\u0bff', | |
'InTelugu' => '\u0c00\u0c04\u0c0d\u0c11\u0c29\u0c34\u0c3a-\u0c3c\u0c45\u0c49\u0c4e-\u0c54\u0c57\u0c5a-\u0c5f\u0c64\u0c65\u0c70-\u0c77', | |
'InKannada' => '\u0c80\u0c81\u0c84\u0c8d\u0c91\u0ca9\u0cb4\u0cba\u0cbb\u0cc5\u0cc9\u0cce-\u0cd4\u0cd7-\u0cdd\u0cdf\u0ce4\u0ce5\u0cf0\u0cf3-\u0cff', | |
'InMalayalam' => '\u0d00\u0d01\u0d04\u0d0d\u0d11\u0d29\u0d3a-\u0d3c\u0d45\u0d49\u0d4e-\u0d56\u0d58-\u0d5f\u0d64\u0d65\u0d76-\u0d78', | |
'InSinhala' => '\u0d80\u0d81\u0d84\u0d97-\u0d99\u0db2\u0dbc\u0dbe\u0dbf\u0dc7-\u0dc9\u0dcb-\u0dce\u0dd5\u0dd7\u0de0-\u0df1\u0df5-\u0dff', | |
'InThai' => '\u0e00\u0e3b-\u0e3e\u0e5c-\u0e7f', | |
'InLao' => '\u0e80\u0e83\u0e85\u0e86\u0e89\u0e8b\u0e8c\u0e8e-\u0e93\u0e98\u0ea0\u0ea4\u0ea6\u0ea8\u0ea9\u0eac\u0eba\u0ebe\u0ebf\u0ec5\u0ec7\u0ece\u0ecf\u0eda\u0edb\u0ede-\u0eff', | |
'InTibetan' => '\u0f48\u0f6d-\u0f70\u0f8c-\u0f8f\u0f98\u0fbd\u0fcd\u0fd9-\u0fff', | |
'InGeorgian' => '\u10c6-\u10cf\u10fd-\u10ff', | |
'InEthiopic' => '\u1249\u124e\u124f\u1257\u1259\u125e\u125f\u1289\u128e\u128f\u12b1\u12b6\u12b7\u12bf\u12c1\u12c6\u12c7\u12d7\u1311\u1316\u1317\u135b-\u135e\u137d-\u137f', | |
'InCherokee' => '\u13f5-\u13ff', | |
'InOgham' => '\u169d-\u169f', | |
'InRunic' => '\u16f1-\u16ff', | |
'InTagalog' => '\u170d\u1715-\u171f', | |
'InHanunoo' => '\u1737-\u173f', | |
'InBuhid' => '\u1754-\u175f', | |
'InTagbanwa' => '\u176d\u1771\u1774-\u177f', | |
'InKhmer' => '\u17de\u17df\u17ea-\u17ef\u17fa-\u17ff', | |
'InMongolian' => '\u180f\u181a-\u181f\u1878-\u187f\u18ab-\u18af', | |
'InLimbu' => '\u191d-\u191f\u192c-\u192f\u193c-\u193f\u1941-\u1943', | |
'InTai_Le' => '\u196e\u196f\u1975-\u197f', | |
'InGreek_Extended' => '\u1f16\u1f17\u1f1e\u1f1f\u1f46\u1f47\u1f4e\u1f4f\u1f58\u1f5a\u1f5c\u1f5e\u1f7e\u1f7f\u1fb5\u1fc5\u1fd4\u1fd5\u1fdc\u1ff0\u1ff1\u1ff5\u1fff', | |
'InGeneral_Punctuation' => '\u2065-\u2069', | |
'InSuperscripts_and_Subscripts' => '\u2072\u2073\u208f\u2095-\u209f', | |
'InCurrency_Symbols' => '\u20b9-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20f1-\u20ff', | |
'InNumber_Forms' => '\u218a-\u218f', | |
'InMiscellaneous_Technical' => '\u23e9-\u23ff', | |
'InControl_Pictures' => '\u2427-\u243f', | |
'InOptical_Character_Recognition' => '\u244b-\u245f', | |
'InMiscellaneous_Symbols' => '\u26ce\u26e2\u26e4-\u26e7', | |
'InDingbats' => '\u2700\u2705\u270a\u270b\u2728\u274c\u274e\u2753-\u2755\u275f\u2760\u2795-\u2797\u27b0\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27cb\u27cd-\u27cf', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b4d-\u2b4f\u2b5a-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e9a\u2ef4-\u2eff', | |
'InKangxi_Radicals' => '\u2fd6-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ffc-\u2fff', | |
'InHiragana' => '\u3040\u3097\u3098', | |
'InBopomofo' => '\u3100-\u3104\u312e\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130\u318f', | |
'InBopomofo_Extended' => '\u31b8-\u31bf', | |
'InEnclosed_CJK_Letters_and_Months' => '\u321f\u32ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u4db6-\u4dbf', | |
'InCJK_Unified_Ideographs' => '\u9fcc-\u9fff', | |
'InYi_Syllables' => '\ua48d-\ua48f', | |
'InYi_Radicals' => '\ua4c7-\ua4cf', | |
'InHangul_Syllables' => '\ud7a4-\ud7af', | |
'InCJK_Compatibility_Ideographs' => '\ufa2e\ufa2f\ufa6e\ufa6f\ufada-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb07-\ufb12\ufb18-\ufb1c\ufb37\ufb3d\ufb3f\ufb42\ufb45', | |
'InArabic_Presentation_Forms-A' => '\ufbb2-\ufbd2\ufd40-\ufd4f\ufd90\ufd91\ufdc8-\ufdef\ufdfe\ufdff', | |
'InCombining_Half_Marks' => '\ufe27-\ufe2f', | |
'InSmall_Form_Variants' => '\ufe53\ufe67\ufe6c-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe75\ufefd\ufefe', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00\uffbf-\uffc1\uffc8\uffc9\uffd0\uffd1\uffd8\uffd9\uffdd-\uffdf\uffe7\uffef', | |
'InSpecials' => '\ufff0-\ufff8\ufffe\uffff', | |
), | |
'Co' => array( | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
), | |
'Cs' => array( | |
), | |
'L' => array( | |
'InBasic_Latin' => '\u0041-\u005a\u0061-\u007a', | |
'InLatin-1_Supplement' => '\u00aa\u00b5\u00ba\u00c0-\u00d6\u00d8-\u00f6\u00f8-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02c1\u02c6-\u02d1\u02e0-\u02e4\u02ec\u02ee', | |
'InGreek_and_Coptic' => '\u0370-\u0374\u0376\u0377\u037a-\u037d\u0386\u0388-\u038a\u038c\u038e-\u03a1\u03a3-\u03f5\u03f7-\u03ff', | |
'InCyrillic' => '\u0400-\u0481\u048a-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u0525', | |
'InArmenian' => '\u0531-\u0556\u0559\u0561-\u0587', | |
'InHebrew' => '\u05d0-\u05ea\u05f0-\u05f2', | |
'InArabic' => '\u0621-\u064a\u066e\u066f\u0671-\u06d3\u06d5\u06e5\u06e6\u06ee\u06ef\u06fa-\u06fc\u06ff', | |
'InSyriac' => '\u0710\u0712-\u072f\u074d-\u074f', | |
'InThaana' => '\u0780-\u07a5\u07b1', | |
'InDevanagari' => '\u0904-\u0939\u093d\u0950\u0958-\u0961\u0971\u0972\u0979-\u097f', | |
'InBengali' => '\u0985-\u098c\u098f\u0990\u0993-\u09a8\u09aa-\u09b0\u09b2\u09b6-\u09b9\u09bd\u09ce\u09dc\u09dd\u09df-\u09e1\u09f0\u09f1', | |
'InGurmukhi' => '\u0a05-\u0a0a\u0a0f\u0a10\u0a13-\u0a28\u0a2a-\u0a30\u0a32\u0a33\u0a35\u0a36\u0a38\u0a39\u0a59-\u0a5c\u0a5e\u0a72-\u0a74', | |
'InGujarati' => '\u0a85-\u0a8d\u0a8f-\u0a91\u0a93-\u0aa8\u0aaa-\u0ab0\u0ab2\u0ab3\u0ab5-\u0ab9\u0abd\u0ad0\u0ae0\u0ae1', | |
'InOriya' => '\u0b05-\u0b0c\u0b0f\u0b10\u0b13-\u0b28\u0b2a-\u0b30\u0b32\u0b33\u0b35-\u0b39\u0b3d\u0b5c\u0b5d\u0b5f-\u0b61\u0b71', | |
'InTamil' => '\u0b83\u0b85-\u0b8a\u0b8e-\u0b90\u0b92-\u0b95\u0b99\u0b9a\u0b9c\u0b9e\u0b9f\u0ba3\u0ba4\u0ba8-\u0baa\u0bae-\u0bb9\u0bd0', | |
'InTelugu' => '\u0c05-\u0c0c\u0c0e-\u0c10\u0c12-\u0c28\u0c2a-\u0c33\u0c35-\u0c39\u0c3d\u0c58\u0c59\u0c60\u0c61', | |
'InKannada' => '\u0c85-\u0c8c\u0c8e-\u0c90\u0c92-\u0ca8\u0caa-\u0cb3\u0cb5-\u0cb9\u0cbd\u0cde\u0ce0\u0ce1', | |
'InMalayalam' => '\u0d05-\u0d0c\u0d0e-\u0d10\u0d12-\u0d28\u0d2a-\u0d39\u0d3d\u0d60\u0d61\u0d7a-\u0d7f', | |
'InSinhala' => '\u0d85-\u0d96\u0d9a-\u0db1\u0db3-\u0dbb\u0dbd\u0dc0-\u0dc6', | |
'InThai' => '\u0e01-\u0e30\u0e32\u0e33\u0e40-\u0e46', | |
'InLao' => '\u0e81\u0e82\u0e84\u0e87\u0e88\u0e8a\u0e8d\u0e94-\u0e97\u0e99-\u0e9f\u0ea1-\u0ea3\u0ea5\u0ea7\u0eaa\u0eab\u0ead-\u0eb0\u0eb2\u0eb3\u0ebd\u0ec0-\u0ec4\u0ec6\u0edc\u0edd', | |
'InTibetan' => '\u0f00\u0f40-\u0f47\u0f49-\u0f6c\u0f88-\u0f8b', | |
'InMyanmar' => '\u1000-\u102a\u103f\u1050-\u1055\u105a-\u105d\u1061\u1065\u1066\u106e-\u1070\u1075-\u1081\u108e', | |
'InGeorgian' => '\u10a0-\u10c5\u10d0-\u10fa\u10fc', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u1248\u124a-\u124d\u1250-\u1256\u1258\u125a-\u125d\u1260-\u1288\u128a-\u128d\u1290-\u12b0\u12b2-\u12b5\u12b8-\u12be\u12c0\u12c2-\u12c5\u12c8-\u12d6\u12d8-\u1310\u1312-\u1315\u1318-\u135a', | |
'InCherokee' => '\u13a0-\u13f4', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1401-\u166c\u166f-\u167f', | |
'InOgham' => '\u1681-\u169a', | |
'InRunic' => '\u16a0-\u16ea', | |
'InTagalog' => '\u1700-\u170c\u170e-\u1711', | |
'InHanunoo' => '\u1720-\u1731', | |
'InBuhid' => '\u1740-\u1751', | |
'InTagbanwa' => '\u1760-\u176c\u176e-\u1770', | |
'InKhmer' => '\u1780-\u17b3\u17d7\u17dc', | |
'InMongolian' => '\u1820-\u1877\u1880-\u18a8\u18aa', | |
'InLimbu' => '\u1900-\u191c', | |
'InTai_Le' => '\u1950-\u196d\u1970-\u1974', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1f15\u1f18-\u1f1d\u1f20-\u1f45\u1f48-\u1f4d\u1f50-\u1f57\u1f59\u1f5b\u1f5d\u1f5f-\u1f7d\u1f80-\u1fb4\u1fb6-\u1fbc\u1fbe\u1fc2-\u1fc4\u1fc6-\u1fcc\u1fd0-\u1fd3\u1fd6-\u1fdb\u1fe0-\u1fec\u1ff2-\u1ff4\u1ff6-\u1ffc', | |
'InSuperscripts_and_Subscripts' => '\u2071\u207f\u2090-\u2094', | |
'InLetterlike_Symbols' => '\u2102\u2107\u210a-\u2113\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u212f-\u2139\u213c-\u213f\u2145-\u2149\u214e', | |
'InNumber_Forms' => '\u2183\u2184', | |
'InCJK_Symbols_and_Punctuation' => '\u3005\u3006\u3031-\u3035\u303b\u303c', | |
'InHiragana' => '\u3041-\u3096\u309d-\u309f', | |
'InKatakana' => '\u30a1-\u30fa\u30fc-\u30ff', | |
'InBopomofo' => '\u3105-\u312d', | |
'InHangul_Compatibility_Jamo' => '\u3131-\u318e', | |
'InBopomofo_Extended' => '\u31a0-\u31b7', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4db5', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fcb', | |
'InYi_Syllables' => '\ua000-\ua48c', | |
'InHangul_Syllables' => '\uac00-\ud7a3', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufa2d\ufa30-\ufa6d\ufa70-\ufad9', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb06\ufb13-\ufb17\ufb1d\ufb1f-\ufb28\ufb2a-\ufb36\ufb38-\ufb3c\ufb3e\ufb40\ufb41\ufb43\ufb44\ufb46-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufbb1\ufbd3-\ufd3d\ufd50-\ufd8f\ufd92-\ufdc7\ufdf0-\ufdfb', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufe74\ufe76-\ufefc', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff21-\uff3a\uff41-\uff5a\uff66-\uffbe\uffc2-\uffc7\uffca-\uffcf\uffd2-\uffd7\uffda-\uffdc', | |
), | |
'Ll' => array( | |
'InBasic_Latin' => '\u0061-\u007a', | |
'InLatin-1_Supplement' => '\u00aa\u00b5\u00ba\u00df-\u00f6\u00f8-\u00ff', | |
'InLatin_Extended-A' => '\u0101\u0103\u0105\u0107\u0109\u010b\u010d\u010f\u0111\u0113\u0115\u0117\u0119\u011b\u011d\u011f\u0121\u0123\u0125\u0127\u0129\u012b\u012d\u012f\u0131\u0133\u0135\u0137\u0138\u013a\u013c\u013e\u0140\u0142\u0144\u0146\u0148\u0149\u014b\u014d\u014f\u0151\u0153\u0155\u0157\u0159\u015b\u015d\u015f\u0161\u0163\u0165\u0167\u0169\u016b\u016d\u016f\u0171\u0173\u0175\u0177\u017a\u017c\u017e\u017f', | |
'InLatin_Extended-B' => '\u0180\u0183\u0185\u0188\u018c\u018d\u0192\u0195\u0199-\u019b\u019e\u01a1\u01a3\u01a5\u01a8\u01aa\u01ab\u01ad\u01b0\u01b4\u01b6\u01b9\u01ba\u01bd-\u01bf\u01c6\u01c9\u01cc\u01ce\u01d0\u01d2\u01d4\u01d6\u01d8\u01da\u01dc\u01dd\u01df\u01e1\u01e3\u01e5\u01e7\u01e9\u01eb\u01ed\u01ef\u01f0\u01f3\u01f5\u01f9\u01fb\u01fd\u01ff\u0201\u0203\u0205\u0207\u0209\u020b\u020d\u020f\u0211\u0213\u0215\u0217\u0219\u021b\u021d\u021f\u0221\u0223\u0225\u0227\u0229\u022b\u022d\u022f\u0231\u0233-\u0239\u023c\u023f\u0240\u0242\u0247\u0249\u024b\u024d\u024f', | |
'InIPA_Extensions' => '\u0250-\u0293\u0295-\u02af', | |
'InGreek_and_Coptic' => '\u0371\u0373\u0377\u037b-\u037d\u0390\u03ac-\u03ce\u03d0\u03d1\u03d5-\u03d7\u03d9\u03db\u03dd\u03df\u03e1\u03e3\u03e5\u03e7\u03e9\u03eb\u03ed\u03ef-\u03f3\u03f5\u03f8\u03fb\u03fc', | |
'InCyrillic' => '\u0430-\u045f\u0461\u0463\u0465\u0467\u0469\u046b\u046d\u046f\u0471\u0473\u0475\u0477\u0479\u047b\u047d\u047f\u0481\u048b\u048d\u048f\u0491\u0493\u0495\u0497\u0499\u049b\u049d\u049f\u04a1\u04a3\u04a5\u04a7\u04a9\u04ab\u04ad\u04af\u04b1\u04b3\u04b5\u04b7\u04b9\u04bb\u04bd\u04bf\u04c2\u04c4\u04c6\u04c8\u04ca\u04cc\u04ce\u04cf\u04d1\u04d3\u04d5\u04d7\u04d9\u04db\u04dd\u04df\u04e1\u04e3\u04e5\u04e7\u04e9\u04eb\u04ed\u04ef\u04f1\u04f3\u04f5\u04f7\u04f9\u04fb\u04fd\u04ff', | |
'InCyrillic_Supplementary' => '\u0501\u0503\u0505\u0507\u0509\u050b\u050d\u050f\u0511\u0513\u0515\u0517\u0519\u051b\u051d\u051f\u0521\u0523\u0525', | |
'InArmenian' => '\u0561-\u0587', | |
'InPhonetic_Extensions' => '\u1d00-\u1d2b\u1d62-\u1d77\u1d79-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e01\u1e03\u1e05\u1e07\u1e09\u1e0b\u1e0d\u1e0f\u1e11\u1e13\u1e15\u1e17\u1e19\u1e1b\u1e1d\u1e1f\u1e21\u1e23\u1e25\u1e27\u1e29\u1e2b\u1e2d\u1e2f\u1e31\u1e33\u1e35\u1e37\u1e39\u1e3b\u1e3d\u1e3f\u1e41\u1e43\u1e45\u1e47\u1e49\u1e4b\u1e4d\u1e4f\u1e51\u1e53\u1e55\u1e57\u1e59\u1e5b\u1e5d\u1e5f\u1e61\u1e63\u1e65\u1e67\u1e69\u1e6b\u1e6d\u1e6f\u1e71\u1e73\u1e75\u1e77\u1e79\u1e7b\u1e7d\u1e7f\u1e81\u1e83\u1e85\u1e87\u1e89\u1e8b\u1e8d\u1e8f\u1e91\u1e93\u1e95-\u1e9d\u1e9f\u1ea1\u1ea3\u1ea5\u1ea7\u1ea9\u1eab\u1ead\u1eaf\u1eb1\u1eb3\u1eb5\u1eb7\u1eb9\u1ebb\u1ebd\u1ebf\u1ec1\u1ec3\u1ec5\u1ec7\u1ec9\u1ecb\u1ecd\u1ecf\u1ed1\u1ed3\u1ed5\u1ed7\u1ed9\u1edb\u1edd\u1edf\u1ee1\u1ee3\u1ee5\u1ee7\u1ee9\u1eeb\u1eed\u1eef\u1ef1\u1ef3\u1ef5\u1ef7\u1ef9\u1efb\u1efd\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1f07\u1f10-\u1f15\u1f20-\u1f27\u1f30-\u1f37\u1f40-\u1f45\u1f50-\u1f57\u1f60-\u1f67\u1f70-\u1f7d\u1f80-\u1f87\u1f90-\u1f97\u1fa0-\u1fa7\u1fb0-\u1fb4\u1fb6\u1fb7\u1fbe\u1fc2-\u1fc4\u1fc6\u1fc7\u1fd0-\u1fd3\u1fd6\u1fd7\u1fe0-\u1fe7\u1ff2-\u1ff4\u1ff6\u1ff7', | |
'InLetterlike_Symbols' => '\u210a\u210e\u210f\u2113\u212f\u2134\u2139\u213c\u213d\u2146-\u2149\u214e', | |
'InNumber_Forms' => '\u2184', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb06\ufb13-\ufb17', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff41-\uff5a', | |
), | |
'Lm' => array( | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02c1\u02c6-\u02d1\u02e0-\u02e4\u02ec\u02ee', | |
'InGreek_and_Coptic' => '\u0374\u037a', | |
'InArmenian' => '\u0559', | |
'InArabic' => '\u0640\u06e5\u06e6', | |
'InDevanagari' => '\u0971', | |
'InThai' => '\u0e46', | |
'InLao' => '\u0ec6', | |
'InGeorgian' => '\u10fc', | |
'InKhmer' => '\u17d7', | |
'InMongolian' => '\u1843', | |
'InPhonetic_Extensions' => '\u1d2c-\u1d61\u1d78', | |
'InSuperscripts_and_Subscripts' => '\u2071\u207f\u2090-\u2094', | |
'InCJK_Symbols_and_Punctuation' => '\u3005\u3031-\u3035\u303b', | |
'InHiragana' => '\u309d\u309e', | |
'InKatakana' => '\u30fc-\u30fe', | |
'InYi_Syllables' => '\ua015', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff70\uff9e\uff9f', | |
), | |
'Lo' => array( | |
'InLatin_Extended-B' => '\u01bb\u01c0-\u01c3', | |
'InIPA_Extensions' => '\u0294', | |
'InHebrew' => '\u05d0-\u05ea\u05f0-\u05f2', | |
'InArabic' => '\u0621-\u063f\u0641-\u064a\u066e\u066f\u0671-\u06d3\u06d5\u06ee\u06ef\u06fa-\u06fc\u06ff', | |
'InSyriac' => '\u0710\u0712-\u072f\u074d-\u074f', | |
'InThaana' => '\u0780-\u07a5\u07b1', | |
'InDevanagari' => '\u0904-\u0939\u093d\u0950\u0958-\u0961\u0972\u0979-\u097f', | |
'InBengali' => '\u0985-\u098c\u098f\u0990\u0993-\u09a8\u09aa-\u09b0\u09b2\u09b6-\u09b9\u09bd\u09ce\u09dc\u09dd\u09df-\u09e1\u09f0\u09f1', | |
'InGurmukhi' => '\u0a05-\u0a0a\u0a0f\u0a10\u0a13-\u0a28\u0a2a-\u0a30\u0a32\u0a33\u0a35\u0a36\u0a38\u0a39\u0a59-\u0a5c\u0a5e\u0a72-\u0a74', | |
'InGujarati' => '\u0a85-\u0a8d\u0a8f-\u0a91\u0a93-\u0aa8\u0aaa-\u0ab0\u0ab2\u0ab3\u0ab5-\u0ab9\u0abd\u0ad0\u0ae0\u0ae1', | |
'InOriya' => '\u0b05-\u0b0c\u0b0f\u0b10\u0b13-\u0b28\u0b2a-\u0b30\u0b32\u0b33\u0b35-\u0b39\u0b3d\u0b5c\u0b5d\u0b5f-\u0b61\u0b71', | |
'InTamil' => '\u0b83\u0b85-\u0b8a\u0b8e-\u0b90\u0b92-\u0b95\u0b99\u0b9a\u0b9c\u0b9e\u0b9f\u0ba3\u0ba4\u0ba8-\u0baa\u0bae-\u0bb9\u0bd0', | |
'InTelugu' => '\u0c05-\u0c0c\u0c0e-\u0c10\u0c12-\u0c28\u0c2a-\u0c33\u0c35-\u0c39\u0c3d\u0c58\u0c59\u0c60\u0c61', | |
'InKannada' => '\u0c85-\u0c8c\u0c8e-\u0c90\u0c92-\u0ca8\u0caa-\u0cb3\u0cb5-\u0cb9\u0cbd\u0cde\u0ce0\u0ce1', | |
'InMalayalam' => '\u0d05-\u0d0c\u0d0e-\u0d10\u0d12-\u0d28\u0d2a-\u0d39\u0d3d\u0d60\u0d61\u0d7a-\u0d7f', | |
'InSinhala' => '\u0d85-\u0d96\u0d9a-\u0db1\u0db3-\u0dbb\u0dbd\u0dc0-\u0dc6', | |
'InThai' => '\u0e01-\u0e30\u0e32\u0e33\u0e40-\u0e45', | |
'InLao' => '\u0e81\u0e82\u0e84\u0e87\u0e88\u0e8a\u0e8d\u0e94-\u0e97\u0e99-\u0e9f\u0ea1-\u0ea3\u0ea5\u0ea7\u0eaa\u0eab\u0ead-\u0eb0\u0eb2\u0eb3\u0ebd\u0ec0-\u0ec4\u0edc\u0edd', | |
'InTibetan' => '\u0f00\u0f40-\u0f47\u0f49-\u0f6c\u0f88-\u0f8b', | |
'InMyanmar' => '\u1000-\u102a\u103f\u1050-\u1055\u105a-\u105d\u1061\u1065\u1066\u106e-\u1070\u1075-\u1081\u108e', | |
'InGeorgian' => '\u10d0-\u10fa', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u1248\u124a-\u124d\u1250-\u1256\u1258\u125a-\u125d\u1260-\u1288\u128a-\u128d\u1290-\u12b0\u12b2-\u12b5\u12b8-\u12be\u12c0\u12c2-\u12c5\u12c8-\u12d6\u12d8-\u1310\u1312-\u1315\u1318-\u135a', | |
'InCherokee' => '\u13a0-\u13f4', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1401-\u166c\u166f-\u167f', | |
'InOgham' => '\u1681-\u169a', | |
'InRunic' => '\u16a0-\u16ea', | |
'InTagalog' => '\u1700-\u170c\u170e-\u1711', | |
'InHanunoo' => '\u1720-\u1731', | |
'InBuhid' => '\u1740-\u1751', | |
'InTagbanwa' => '\u1760-\u176c\u176e-\u1770', | |
'InKhmer' => '\u1780-\u17b3\u17dc', | |
'InMongolian' => '\u1820-\u1842\u1844-\u1877\u1880-\u18a8\u18aa', | |
'InLimbu' => '\u1900-\u191c', | |
'InTai_Le' => '\u1950-\u196d\u1970-\u1974', | |
'InLetterlike_Symbols' => '\u2135-\u2138', | |
'InCJK_Symbols_and_Punctuation' => '\u3006\u303c', | |
'InHiragana' => '\u3041-\u3096\u309f', | |
'InKatakana' => '\u30a1-\u30fa\u30ff', | |
'InBopomofo' => '\u3105-\u312d', | |
'InHangul_Compatibility_Jamo' => '\u3131-\u318e', | |
'InBopomofo_Extended' => '\u31a0-\u31b7', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4db5', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fcb', | |
'InYi_Syllables' => '\ua000-\ua014\ua016-\ua48c', | |
'InHangul_Syllables' => '\uac00-\ud7a3', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufa2d\ufa30-\ufa6d\ufa70-\ufad9', | |
'InAlphabetic_Presentation_Forms' => '\ufb1d\ufb1f-\ufb28\ufb2a-\ufb36\ufb38-\ufb3c\ufb3e\ufb40\ufb41\ufb43\ufb44\ufb46-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufbb1\ufbd3-\ufd3d\ufd50-\ufd8f\ufd92-\ufdc7\ufdf0-\ufdfb', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufe74\ufe76-\ufefc', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff66-\uff6f\uff71-\uff9d\uffa0-\uffbe\uffc2-\uffc7\uffca-\uffcf\uffd2-\uffd7\uffda-\uffdc', | |
), | |
'Lt' => array( | |
'InLatin_Extended-B' => '\u01c5\u01c8\u01cb\u01f2', | |
'InGreek_Extended' => '\u1f88-\u1f8f\u1f98-\u1f9f\u1fa8-\u1faf\u1fbc\u1fcc\u1ffc', | |
), | |
'Lu' => array( | |
'InBasic_Latin' => '\u0041-\u005a', | |
'InLatin-1_Supplement' => '\u00c0-\u00d6\u00d8-\u00de', | |
'InLatin_Extended-A' => '\u0100\u0102\u0104\u0106\u0108\u010a\u010c\u010e\u0110\u0112\u0114\u0116\u0118\u011a\u011c\u011e\u0120\u0122\u0124\u0126\u0128\u012a\u012c\u012e\u0130\u0132\u0134\u0136\u0139\u013b\u013d\u013f\u0141\u0143\u0145\u0147\u014a\u014c\u014e\u0150\u0152\u0154\u0156\u0158\u015a\u015c\u015e\u0160\u0162\u0164\u0166\u0168\u016a\u016c\u016e\u0170\u0172\u0174\u0176\u0178\u0179\u017b\u017d', | |
'InLatin_Extended-B' => '\u0181\u0182\u0184\u0186\u0187\u0189-\u018b\u018e-\u0191\u0193\u0194\u0196-\u0198\u019c\u019d\u019f\u01a0\u01a2\u01a4\u01a6\u01a7\u01a9\u01ac\u01ae\u01af\u01b1-\u01b3\u01b5\u01b7\u01b8\u01bc\u01c4\u01c7\u01ca\u01cd\u01cf\u01d1\u01d3\u01d5\u01d7\u01d9\u01db\u01de\u01e0\u01e2\u01e4\u01e6\u01e8\u01ea\u01ec\u01ee\u01f1\u01f4\u01f6-\u01f8\u01fa\u01fc\u01fe\u0200\u0202\u0204\u0206\u0208\u020a\u020c\u020e\u0210\u0212\u0214\u0216\u0218\u021a\u021c\u021e\u0220\u0222\u0224\u0226\u0228\u022a\u022c\u022e\u0230\u0232\u023a\u023b\u023d\u023e\u0241\u0243-\u0246\u0248\u024a\u024c\u024e', | |
'InGreek_and_Coptic' => '\u0370\u0372\u0376\u0386\u0388-\u038a\u038c\u038e\u038f\u0391-\u03a1\u03a3-\u03ab\u03cf\u03d2-\u03d4\u03d8\u03da\u03dc\u03de\u03e0\u03e2\u03e4\u03e6\u03e8\u03ea\u03ec\u03ee\u03f4\u03f7\u03f9\u03fa\u03fd-\u03ff', | |
'InCyrillic' => '\u0400-\u042f\u0460\u0462\u0464\u0466\u0468\u046a\u046c\u046e\u0470\u0472\u0474\u0476\u0478\u047a\u047c\u047e\u0480\u048a\u048c\u048e\u0490\u0492\u0494\u0496\u0498\u049a\u049c\u049e\u04a0\u04a2\u04a4\u04a6\u04a8\u04aa\u04ac\u04ae\u04b0\u04b2\u04b4\u04b6\u04b8\u04ba\u04bc\u04be\u04c0\u04c1\u04c3\u04c5\u04c7\u04c9\u04cb\u04cd\u04d0\u04d2\u04d4\u04d6\u04d8\u04da\u04dc\u04de\u04e0\u04e2\u04e4\u04e6\u04e8\u04ea\u04ec\u04ee\u04f0\u04f2\u04f4\u04f6\u04f8\u04fa\u04fc\u04fe', | |
'InCyrillic_Supplementary' => '\u0500\u0502\u0504\u0506\u0508\u050a\u050c\u050e\u0510\u0512\u0514\u0516\u0518\u051a\u051c\u051e\u0520\u0522\u0524', | |
'InArmenian' => '\u0531-\u0556', | |
'InGeorgian' => '\u10a0-\u10c5', | |
'InLatin_Extended_Additional' => '\u1e00\u1e02\u1e04\u1e06\u1e08\u1e0a\u1e0c\u1e0e\u1e10\u1e12\u1e14\u1e16\u1e18\u1e1a\u1e1c\u1e1e\u1e20\u1e22\u1e24\u1e26\u1e28\u1e2a\u1e2c\u1e2e\u1e30\u1e32\u1e34\u1e36\u1e38\u1e3a\u1e3c\u1e3e\u1e40\u1e42\u1e44\u1e46\u1e48\u1e4a\u1e4c\u1e4e\u1e50\u1e52\u1e54\u1e56\u1e58\u1e5a\u1e5c\u1e5e\u1e60\u1e62\u1e64\u1e66\u1e68\u1e6a\u1e6c\u1e6e\u1e70\u1e72\u1e74\u1e76\u1e78\u1e7a\u1e7c\u1e7e\u1e80\u1e82\u1e84\u1e86\u1e88\u1e8a\u1e8c\u1e8e\u1e90\u1e92\u1e94\u1e9e\u1ea0\u1ea2\u1ea4\u1ea6\u1ea8\u1eaa\u1eac\u1eae\u1eb0\u1eb2\u1eb4\u1eb6\u1eb8\u1eba\u1ebc\u1ebe\u1ec0\u1ec2\u1ec4\u1ec6\u1ec8\u1eca\u1ecc\u1ece\u1ed0\u1ed2\u1ed4\u1ed6\u1ed8\u1eda\u1edc\u1ede\u1ee0\u1ee2\u1ee4\u1ee6\u1ee8\u1eea\u1eec\u1eee\u1ef0\u1ef2\u1ef4\u1ef6\u1ef8\u1efa\u1efc\u1efe', | |
'InGreek_Extended' => '\u1f08-\u1f0f\u1f18-\u1f1d\u1f28-\u1f2f\u1f38-\u1f3f\u1f48-\u1f4d\u1f59\u1f5b\u1f5d\u1f5f\u1f68-\u1f6f\u1fb8-\u1fbb\u1fc8-\u1fcb\u1fd8-\u1fdb\u1fe8-\u1fec\u1ff8-\u1ffb', | |
'InLetterlike_Symbols' => '\u2102\u2107\u210b-\u210d\u2110-\u2112\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u2130-\u2133\u213e\u213f\u2145', | |
'InNumber_Forms' => '\u2183', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff21-\uff3a', | |
), | |
'M' => array( | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InCyrillic' => '\u0483-\u0489', | |
'InHebrew' => '\u0591-\u05bd\u05bf\u05c1\u05c2\u05c4\u05c5\u05c7', | |
'InArabic' => '\u0610-\u061a\u064b-\u065e\u0670\u06d6-\u06dc\u06de-\u06e4\u06e7\u06e8\u06ea-\u06ed', | |
'InSyriac' => '\u0711\u0730-\u074a', | |
'InThaana' => '\u07a6-\u07b0', | |
'InDevanagari' => '\u0900-\u0903\u093c\u093e-\u094e\u0951-\u0955\u0962\u0963', | |
'InBengali' => '\u0981-\u0983\u09bc\u09be-\u09c4\u09c7\u09c8\u09cb-\u09cd\u09d7\u09e2\u09e3', | |
'InGurmukhi' => '\u0a01-\u0a03\u0a3c\u0a3e-\u0a42\u0a47\u0a48\u0a4b-\u0a4d\u0a51\u0a70\u0a71\u0a75', | |
'InGujarati' => '\u0a81-\u0a83\u0abc\u0abe-\u0ac5\u0ac7-\u0ac9\u0acb-\u0acd\u0ae2\u0ae3', | |
'InOriya' => '\u0b01-\u0b03\u0b3c\u0b3e-\u0b44\u0b47\u0b48\u0b4b-\u0b4d\u0b56\u0b57\u0b62\u0b63', | |
'InTamil' => '\u0b82\u0bbe-\u0bc2\u0bc6-\u0bc8\u0bca-\u0bcd\u0bd7', | |
'InTelugu' => '\u0c01-\u0c03\u0c3e-\u0c44\u0c46-\u0c48\u0c4a-\u0c4d\u0c55\u0c56\u0c62\u0c63', | |
'InKannada' => '\u0c82\u0c83\u0cbc\u0cbe-\u0cc4\u0cc6-\u0cc8\u0cca-\u0ccd\u0cd5\u0cd6\u0ce2\u0ce3', | |
'InMalayalam' => '\u0d02\u0d03\u0d3e-\u0d44\u0d46-\u0d48\u0d4a-\u0d4d\u0d57\u0d62\u0d63', | |
'InSinhala' => '\u0d82\u0d83\u0dca\u0dcf-\u0dd4\u0dd6\u0dd8-\u0ddf\u0df2\u0df3', | |
'InThai' => '\u0e31\u0e34-\u0e3a\u0e47-\u0e4e', | |
'InLao' => '\u0eb1\u0eb4-\u0eb9\u0ebb\u0ebc\u0ec8-\u0ecd', | |
'InTibetan' => '\u0f18\u0f19\u0f35\u0f37\u0f39\u0f3e\u0f3f\u0f71-\u0f84\u0f86\u0f87\u0f90-\u0f97\u0f99-\u0fbc\u0fc6', | |
'InMyanmar' => '\u102b-\u103e\u1056-\u1059\u105e-\u1060\u1062-\u1064\u1067-\u106d\u1071-\u1074\u1082-\u108d\u108f\u109a-\u109d', | |
'InEthiopic' => '\u135f', | |
'InTagalog' => '\u1712-\u1714', | |
'InHanunoo' => '\u1732-\u1734', | |
'InBuhid' => '\u1752\u1753', | |
'InTagbanwa' => '\u1772\u1773', | |
'InKhmer' => '\u17b6-\u17d3\u17dd', | |
'InMongolian' => '\u180b-\u180d\u18a9', | |
'InLimbu' => '\u1920-\u192b\u1930-\u193b', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20f0', | |
'InCJK_Symbols_and_Punctuation' => '\u302a-\u302f', | |
'InHiragana' => '\u3099\u309a', | |
'InAlphabetic_Presentation_Forms' => '\ufb1e', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe26', | |
), | |
'Mc' => array( | |
'InDevanagari' => '\u0903\u093e-\u0940\u0949-\u094c\u094e', | |
'InBengali' => '\u0982\u0983\u09be-\u09c0\u09c7\u09c8\u09cb\u09cc\u09d7', | |
'InGurmukhi' => '\u0a03\u0a3e-\u0a40', | |
'InGujarati' => '\u0a83\u0abe-\u0ac0\u0ac9\u0acb\u0acc', | |
'InOriya' => '\u0b02\u0b03\u0b3e\u0b40\u0b47\u0b48\u0b4b\u0b4c\u0b57', | |
'InTamil' => '\u0bbe\u0bbf\u0bc1\u0bc2\u0bc6-\u0bc8\u0bca-\u0bcc\u0bd7', | |
'InTelugu' => '\u0c01-\u0c03\u0c41-\u0c44', | |
'InKannada' => '\u0c82\u0c83\u0cbe\u0cc0-\u0cc4\u0cc7\u0cc8\u0cca\u0ccb\u0cd5\u0cd6', | |
'InMalayalam' => '\u0d02\u0d03\u0d3e-\u0d40\u0d46-\u0d48\u0d4a-\u0d4c\u0d57', | |
'InSinhala' => '\u0d82\u0d83\u0dcf-\u0dd1\u0dd8-\u0ddf\u0df2\u0df3', | |
'InTibetan' => '\u0f3e\u0f3f\u0f7f', | |
'InMyanmar' => '\u102b\u102c\u1031\u1038\u103b\u103c\u1056\u1057\u1062-\u1064\u1067-\u106d\u1083\u1084\u1087-\u108c\u108f\u109a-\u109c', | |
'InKhmer' => '\u17b6\u17be-\u17c5\u17c7\u17c8', | |
'InLimbu' => '\u1923-\u1926\u1929-\u192b\u1930\u1931\u1933-\u1938', | |
), | |
'Me' => array( | |
'InCyrillic' => '\u0488\u0489', | |
'InArabic' => '\u06de', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20dd-\u20e0\u20e2-\u20e4', | |
), | |
'Mn' => array( | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InCyrillic' => '\u0483-\u0487', | |
'InHebrew' => '\u0591-\u05bd\u05bf\u05c1\u05c2\u05c4\u05c5\u05c7', | |
'InArabic' => '\u0610-\u061a\u064b-\u065e\u0670\u06d6-\u06dc\u06df-\u06e4\u06e7\u06e8\u06ea-\u06ed', | |
'InSyriac' => '\u0711\u0730-\u074a', | |
'InThaana' => '\u07a6-\u07b0', | |
'InDevanagari' => '\u0900-\u0902\u093c\u0941-\u0948\u094d\u0951-\u0955\u0962\u0963', | |
'InBengali' => '\u0981\u09bc\u09c1-\u09c4\u09cd\u09e2\u09e3', | |
'InGurmukhi' => '\u0a01\u0a02\u0a3c\u0a41\u0a42\u0a47\u0a48\u0a4b-\u0a4d\u0a51\u0a70\u0a71\u0a75', | |
'InGujarati' => '\u0a81\u0a82\u0abc\u0ac1-\u0ac5\u0ac7\u0ac8\u0acd\u0ae2\u0ae3', | |
'InOriya' => '\u0b01\u0b3c\u0b3f\u0b41-\u0b44\u0b4d\u0b56\u0b62\u0b63', | |
'InTamil' => '\u0b82\u0bc0\u0bcd', | |
'InTelugu' => '\u0c3e-\u0c40\u0c46-\u0c48\u0c4a-\u0c4d\u0c55\u0c56\u0c62\u0c63', | |
'InKannada' => '\u0cbc\u0cbf\u0cc6\u0ccc\u0ccd\u0ce2\u0ce3', | |
'InMalayalam' => '\u0d41-\u0d44\u0d4d\u0d62\u0d63', | |
'InSinhala' => '\u0dca\u0dd2-\u0dd4\u0dd6', | |
'InThai' => '\u0e31\u0e34-\u0e3a\u0e47-\u0e4e', | |
'InLao' => '\u0eb1\u0eb4-\u0eb9\u0ebb\u0ebc\u0ec8-\u0ecd', | |
'InTibetan' => '\u0f18\u0f19\u0f35\u0f37\u0f39\u0f71-\u0f7e\u0f80-\u0f84\u0f86\u0f87\u0f90-\u0f97\u0f99-\u0fbc\u0fc6', | |
'InMyanmar' => '\u102d-\u1030\u1032-\u1037\u1039\u103a\u103d\u103e\u1058\u1059\u105e-\u1060\u1071-\u1074\u1082\u1085\u1086\u108d\u109d', | |
'InEthiopic' => '\u135f', | |
'InTagalog' => '\u1712-\u1714', | |
'InHanunoo' => '\u1732-\u1734', | |
'InBuhid' => '\u1752\u1753', | |
'InTagbanwa' => '\u1772\u1773', | |
'InKhmer' => '\u17b7-\u17bd\u17c6\u17c9-\u17d3\u17dd', | |
'InMongolian' => '\u180b-\u180d\u18a9', | |
'InLimbu' => '\u1920-\u1922\u1927\u1928\u1932\u1939-\u193b', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20dc\u20e1\u20e5-\u20f0', | |
'InCJK_Symbols_and_Punctuation' => '\u302a-\u302f', | |
'InHiragana' => '\u3099\u309a', | |
'InAlphabetic_Presentation_Forms' => '\ufb1e', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe26', | |
), | |
'N' => array( | |
'InBasic_Latin' => '\u0030-\u0039', | |
'InLatin-1_Supplement' => '\u00b2\u00b3\u00b9\u00bc-\u00be', | |
'InArabic' => '\u0660-\u0669\u06f0-\u06f9', | |
'InDevanagari' => '\u0966-\u096f', | |
'InBengali' => '\u09e6-\u09ef\u09f4-\u09f9', | |
'InGurmukhi' => '\u0a66-\u0a6f', | |
'InGujarati' => '\u0ae6-\u0aef', | |
'InOriya' => '\u0b66-\u0b6f', | |
'InTamil' => '\u0be6-\u0bf2', | |
'InTelugu' => '\u0c66-\u0c6f\u0c78-\u0c7e', | |
'InKannada' => '\u0ce6-\u0cef', | |
'InMalayalam' => '\u0d66-\u0d75', | |
'InThai' => '\u0e50-\u0e59', | |
'InLao' => '\u0ed0-\u0ed9', | |
'InTibetan' => '\u0f20-\u0f33', | |
'InMyanmar' => '\u1040-\u1049\u1090-\u1099', | |
'InEthiopic' => '\u1369-\u137c', | |
'InRunic' => '\u16ee-\u16f0', | |
'InKhmer' => '\u17e0-\u17e9\u17f0-\u17f9', | |
'InMongolian' => '\u1810-\u1819', | |
'InLimbu' => '\u1946-\u194f', | |
'InSuperscripts_and_Subscripts' => '\u2070\u2074-\u2079\u2080-\u2089', | |
'InNumber_Forms' => '\u2150-\u2182\u2185-\u2189', | |
'InEnclosed_Alphanumerics' => '\u2460-\u249b\u24ea-\u24ff', | |
'InDingbats' => '\u2776-\u2793', | |
'InCJK_Symbols_and_Punctuation' => '\u3007\u3021-\u3029\u3038-\u303a', | |
'InKanbun' => '\u3192-\u3195', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3220-\u3229\u3251-\u325f\u3280-\u3289\u32b1-\u32bf', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff10-\uff19', | |
), | |
'Nd' => array( | |
'InBasic_Latin' => '\u0030-\u0039', | |
'InArabic' => '\u0660-\u0669\u06f0-\u06f9', | |
'InDevanagari' => '\u0966-\u096f', | |
'InBengali' => '\u09e6-\u09ef', | |
'InGurmukhi' => '\u0a66-\u0a6f', | |
'InGujarati' => '\u0ae6-\u0aef', | |
'InOriya' => '\u0b66-\u0b6f', | |
'InTamil' => '\u0be6-\u0bef', | |
'InTelugu' => '\u0c66-\u0c6f', | |
'InKannada' => '\u0ce6-\u0cef', | |
'InMalayalam' => '\u0d66-\u0d6f', | |
'InThai' => '\u0e50-\u0e59', | |
'InLao' => '\u0ed0-\u0ed9', | |
'InTibetan' => '\u0f20-\u0f29', | |
'InMyanmar' => '\u1040-\u1049\u1090-\u1099', | |
'InKhmer' => '\u17e0-\u17e9', | |
'InMongolian' => '\u1810-\u1819', | |
'InLimbu' => '\u1946-\u194f', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff10-\uff19', | |
), | |
'Nl' => array( | |
'InRunic' => '\u16ee-\u16f0', | |
'InNumber_Forms' => '\u2160-\u2182\u2185-\u2188', | |
'InCJK_Symbols_and_Punctuation' => '\u3007\u3021-\u3029\u3038-\u303a', | |
), | |
'No' => array( | |
'InLatin-1_Supplement' => '\u00b2\u00b3\u00b9\u00bc-\u00be', | |
'InBengali' => '\u09f4-\u09f9', | |
'InTamil' => '\u0bf0-\u0bf2', | |
'InTelugu' => '\u0c78-\u0c7e', | |
'InMalayalam' => '\u0d70-\u0d75', | |
'InTibetan' => '\u0f2a-\u0f33', | |
'InEthiopic' => '\u1369-\u137c', | |
'InKhmer' => '\u17f0-\u17f9', | |
'InSuperscripts_and_Subscripts' => '\u2070\u2074-\u2079\u2080-\u2089', | |
'InNumber_Forms' => '\u2150-\u215f\u2189', | |
'InEnclosed_Alphanumerics' => '\u2460-\u249b\u24ea-\u24ff', | |
'InDingbats' => '\u2776-\u2793', | |
'InKanbun' => '\u3192-\u3195', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3220-\u3229\u3251-\u325f\u3280-\u3289\u32b1-\u32bf', | |
), | |
'P' => array( | |
'InBasic_Latin' => '\u0021-\u0023\u0025-\u002a\u002c-\u002f\u003a\u003b\u003f\u0040\u005b-\u005d\u005f\u007b\u007d', | |
'InLatin-1_Supplement' => '\u00a1\u00ab\u00b7\u00bb\u00bf', | |
'InGreek_and_Coptic' => '\u037e\u0387', | |
'InArmenian' => '\u055a-\u055f\u0589\u058a', | |
'InHebrew' => '\u05be\u05c0\u05c3\u05c6\u05f3\u05f4', | |
'InArabic' => '\u0609\u060a\u060c\u060d\u061b\u061e\u061f\u066a-\u066d\u06d4', | |
'InSyriac' => '\u0700-\u070d', | |
'InDevanagari' => '\u0964\u0965\u0970', | |
'InSinhala' => '\u0df4', | |
'InThai' => '\u0e4f\u0e5a\u0e5b', | |
'InTibetan' => '\u0f04-\u0f12\u0f3a-\u0f3d\u0f85\u0fd0-\u0fd4', | |
'InMyanmar' => '\u104a-\u104f', | |
'InGeorgian' => '\u10fb', | |
'InEthiopic' => '\u1361-\u1368', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400\u166d\u166e', | |
'InOgham' => '\u169b\u169c', | |
'InRunic' => '\u16eb-\u16ed', | |
'InHanunoo' => '\u1735\u1736', | |
'InKhmer' => '\u17d4-\u17d6\u17d8-\u17da', | |
'InMongolian' => '\u1800-\u180a', | |
'InLimbu' => '\u1944\u1945', | |
'InGeneral_Punctuation' => '\u2010-\u2027\u2030-\u2043\u2045-\u2051\u2053-\u205e', | |
'InSuperscripts_and_Subscripts' => '\u207d\u207e\u208d\u208e', | |
'InMiscellaneous_Technical' => '\u2329\u232a', | |
'InDingbats' => '\u2768-\u2775', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c5\u27c6\u27e6-\u27ef', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2983-\u2998\u29d8-\u29db\u29fc\u29fd', | |
'InCJK_Symbols_and_Punctuation' => '\u3001-\u3003\u3008-\u3011\u3014-\u301f\u3030\u303d', | |
'InKatakana' => '\u30a0\u30fb', | |
'InArabic_Presentation_Forms-A' => '\ufd3e\ufd3f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe52\ufe54-\ufe61\ufe63\ufe68\ufe6a\ufe6b', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff01-\uff03\uff05-\uff0a\uff0c-\uff0f\uff1a\uff1b\uff1f\uff20\uff3b-\uff3d\uff3f\uff5b\uff5d\uff5f-\uff65', | |
), | |
'Pc' => array( | |
'InBasic_Latin' => '\u005f', | |
'InGeneral_Punctuation' => '\u203f\u2040\u2054', | |
'InCJK_Compatibility_Forms' => '\ufe33\ufe34\ufe4d-\ufe4f', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff3f', | |
), | |
'Pd' => array( | |
'InBasic_Latin' => '\u002d', | |
'InArmenian' => '\u058a', | |
'InHebrew' => '\u05be', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400', | |
'InMongolian' => '\u1806', | |
'InGeneral_Punctuation' => '\u2010-\u2015', | |
'InCJK_Symbols_and_Punctuation' => '\u301c\u3030', | |
'InKatakana' => '\u30a0', | |
'InCJK_Compatibility_Forms' => '\ufe31\ufe32', | |
'InSmall_Form_Variants' => '\ufe58\ufe63', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff0d', | |
), | |
'Pe' => array( | |
'InBasic_Latin' => '\u0029\u005d\u007d', | |
'InTibetan' => '\u0f3b\u0f3d', | |
'InOgham' => '\u169c', | |
'InGeneral_Punctuation' => '\u2046', | |
'InSuperscripts_and_Subscripts' => '\u207e\u208e', | |
'InMiscellaneous_Technical' => '\u232a', | |
'InDingbats' => '\u2769\u276b\u276d\u276f\u2771\u2773\u2775', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c6\u27e7\u27e9\u27eb\u27ed\u27ef', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2984\u2986\u2988\u298a\u298c\u298e\u2990\u2992\u2994\u2996\u2998\u29d9\u29db\u29fd', | |
'InCJK_Symbols_and_Punctuation' => '\u3009\u300b\u300d\u300f\u3011\u3015\u3017\u3019\u301b\u301e\u301f', | |
'InArabic_Presentation_Forms-A' => '\ufd3f', | |
'InCJK_Compatibility_Forms' => '\ufe36\ufe38\ufe3a\ufe3c\ufe3e\ufe40\ufe42\ufe44\ufe48', | |
'InSmall_Form_Variants' => '\ufe5a\ufe5c\ufe5e', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff09\uff3d\uff5d\uff60\uff63', | |
), | |
'Pf' => array( | |
'InLatin-1_Supplement' => '\u00bb', | |
'InGeneral_Punctuation' => '\u2019\u201d\u203a', | |
), | |
'Pi' => array( | |
'InLatin-1_Supplement' => '\u00ab', | |
'InGeneral_Punctuation' => '\u2018\u201b\u201c\u201f\u2039', | |
), | |
'Po' => array( | |
'InBasic_Latin' => '\u0021-\u0023\u0025-\u0027\u002a\u002c\u002e\u002f\u003a\u003b\u003f\u0040\u005c', | |
'InLatin-1_Supplement' => '\u00a1\u00b7\u00bf', | |
'InGreek_and_Coptic' => '\u037e\u0387', | |
'InArmenian' => '\u055a-\u055f\u0589', | |
'InHebrew' => '\u05c0\u05c3\u05c6\u05f3\u05f4', | |
'InArabic' => '\u0609\u060a\u060c\u060d\u061b\u061e\u061f\u066a-\u066d\u06d4', | |
'InSyriac' => '\u0700-\u070d', | |
'InDevanagari' => '\u0964\u0965\u0970', | |
'InSinhala' => '\u0df4', | |
'InThai' => '\u0e4f\u0e5a\u0e5b', | |
'InTibetan' => '\u0f04-\u0f12\u0f85\u0fd0-\u0fd4', | |
'InMyanmar' => '\u104a-\u104f', | |
'InGeorgian' => '\u10fb', | |
'InEthiopic' => '\u1361-\u1368', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u166d\u166e', | |
'InRunic' => '\u16eb-\u16ed', | |
'InHanunoo' => '\u1735\u1736', | |
'InKhmer' => '\u17d4-\u17d6\u17d8-\u17da', | |
'InMongolian' => '\u1800-\u1805\u1807-\u180a', | |
'InLimbu' => '\u1944\u1945', | |
'InGeneral_Punctuation' => '\u2016\u2017\u2020-\u2027\u2030-\u2038\u203b-\u203e\u2041-\u2043\u2047-\u2051\u2053\u2055-\u205e', | |
'InCJK_Symbols_and_Punctuation' => '\u3001-\u3003\u303d', | |
'InKatakana' => '\u30fb', | |
'InCJK_Compatibility_Forms' => '\ufe30\ufe45\ufe46\ufe49-\ufe4c', | |
'InSmall_Form_Variants' => '\ufe50-\ufe52\ufe54-\ufe57\ufe5f-\ufe61\ufe68\ufe6a\ufe6b', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff01-\uff03\uff05-\uff07\uff0a\uff0c\uff0e\uff0f\uff1a\uff1b\uff1f\uff20\uff3c\uff61\uff64\uff65', | |
), | |
'Ps' => array( | |
'InBasic_Latin' => '\u0028\u005b\u007b', | |
'InTibetan' => '\u0f3a\u0f3c', | |
'InOgham' => '\u169b', | |
'InGeneral_Punctuation' => '\u201a\u201e\u2045', | |
'InSuperscripts_and_Subscripts' => '\u207d\u208d', | |
'InMiscellaneous_Technical' => '\u2329', | |
'InDingbats' => '\u2768\u276a\u276c\u276e\u2770\u2772\u2774', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c5\u27e6\u27e8\u27ea\u27ec\u27ee', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2983\u2985\u2987\u2989\u298b\u298d\u298f\u2991\u2993\u2995\u2997\u29d8\u29da\u29fc', | |
'InCJK_Symbols_and_Punctuation' => '\u3008\u300a\u300c\u300e\u3010\u3014\u3016\u3018\u301a\u301d', | |
'InArabic_Presentation_Forms-A' => '\ufd3e', | |
'InCJK_Compatibility_Forms' => '\ufe35\ufe37\ufe39\ufe3b\ufe3d\ufe3f\ufe41\ufe43\ufe47', | |
'InSmall_Form_Variants' => '\ufe59\ufe5b\ufe5d', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff08\uff3b\uff5b\uff5f\uff62', | |
), | |
'S' => array( | |
'InBasic_Latin' => '\u0024\u002b\u003c-\u003e\u005e\u0060\u007c\u007e', | |
'InLatin-1_Supplement' => '\u00a2-\u00a9\u00ac\u00ae-\u00b1\u00b4\u00b6\u00b8\u00d7\u00f7', | |
'InSpacing_Modifier_Letters' => '\u02c2-\u02c5\u02d2-\u02df\u02e5-\u02eb\u02ed\u02ef-\u02ff', | |
'InGreek_and_Coptic' => '\u0375\u0384\u0385\u03f6', | |
'InCyrillic' => '\u0482', | |
'InArabic' => '\u0606-\u0608\u060b\u060e\u060f\u06e9\u06fd\u06fe', | |
'InBengali' => '\u09f2\u09f3\u09fa\u09fb', | |
'InGujarati' => '\u0af1', | |
'InOriya' => '\u0b70', | |
'InTamil' => '\u0bf3-\u0bfa', | |
'InTelugu' => '\u0c7f', | |
'InKannada' => '\u0cf1\u0cf2', | |
'InMalayalam' => '\u0d79', | |
'InThai' => '\u0e3f', | |
'InTibetan' => '\u0f01-\u0f03\u0f13-\u0f17\u0f1a-\u0f1f\u0f34\u0f36\u0f38\u0fbe-\u0fc5\u0fc7-\u0fcc\u0fce\u0fcf\u0fd5-\u0fd8', | |
'InMyanmar' => '\u109e\u109f', | |
'InEthiopic' => '\u1360', | |
'InKhmer' => '\u17db', | |
'InLimbu' => '\u1940', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InGreek_Extended' => '\u1fbd\u1fbf-\u1fc1\u1fcd-\u1fcf\u1fdd-\u1fdf\u1fed-\u1fef\u1ffd\u1ffe', | |
'InGeneral_Punctuation' => '\u2044\u2052', | |
'InSuperscripts_and_Subscripts' => '\u207a-\u207c\u208a-\u208c', | |
'InCurrency_Symbols' => '\u20a0-\u20b8', | |
'InLetterlike_Symbols' => '\u2100\u2101\u2103-\u2106\u2108\u2109\u2114\u2116-\u2118\u211e-\u2123\u2125\u2127\u2129\u212e\u213a\u213b\u2140-\u2144\u214a-\u214d\u214f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u2328\u232b-\u23e8', | |
'InControl_Pictures' => '\u2400-\u2426', | |
'InOptical_Character_Recognition' => '\u2440-\u244a', | |
'InEnclosed_Alphanumerics' => '\u249c-\u24e9', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26cd\u26cf-\u26e1\u26e3\u26e8-\u26ff', | |
'InDingbats' => '\u2701-\u2704\u2706-\u2709\u270c-\u2727\u2729-\u274b\u274d\u274f-\u2752\u2756-\u275e\u2761-\u2767\u2794\u2798-\u27af\u27b1-\u27be', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27c4\u27c7-\u27ca\u27cc\u27d0-\u27e5', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u2982\u2999-\u29d7\u29dc-\u29fb\u29fe\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2b4c\u2b50-\u2b59', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2e99\u2e9b-\u2ef3', | |
'InKangxi_Radicals' => '\u2f00-\u2fd5', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2ffb', | |
'InCJK_Symbols_and_Punctuation' => '\u3004\u3012\u3013\u3020\u3036\u3037\u303e\u303f', | |
'InHiragana' => '\u309b\u309c', | |
'InKanbun' => '\u3190\u3191\u3196-\u319f', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u321e\u322a-\u3250\u3260-\u327f\u328a-\u32b0\u32c0-\u32fe', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InYi_Radicals' => '\ua490-\ua4c6', | |
'InAlphabetic_Presentation_Forms' => '\ufb29', | |
'InArabic_Presentation_Forms-A' => '\ufdfc\ufdfd', | |
'InSmall_Form_Variants' => '\ufe62\ufe64-\ufe66\ufe69', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff04\uff0b\uff1c-\uff1e\uff3e\uff40\uff5c\uff5e\uffe0-\uffe6\uffe8-\uffee', | |
'InSpecials' => '\ufffc\ufffd', | |
), | |
'Sc' => array( | |
'InBasic_Latin' => '\u0024', | |
'InLatin-1_Supplement' => '\u00a2-\u00a5', | |
'InArabic' => '\u060b', | |
'InBengali' => '\u09f2\u09f3\u09fb', | |
'InGujarati' => '\u0af1', | |
'InTamil' => '\u0bf9', | |
'InThai' => '\u0e3f', | |
'InKhmer' => '\u17db', | |
'InCurrency_Symbols' => '\u20a0-\u20b8', | |
'InArabic_Presentation_Forms-A' => '\ufdfc', | |
'InSmall_Form_Variants' => '\ufe69', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff04\uffe0\uffe1\uffe5\uffe6', | |
), | |
'Sk' => array( | |
'InBasic_Latin' => '\u005e\u0060', | |
'InLatin-1_Supplement' => '\u00a8\u00af\u00b4\u00b8', | |
'InSpacing_Modifier_Letters' => '\u02c2-\u02c5\u02d2-\u02df\u02e5-\u02eb\u02ed\u02ef-\u02ff', | |
'InGreek_and_Coptic' => '\u0375\u0384\u0385', | |
'InGreek_Extended' => '\u1fbd\u1fbf-\u1fc1\u1fcd-\u1fcf\u1fdd-\u1fdf\u1fed-\u1fef\u1ffd\u1ffe', | |
'InHiragana' => '\u309b\u309c', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff3e\uff40\uffe3', | |
), | |
'Sm' => array( | |
'InBasic_Latin' => '\u002b\u003c-\u003e\u007c\u007e', | |
'InLatin-1_Supplement' => '\u00ac\u00b1\u00d7\u00f7', | |
'InGreek_and_Coptic' => '\u03f6', | |
'InArabic' => '\u0606-\u0608', | |
'InGeneral_Punctuation' => '\u2044\u2052', | |
'InSuperscripts_and_Subscripts' => '\u207a-\u207c\u208a-\u208c', | |
'InLetterlike_Symbols' => '\u2140-\u2144\u214b', | |
'InArrows' => '\u2190-\u2194\u219a\u219b\u21a0\u21a3\u21a6\u21ae\u21ce\u21cf\u21d2\u21d4\u21f4-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2308-\u230b\u2320\u2321\u237c\u239b-\u23b3\u23dc-\u23e1', | |
'InGeometric_Shapes' => '\u25b7\u25c1\u25f8-\u25ff', | |
'InMiscellaneous_Symbols' => '\u266f', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27c4\u27c7-\u27ca\u27cc\u27d0-\u27e5', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u2982\u2999-\u29d7\u29dc-\u29fb\u29fe\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b30-\u2b44\u2b47-\u2b4c', | |
'InAlphabetic_Presentation_Forms' => '\ufb29', | |
'InSmall_Form_Variants' => '\ufe62\ufe64-\ufe66', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff0b\uff1c-\uff1e\uff5c\uff5e\uffe2\uffe9-\uffec', | |
), | |
'So' => array( | |
'InLatin-1_Supplement' => '\u00a6\u00a7\u00a9\u00ae\u00b0\u00b6', | |
'InCyrillic' => '\u0482', | |
'InArabic' => '\u060e\u060f\u06e9\u06fd\u06fe', | |
'InBengali' => '\u09fa', | |
'InOriya' => '\u0b70', | |
'InTamil' => '\u0bf3-\u0bf8\u0bfa', | |
'InTelugu' => '\u0c7f', | |
'InKannada' => '\u0cf1\u0cf2', | |
'InMalayalam' => '\u0d79', | |
'InTibetan' => '\u0f01-\u0f03\u0f13-\u0f17\u0f1a-\u0f1f\u0f34\u0f36\u0f38\u0fbe-\u0fc5\u0fc7-\u0fcc\u0fce\u0fcf\u0fd5-\u0fd8', | |
'InMyanmar' => '\u109e\u109f', | |
'InEthiopic' => '\u1360', | |
'InLimbu' => '\u1940', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InLetterlike_Symbols' => '\u2100\u2101\u2103-\u2106\u2108\u2109\u2114\u2116-\u2118\u211e-\u2123\u2125\u2127\u2129\u212e\u213a\u213b\u214a\u214c\u214d\u214f', | |
'InArrows' => '\u2195-\u2199\u219c-\u219f\u21a1\u21a2\u21a4\u21a5\u21a7-\u21ad\u21af-\u21cd\u21d0\u21d1\u21d3\u21d5-\u21f3', | |
'InMiscellaneous_Technical' => '\u2300-\u2307\u230c-\u231f\u2322-\u2328\u232b-\u237b\u237d-\u239a\u23b4-\u23db\u23e2-\u23e8', | |
'InControl_Pictures' => '\u2400-\u2426', | |
'InOptical_Character_Recognition' => '\u2440-\u244a', | |
'InEnclosed_Alphanumerics' => '\u249c-\u24e9', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25b6\u25b8-\u25c0\u25c2-\u25f7', | |
'InMiscellaneous_Symbols' => '\u2600-\u266e\u2670-\u26cd\u26cf-\u26e1\u26e3\u26e8-\u26ff', | |
'InDingbats' => '\u2701-\u2704\u2706-\u2709\u270c-\u2727\u2729-\u274b\u274d\u274f-\u2752\u2756-\u275e\u2761-\u2767\u2794\u2798-\u27af\u27b1-\u27be', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2b2f\u2b45\u2b46\u2b50-\u2b59', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2e99\u2e9b-\u2ef3', | |
'InKangxi_Radicals' => '\u2f00-\u2fd5', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2ffb', | |
'InCJK_Symbols_and_Punctuation' => '\u3004\u3012\u3013\u3020\u3036\u3037\u303e\u303f', | |
'InKanbun' => '\u3190\u3191\u3196-\u319f', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u321e\u322a-\u3250\u3260-\u327f\u328a-\u32b0\u32c0-\u32fe', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InYi_Radicals' => '\ua490-\ua4c6', | |
'InArabic_Presentation_Forms-A' => '\ufdfd', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uffe4\uffe8\uffed\uffee', | |
'InSpecials' => '\ufffc\ufffd', | |
), | |
'Z' => array( | |
'InBasic_Latin' => '\u0020', | |
'InLatin-1_Supplement' => '\u00a0', | |
'InOgham' => '\u1680', | |
'InMongolian' => '\u180e', | |
'InGeneral_Punctuation' => '\u2000-\u200a\u2028\u2029\u202f\u205f', | |
'InCJK_Symbols_and_Punctuation' => '\u3000', | |
), | |
'Zl' => array( | |
'InGeneral_Punctuation' => '\u2028', | |
), | |
'Zp' => array( | |
'InGeneral_Punctuation' => '\u2029', | |
), | |
'Zs' => array( | |
'InBasic_Latin' => '\u0020', | |
'InLatin-1_Supplement' => '\u00a0', | |
'InOgham' => '\u1680', | |
'InMongolian' => '\u180e', | |
'InGeneral_Punctuation' => '\u2000-\u200a\u202f\u205f', | |
'InCJK_Symbols_and_Punctuation' => '\u3000', | |
), | |
'X' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u0482\u048a-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590\u05be\u05c0\u05c3\u05c6\u05c8-\u05ff', | |
'InArabic' => '\u0600-\u060f\u061b-\u064a\u065f-\u066f\u0671-\u06d5\u06dd\u06e5\u06e6\u06e9\u06ee-\u06ff', | |
'InSyriac' => '\u0700-\u0710\u0712-\u072f\u074b-\u074f', | |
'InThaana' => '\u0780-\u07a5\u07b1-\u07bf', | |
'InDevanagari' => '\u0904-\u093b\u093d\u094f\u0950\u0956-\u0961\u0964-\u097f', | |
'InBengali' => '\u0980\u0984-\u09bb\u09bd\u09c5\u09c6\u09c9\u09ca\u09ce-\u09d6\u09d8-\u09e1\u09e4-\u09ff', | |
'InGurmukhi' => '\u0a00\u0a04-\u0a3b\u0a3d\u0a43-\u0a46\u0a49\u0a4a\u0a4e-\u0a50\u0a52-\u0a6f\u0a72-\u0a74\u0a76-\u0a7f', | |
'InGujarati' => '\u0a80\u0a84-\u0abb\u0abd\u0ac6\u0aca\u0ace-\u0ae1\u0ae4-\u0aff', | |
'InOriya' => '\u0b00\u0b04-\u0b3b\u0b3d\u0b45\u0b46\u0b49\u0b4a\u0b4e-\u0b55\u0b58-\u0b61\u0b64-\u0b7f', | |
'InTamil' => '\u0b80\u0b81\u0b83-\u0bbd\u0bc3-\u0bc5\u0bc9\u0bce-\u0bd6\u0bd8-\u0bff', | |
'InTelugu' => '\u0c00\u0c04-\u0c3d\u0c45\u0c49\u0c4e-\u0c54\u0c57-\u0c61\u0c64-\u0c7f', | |
'InKannada' => '\u0c80\u0c81\u0c84-\u0cbb\u0cbd\u0cc5\u0cc9\u0cce-\u0cd4\u0cd7-\u0ce1\u0ce4-\u0cff', | |
'InMalayalam' => '\u0d00\u0d01\u0d04-\u0d3d\u0d45\u0d49\u0d4e-\u0d56\u0d58-\u0d61\u0d64-\u0d7f', | |
'InSinhala' => '\u0d80\u0d81\u0d84-\u0dc9\u0dcb-\u0dce\u0dd5\u0dd7\u0de0-\u0df1\u0df4-\u0dff', | |
'InThai' => '\u0e00-\u0e30\u0e32\u0e33\u0e3b-\u0e46\u0e4f-\u0e7f', | |
'InLao' => '\u0e80-\u0eb0\u0eb2\u0eb3\u0eba\u0ebd-\u0ec7\u0ece-\u0eff', | |
'InTibetan' => '\u0f00-\u0f17\u0f1a-\u0f34\u0f36\u0f38\u0f3a-\u0f3d\u0f40-\u0f70\u0f85\u0f88-\u0f8f\u0f98\u0fbd-\u0fc5\u0fc7-\u0fff', | |
'InMyanmar' => '\u1000-\u102a\u103f-\u1055\u105a-\u105d\u1061\u1065\u1066\u106e-\u1070\u1075-\u1081\u108e\u1090-\u1099\u109e\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u135e\u1360-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u1711\u1715-\u171f', | |
'InHanunoo' => '\u1720-\u1731\u1735-\u173f', | |
'InBuhid' => '\u1740-\u1751\u1754-\u175f', | |
'InTagbanwa' => '\u1760-\u1771\u1774-\u177f', | |
'InKhmer' => '\u1780-\u17b5\u17d4-\u17dc\u17de-\u17ff', | |
'InMongolian' => '\u1800-\u180a\u180e-\u18a8\u18aa-\u18af', | |
'InLimbu' => '\u1900-\u191f\u192c-\u192f\u193c-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20f1-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3029\u3030-\u303f', | |
'InHiragana' => '\u3040-\u3098\u309b-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb1d\ufb1f-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InCombining_Half_Marks' => '\ufe27-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
) | |
); | |
public static $notEqualRules = array( | |
'C' => array( | |
'InBasic_Latin' => '\u0020-\u007e', | |
'InLatin-1_Supplement' => '\u00a0-\u00ac\u00ae-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u0377\u037a-\u037e\u0384-\u038a\u038c\u038e-\u03a1\u03a3-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u0525', | |
'InArmenian' => '\u0531-\u0556\u0559-\u055f\u0561-\u0587\u0589\u058a', | |
'InHebrew' => '\u0591-\u05c7\u05d0-\u05ea\u05f0-\u05f4', | |
'InArabic' => '\u0606-\u061b\u061e\u061f\u0621-\u065e\u0660-\u06dc\u06de-\u06ff', | |
'InSyriac' => '\u0700-\u070d\u0710-\u074a\u074d-\u074f', | |
'InThaana' => '\u0780-\u07b1', | |
'InDevanagari' => '\u0900-\u0939\u093c-\u094e\u0950-\u0955\u0958-\u0972\u0979-\u097f', | |
'InBengali' => '\u0981-\u0983\u0985-\u098c\u098f\u0990\u0993-\u09a8\u09aa-\u09b0\u09b2\u09b6-\u09b9\u09bc-\u09c4\u09c7\u09c8\u09cb-\u09ce\u09d7\u09dc\u09dd\u09df-\u09e3\u09e6-\u09fb', | |
'InGurmukhi' => '\u0a01-\u0a03\u0a05-\u0a0a\u0a0f\u0a10\u0a13-\u0a28\u0a2a-\u0a30\u0a32\u0a33\u0a35\u0a36\u0a38\u0a39\u0a3c\u0a3e-\u0a42\u0a47\u0a48\u0a4b-\u0a4d\u0a51\u0a59-\u0a5c\u0a5e\u0a66-\u0a75', | |
'InGujarati' => '\u0a81-\u0a83\u0a85-\u0a8d\u0a8f-\u0a91\u0a93-\u0aa8\u0aaa-\u0ab0\u0ab2\u0ab3\u0ab5-\u0ab9\u0abc-\u0ac5\u0ac7-\u0ac9\u0acb-\u0acd\u0ad0\u0ae0-\u0ae3\u0ae6-\u0aef\u0af1', | |
'InOriya' => '\u0b01-\u0b03\u0b05-\u0b0c\u0b0f\u0b10\u0b13-\u0b28\u0b2a-\u0b30\u0b32\u0b33\u0b35-\u0b39\u0b3c-\u0b44\u0b47\u0b48\u0b4b-\u0b4d\u0b56\u0b57\u0b5c\u0b5d\u0b5f-\u0b63\u0b66-\u0b71', | |
'InTamil' => '\u0b82\u0b83\u0b85-\u0b8a\u0b8e-\u0b90\u0b92-\u0b95\u0b99\u0b9a\u0b9c\u0b9e\u0b9f\u0ba3\u0ba4\u0ba8-\u0baa\u0bae-\u0bb9\u0bbe-\u0bc2\u0bc6-\u0bc8\u0bca-\u0bcd\u0bd0\u0bd7\u0be6-\u0bfa', | |
'InTelugu' => '\u0c01-\u0c03\u0c05-\u0c0c\u0c0e-\u0c10\u0c12-\u0c28\u0c2a-\u0c33\u0c35-\u0c39\u0c3d-\u0c44\u0c46-\u0c48\u0c4a-\u0c4d\u0c55\u0c56\u0c58\u0c59\u0c60-\u0c63\u0c66-\u0c6f\u0c78-\u0c7f', | |
'InKannada' => '\u0c82\u0c83\u0c85-\u0c8c\u0c8e-\u0c90\u0c92-\u0ca8\u0caa-\u0cb3\u0cb5-\u0cb9\u0cbc-\u0cc4\u0cc6-\u0cc8\u0cca-\u0ccd\u0cd5\u0cd6\u0cde\u0ce0-\u0ce3\u0ce6-\u0cef\u0cf1\u0cf2', | |
'InMalayalam' => '\u0d02\u0d03\u0d05-\u0d0c\u0d0e-\u0d10\u0d12-\u0d28\u0d2a-\u0d39\u0d3d-\u0d44\u0d46-\u0d48\u0d4a-\u0d4d\u0d57\u0d60-\u0d63\u0d66-\u0d75\u0d79-\u0d7f', | |
'InSinhala' => '\u0d82\u0d83\u0d85-\u0d96\u0d9a-\u0db1\u0db3-\u0dbb\u0dbd\u0dc0-\u0dc6\u0dca\u0dcf-\u0dd4\u0dd6\u0dd8-\u0ddf\u0df2-\u0df4', | |
'InThai' => '\u0e01-\u0e3a\u0e3f-\u0e5b', | |
'InLao' => '\u0e81\u0e82\u0e84\u0e87\u0e88\u0e8a\u0e8d\u0e94-\u0e97\u0e99-\u0e9f\u0ea1-\u0ea3\u0ea5\u0ea7\u0eaa\u0eab\u0ead-\u0eb9\u0ebb-\u0ebd\u0ec0-\u0ec4\u0ec6\u0ec8-\u0ecd\u0ed0-\u0ed9\u0edc\u0edd', | |
'InTibetan' => '\u0f00-\u0f47\u0f49-\u0f6c\u0f71-\u0f8b\u0f90-\u0f97\u0f99-\u0fbc\u0fbe-\u0fcc\u0fce-\u0fd8', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10c5\u10d0-\u10fc', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u1248\u124a-\u124d\u1250-\u1256\u1258\u125a-\u125d\u1260-\u1288\u128a-\u128d\u1290-\u12b0\u12b2-\u12b5\u12b8-\u12be\u12c0\u12c2-\u12c5\u12c8-\u12d6\u12d8-\u1310\u1312-\u1315\u1318-\u135a\u135f-\u137c', | |
'InCherokee' => '\u13a0-\u13f4', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169c', | |
'InRunic' => '\u16a0-\u16f0', | |
'InTagalog' => '\u1700-\u170c\u170e-\u1714', | |
'InHanunoo' => '\u1720-\u1736', | |
'InBuhid' => '\u1740-\u1753', | |
'InTagbanwa' => '\u1760-\u176c\u176e-\u1770\u1772\u1773', | |
'InKhmer' => '\u1780-\u17b3\u17b6-\u17dd\u17e0-\u17e9\u17f0-\u17f9', | |
'InMongolian' => '\u1800-\u180e\u1810-\u1819\u1820-\u1877\u1880-\u18aa', | |
'InLimbu' => '\u1900-\u191c\u1920-\u192b\u1930-\u193b\u1940\u1944-\u194f', | |
'InTai_Le' => '\u1950-\u196d\u1970-\u1974', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1f15\u1f18-\u1f1d\u1f20-\u1f45\u1f48-\u1f4d\u1f50-\u1f57\u1f59\u1f5b\u1f5d\u1f5f-\u1f7d\u1f80-\u1fb4\u1fb6-\u1fc4\u1fc6-\u1fd3\u1fd6-\u1fdb\u1fdd-\u1fef\u1ff2-\u1ff4\u1ff6-\u1ffe', | |
'InGeneral_Punctuation' => '\u2000-\u200a\u2010-\u2029\u202f-\u205f', | |
'InSuperscripts_and_Subscripts' => '\u2070\u2071\u2074-\u208e\u2090-\u2094', | |
'InCurrency_Symbols' => '\u20a0-\u20b8', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20f0', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u2189', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23e8', | |
'InControl_Pictures' => '\u2400-\u2426', | |
'InOptical_Character_Recognition' => '\u2440-\u244a', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26cd\u26cf-\u26e1\u26e3\u26e8-\u26ff', | |
'InDingbats' => '\u2701-\u2704\u2706-\u2709\u270c-\u2727\u2729-\u274b\u274d\u274f-\u2752\u2756-\u275e\u2761-\u2794\u2798-\u27af\u27b1-\u27be', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ca\u27cc\u27d0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2b4c\u2b50-\u2b59', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2e99\u2e9b-\u2ef3', | |
'InKangxi_Radicals' => '\u2f00-\u2fd5', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2ffb', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3041-\u3096\u3099-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3105-\u312d', | |
'InHangul_Compatibility_Jamo' => '\u3131-\u318e', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31b7', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u321e\u3220-\u32fe', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4db5', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fcb', | |
'InYi_Syllables' => '\ua000-\ua48c', | |
'InYi_Radicals' => '\ua490-\ua4c6', | |
'InHangul_Syllables' => '\uac00-\ud7a3', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufa2d\ufa30-\ufa6d\ufa70-\ufad9', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb06\ufb13-\ufb17\ufb1d-\ufb36\ufb38-\ufb3c\ufb3e\ufb40\ufb41\ufb43\ufb44\ufb46-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufbb1\ufbd3-\ufd3f\ufd50-\ufd8f\ufd92-\ufdc7\ufdf0-\ufdfd', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe26', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe52\ufe54-\ufe66\ufe68-\ufe6b', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufe74\ufe76-\ufefc', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff01-\uffbe\uffc2-\uffc7\uffca-\uffcf\uffd2-\uffd7\uffda-\uffdc\uffe0-\uffe6\uffe8-\uffee', | |
'InSpecials' => '\ufffc\ufffd', | |
), | |
'Cc' => array( | |
'InBasic_Latin' => '\u0020-\u007e', | |
'InLatin-1_Supplement' => '\u00a0-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Cf' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ac\u00ae-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0604-\u06dc\u06de-\u06ff', | |
'InSyriac' => '\u0700-\u070e\u0710-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17b3\u17b6-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u200a\u2010-\u2029\u202f-\u205f\u2065-\u2069', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufefe', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\ufff8\ufffc-\uffff', | |
), | |
'Cn' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u0377\u037a-\u037e\u0384-\u038a\u038c\u038e-\u03a1\u03a3-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u0525', | |
'InArmenian' => '\u0531-\u0556\u0559-\u055f\u0561-\u0587\u0589\u058a', | |
'InHebrew' => '\u0591-\u05c7\u05d0-\u05ea\u05f0-\u05f4', | |
'InArabic' => '\u0600-\u0603\u0606-\u061b\u061e\u061f\u0621-\u065e\u0660-\u06ff', | |
'InSyriac' => '\u0700-\u070d\u070f-\u074a\u074d-\u074f', | |
'InThaana' => '\u0780-\u07b1', | |
'InDevanagari' => '\u0900-\u0939\u093c-\u094e\u0950-\u0955\u0958-\u0972\u0979-\u097f', | |
'InBengali' => '\u0981-\u0983\u0985-\u098c\u098f\u0990\u0993-\u09a8\u09aa-\u09b0\u09b2\u09b6-\u09b9\u09bc-\u09c4\u09c7\u09c8\u09cb-\u09ce\u09d7\u09dc\u09dd\u09df-\u09e3\u09e6-\u09fb', | |
'InGurmukhi' => '\u0a01-\u0a03\u0a05-\u0a0a\u0a0f\u0a10\u0a13-\u0a28\u0a2a-\u0a30\u0a32\u0a33\u0a35\u0a36\u0a38\u0a39\u0a3c\u0a3e-\u0a42\u0a47\u0a48\u0a4b-\u0a4d\u0a51\u0a59-\u0a5c\u0a5e\u0a66-\u0a75', | |
'InGujarati' => '\u0a81-\u0a83\u0a85-\u0a8d\u0a8f-\u0a91\u0a93-\u0aa8\u0aaa-\u0ab0\u0ab2\u0ab3\u0ab5-\u0ab9\u0abc-\u0ac5\u0ac7-\u0ac9\u0acb-\u0acd\u0ad0\u0ae0-\u0ae3\u0ae6-\u0aef\u0af1', | |
'InOriya' => '\u0b01-\u0b03\u0b05-\u0b0c\u0b0f\u0b10\u0b13-\u0b28\u0b2a-\u0b30\u0b32\u0b33\u0b35-\u0b39\u0b3c-\u0b44\u0b47\u0b48\u0b4b-\u0b4d\u0b56\u0b57\u0b5c\u0b5d\u0b5f-\u0b63\u0b66-\u0b71', | |
'InTamil' => '\u0b82\u0b83\u0b85-\u0b8a\u0b8e-\u0b90\u0b92-\u0b95\u0b99\u0b9a\u0b9c\u0b9e\u0b9f\u0ba3\u0ba4\u0ba8-\u0baa\u0bae-\u0bb9\u0bbe-\u0bc2\u0bc6-\u0bc8\u0bca-\u0bcd\u0bd0\u0bd7\u0be6-\u0bfa', | |
'InTelugu' => '\u0c01-\u0c03\u0c05-\u0c0c\u0c0e-\u0c10\u0c12-\u0c28\u0c2a-\u0c33\u0c35-\u0c39\u0c3d-\u0c44\u0c46-\u0c48\u0c4a-\u0c4d\u0c55\u0c56\u0c58\u0c59\u0c60-\u0c63\u0c66-\u0c6f\u0c78-\u0c7f', | |
'InKannada' => '\u0c82\u0c83\u0c85-\u0c8c\u0c8e-\u0c90\u0c92-\u0ca8\u0caa-\u0cb3\u0cb5-\u0cb9\u0cbc-\u0cc4\u0cc6-\u0cc8\u0cca-\u0ccd\u0cd5\u0cd6\u0cde\u0ce0-\u0ce3\u0ce6-\u0cef\u0cf1\u0cf2', | |
'InMalayalam' => '\u0d02\u0d03\u0d05-\u0d0c\u0d0e-\u0d10\u0d12-\u0d28\u0d2a-\u0d39\u0d3d-\u0d44\u0d46-\u0d48\u0d4a-\u0d4d\u0d57\u0d60-\u0d63\u0d66-\u0d75\u0d79-\u0d7f', | |
'InSinhala' => '\u0d82\u0d83\u0d85-\u0d96\u0d9a-\u0db1\u0db3-\u0dbb\u0dbd\u0dc0-\u0dc6\u0dca\u0dcf-\u0dd4\u0dd6\u0dd8-\u0ddf\u0df2-\u0df4', | |
'InThai' => '\u0e01-\u0e3a\u0e3f-\u0e5b', | |
'InLao' => '\u0e81\u0e82\u0e84\u0e87\u0e88\u0e8a\u0e8d\u0e94-\u0e97\u0e99-\u0e9f\u0ea1-\u0ea3\u0ea5\u0ea7\u0eaa\u0eab\u0ead-\u0eb9\u0ebb-\u0ebd\u0ec0-\u0ec4\u0ec6\u0ec8-\u0ecd\u0ed0-\u0ed9\u0edc\u0edd', | |
'InTibetan' => '\u0f00-\u0f47\u0f49-\u0f6c\u0f71-\u0f8b\u0f90-\u0f97\u0f99-\u0fbc\u0fbe-\u0fcc\u0fce-\u0fd8', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10c5\u10d0-\u10fc', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u1248\u124a-\u124d\u1250-\u1256\u1258\u125a-\u125d\u1260-\u1288\u128a-\u128d\u1290-\u12b0\u12b2-\u12b5\u12b8-\u12be\u12c0\u12c2-\u12c5\u12c8-\u12d6\u12d8-\u1310\u1312-\u1315\u1318-\u135a\u135f-\u137c', | |
'InCherokee' => '\u13a0-\u13f4', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169c', | |
'InRunic' => '\u16a0-\u16f0', | |
'InTagalog' => '\u1700-\u170c\u170e-\u1714', | |
'InHanunoo' => '\u1720-\u1736', | |
'InBuhid' => '\u1740-\u1753', | |
'InTagbanwa' => '\u1760-\u176c\u176e-\u1770\u1772\u1773', | |
'InKhmer' => '\u1780-\u17dd\u17e0-\u17e9\u17f0-\u17f9', | |
'InMongolian' => '\u1800-\u180e\u1810-\u1819\u1820-\u1877\u1880-\u18aa', | |
'InLimbu' => '\u1900-\u191c\u1920-\u192b\u1930-\u193b\u1940\u1944-\u194f', | |
'InTai_Le' => '\u1950-\u196d\u1970-\u1974', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1f15\u1f18-\u1f1d\u1f20-\u1f45\u1f48-\u1f4d\u1f50-\u1f57\u1f59\u1f5b\u1f5d\u1f5f-\u1f7d\u1f80-\u1fb4\u1fb6-\u1fc4\u1fc6-\u1fd3\u1fd6-\u1fdb\u1fdd-\u1fef\u1ff2-\u1ff4\u1ff6-\u1ffe', | |
'InGeneral_Punctuation' => '\u2000-\u2064\u206a-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070\u2071\u2074-\u208e\u2090-\u2094', | |
'InCurrency_Symbols' => '\u20a0-\u20b8', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20f0', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u2189', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23e8', | |
'InControl_Pictures' => '\u2400-\u2426', | |
'InOptical_Character_Recognition' => '\u2440-\u244a', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26cd\u26cf-\u26e1\u26e3\u26e8-\u26ff', | |
'InDingbats' => '\u2701-\u2704\u2706-\u2709\u270c-\u2727\u2729-\u274b\u274d\u274f-\u2752\u2756-\u275e\u2761-\u2794\u2798-\u27af\u27b1-\u27be', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ca\u27cc\u27d0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2b4c\u2b50-\u2b59', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2e99\u2e9b-\u2ef3', | |
'InKangxi_Radicals' => '\u2f00-\u2fd5', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2ffb', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3041-\u3096\u3099-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3105-\u312d', | |
'InHangul_Compatibility_Jamo' => '\u3131-\u318e', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31b7', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u321e\u3220-\u32fe', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4db5', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fcb', | |
'InYi_Syllables' => '\ua000-\ua48c', | |
'InYi_Radicals' => '\ua490-\ua4c6', | |
'InHangul_Syllables' => '\uac00-\ud7a3', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufa2d\ufa30-\ufa6d\ufa70-\ufad9', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb06\ufb13-\ufb17\ufb1d-\ufb36\ufb38-\ufb3c\ufb3e\ufb40\ufb41\ufb43\ufb44\ufb46-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufbb1\ufbd3-\ufd3f\ufd50-\ufd8f\ufd92-\ufdc7\ufdf0-\ufdfd', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe26', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe52\ufe54-\ufe66\ufe68-\ufe6b', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufe74\ufe76-\ufefc\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff01-\uffbe\uffc2-\uffc7\uffca-\uffcf\uffd2-\uffd7\uffda-\uffdc\uffe0-\uffe6\uffe8-\uffee', | |
'InSpecials' => '\ufff9-\ufffd', | |
), | |
'Co' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Cs' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'L' => array( | |
'InBasic_Latin' => '\u0000-\u0040\u005b-\u0060\u007b-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00a9\u00ab-\u00b4\u00b6-\u00b9\u00bb-\u00bf\u00d7\u00f7', | |
'InSpacing_Modifier_Letters' => '\u02c2-\u02c5\u02d2-\u02df\u02e5-\u02eb\u02ed\u02ef-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0375\u0378\u0379\u037e-\u0385\u0387\u038b\u038d\u03a2\u03f6', | |
'InCyrillic' => '\u0482-\u0489', | |
'InCyrillic_Supplementary' => '\u0526-\u052f', | |
'InArmenian' => '\u0530\u0557\u0558\u055a-\u0560\u0588-\u058f', | |
'InHebrew' => '\u0590-\u05cf\u05eb-\u05ef\u05f3-\u05ff', | |
'InArabic' => '\u0600-\u0620\u064b-\u066d\u0670\u06d4\u06d6-\u06e4\u06e7-\u06ed\u06f0-\u06f9\u06fd\u06fe', | |
'InSyriac' => '\u0700-\u070f\u0711\u0730-\u074c', | |
'InThaana' => '\u07a6-\u07b0\u07b2-\u07bf', | |
'InDevanagari' => '\u0900-\u0903\u093a-\u093c\u093e-\u094f\u0951-\u0957\u0962-\u0970\u0973-\u0978', | |
'InBengali' => '\u0980-\u0984\u098d\u098e\u0991\u0992\u09a9\u09b1\u09b3-\u09b5\u09ba-\u09bc\u09be-\u09cd\u09cf-\u09db\u09de\u09e2-\u09ef\u09f2-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a04\u0a0b-\u0a0e\u0a11\u0a12\u0a29\u0a31\u0a34\u0a37\u0a3a-\u0a58\u0a5d\u0a5f-\u0a71\u0a75-\u0a7f', | |
'InGujarati' => '\u0a80-\u0a84\u0a8e\u0a92\u0aa9\u0ab1\u0ab4\u0aba-\u0abc\u0abe-\u0acf\u0ad1-\u0adf\u0ae2-\u0aff', | |
'InOriya' => '\u0b00-\u0b04\u0b0d\u0b0e\u0b11\u0b12\u0b29\u0b31\u0b34\u0b3a-\u0b3c\u0b3e-\u0b5b\u0b5e\u0b62-\u0b70\u0b72-\u0b7f', | |
'InTamil' => '\u0b80-\u0b82\u0b84\u0b8b-\u0b8d\u0b91\u0b96-\u0b98\u0b9b\u0b9d\u0ba0-\u0ba2\u0ba5-\u0ba7\u0bab-\u0bad\u0bba-\u0bcf\u0bd1-\u0bff', | |
'InTelugu' => '\u0c00-\u0c04\u0c0d\u0c11\u0c29\u0c34\u0c3a-\u0c3c\u0c3e-\u0c57\u0c5a-\u0c5f\u0c62-\u0c7f', | |
'InKannada' => '\u0c80-\u0c84\u0c8d\u0c91\u0ca9\u0cb4\u0cba-\u0cbc\u0cbe-\u0cdd\u0cdf\u0ce2-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d04\u0d0d\u0d11\u0d29\u0d3a-\u0d3c\u0d3e-\u0d5f\u0d62-\u0d79', | |
'InSinhala' => '\u0d80-\u0d84\u0d97-\u0d99\u0db2\u0dbc\u0dbe\u0dbf\u0dc7-\u0dff', | |
'InThai' => '\u0e00\u0e31\u0e34-\u0e3f\u0e47-\u0e7f', | |
'InLao' => '\u0e80\u0e83\u0e85\u0e86\u0e89\u0e8b\u0e8c\u0e8e-\u0e93\u0e98\u0ea0\u0ea4\u0ea6\u0ea8\u0ea9\u0eac\u0eb1\u0eb4-\u0ebc\u0ebe\u0ebf\u0ec5\u0ec7-\u0edb\u0ede-\u0eff', | |
'InTibetan' => '\u0f01-\u0f3f\u0f48\u0f6d-\u0f87\u0f8c-\u0fff', | |
'InMyanmar' => '\u102b-\u103e\u1040-\u104f\u1056-\u1059\u105e-\u1060\u1062-\u1064\u1067-\u106d\u1071-\u1074\u1082-\u108d\u108f-\u109f', | |
'InGeorgian' => '\u10c6-\u10cf\u10fb\u10fd-\u10ff', | |
'InEthiopic' => '\u1249\u124e\u124f\u1257\u1259\u125e\u125f\u1289\u128e\u128f\u12b1\u12b6\u12b7\u12bf\u12c1\u12c6\u12c7\u12d7\u1311\u1316\u1317\u135b-\u137f', | |
'InCherokee' => '\u13f5-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400\u166d\u166e', | |
'InOgham' => '\u1680\u169b-\u169f', | |
'InRunic' => '\u16eb-\u16ff', | |
'InTagalog' => '\u170d\u1712-\u171f', | |
'InHanunoo' => '\u1732-\u173f', | |
'InBuhid' => '\u1752-\u175f', | |
'InTagbanwa' => '\u176d\u1771-\u177f', | |
'InKhmer' => '\u17b4-\u17d6\u17d8-\u17db\u17dd-\u17ff', | |
'InMongolian' => '\u1800-\u181f\u1878-\u187f\u18a9\u18ab-\u18af', | |
'InLimbu' => '\u191d-\u194f', | |
'InTai_Le' => '\u196e\u196f\u1975-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InGreek_Extended' => '\u1f16\u1f17\u1f1e\u1f1f\u1f46\u1f47\u1f4e\u1f4f\u1f58\u1f5a\u1f5c\u1f5e\u1f7e\u1f7f\u1fb5\u1fbd\u1fbf-\u1fc1\u1fc5\u1fcd-\u1fcf\u1fd4\u1fd5\u1fdc-\u1fdf\u1fed-\u1ff1\u1ff5\u1ffd-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070\u2072-\u207e\u2080-\u208f\u2095-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100\u2101\u2103-\u2106\u2108\u2109\u2114\u2116-\u2118\u211e-\u2123\u2125\u2127\u2129\u212e\u213a\u213b\u2140-\u2144\u214a-\u214d\u214f', | |
'InNumber_Forms' => '\u2150-\u2182\u2185-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3004\u3007-\u3030\u3036-\u303a\u303d-\u303f', | |
'InHiragana' => '\u3040\u3097-\u309c', | |
'InKatakana' => '\u30a0\u30fb', | |
'InBopomofo' => '\u3100-\u3104\u312e\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31b8-\u31bf', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u4db6-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u9fcc-\u9fff', | |
'InYi_Syllables' => '\ua48d-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\ud7a4-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\ufa2e\ufa2f\ufa6e\ufa6f\ufada-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb07-\ufb12\ufb18-\ufb1c\ufb1e\ufb29\ufb37\ufb3d\ufb3f\ufb42\ufb45', | |
'InArabic_Presentation_Forms-A' => '\ufbb2-\ufbd2\ufd3e-\ufd4f\ufd90\ufd91\ufdc8-\ufdef\ufdfc-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe75\ufefd-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff20\uff3b-\uff40\uff5b-\uff65\uffbf-\uffc1\uffc8\uffc9\uffd0\uffd1\uffd8\uffd9\uffdd-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Ll' => array( | |
'InBasic_Latin' => '\u0000-\u0060\u007b-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00a9\u00ab-\u00b4\u00b6-\u00b9\u00bb-\u00de\u00f7', | |
'InLatin_Extended-A' => '\u0100\u0102\u0104\u0106\u0108\u010a\u010c\u010e\u0110\u0112\u0114\u0116\u0118\u011a\u011c\u011e\u0120\u0122\u0124\u0126\u0128\u012a\u012c\u012e\u0130\u0132\u0134\u0136\u0139\u013b\u013d\u013f\u0141\u0143\u0145\u0147\u014a\u014c\u014e\u0150\u0152\u0154\u0156\u0158\u015a\u015c\u015e\u0160\u0162\u0164\u0166\u0168\u016a\u016c\u016e\u0170\u0172\u0174\u0176\u0178\u0179\u017b\u017d', | |
'InLatin_Extended-B' => '\u0181\u0182\u0184\u0186\u0187\u0189-\u018b\u018e-\u0191\u0193\u0194\u0196-\u0198\u019c\u019d\u019f\u01a0\u01a2\u01a4\u01a6\u01a7\u01a9\u01ac\u01ae\u01af\u01b1-\u01b3\u01b5\u01b7\u01b8\u01bb\u01bc\u01c0-\u01c5\u01c7\u01c8\u01ca\u01cb\u01cd\u01cf\u01d1\u01d3\u01d5\u01d7\u01d9\u01db\u01de\u01e0\u01e2\u01e4\u01e6\u01e8\u01ea\u01ec\u01ee\u01f1\u01f2\u01f4\u01f6-\u01f8\u01fa\u01fc\u01fe\u0200\u0202\u0204\u0206\u0208\u020a\u020c\u020e\u0210\u0212\u0214\u0216\u0218\u021a\u021c\u021e\u0220\u0222\u0224\u0226\u0228\u022a\u022c\u022e\u0230\u0232\u023a\u023b\u023d\u023e\u0241\u0243-\u0246\u0248\u024a\u024c\u024e', | |
'InIPA_Extensions' => '\u0294', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370\u0372\u0374-\u0376\u0378-\u037a\u037e-\u038f\u0391-\u03ab\u03cf\u03d2-\u03d4\u03d8\u03da\u03dc\u03de\u03e0\u03e2\u03e4\u03e6\u03e8\u03ea\u03ec\u03ee\u03f4\u03f6\u03f7\u03f9\u03fa\u03fd-\u03ff', | |
'InCyrillic' => '\u0400-\u042f\u0460\u0462\u0464\u0466\u0468\u046a\u046c\u046e\u0470\u0472\u0474\u0476\u0478\u047a\u047c\u047e\u0480\u0482-\u048a\u048c\u048e\u0490\u0492\u0494\u0496\u0498\u049a\u049c\u049e\u04a0\u04a2\u04a4\u04a6\u04a8\u04aa\u04ac\u04ae\u04b0\u04b2\u04b4\u04b6\u04b8\u04ba\u04bc\u04be\u04c0\u04c1\u04c3\u04c5\u04c7\u04c9\u04cb\u04cd\u04d0\u04d2\u04d4\u04d6\u04d8\u04da\u04dc\u04de\u04e0\u04e2\u04e4\u04e6\u04e8\u04ea\u04ec\u04ee\u04f0\u04f2\u04f4\u04f6\u04f8\u04fa\u04fc\u04fe', | |
'InCyrillic_Supplementary' => '\u0500\u0502\u0504\u0506\u0508\u050a\u050c\u050e\u0510\u0512\u0514\u0516\u0518\u051a\u051c\u051e\u0520\u0522\u0524\u0526-\u052f', | |
'InArmenian' => '\u0530-\u0560\u0588-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d2c-\u1d61\u1d78', | |
'InLatin_Extended_Additional' => '\u1e00\u1e02\u1e04\u1e06\u1e08\u1e0a\u1e0c\u1e0e\u1e10\u1e12\u1e14\u1e16\u1e18\u1e1a\u1e1c\u1e1e\u1e20\u1e22\u1e24\u1e26\u1e28\u1e2a\u1e2c\u1e2e\u1e30\u1e32\u1e34\u1e36\u1e38\u1e3a\u1e3c\u1e3e\u1e40\u1e42\u1e44\u1e46\u1e48\u1e4a\u1e4c\u1e4e\u1e50\u1e52\u1e54\u1e56\u1e58\u1e5a\u1e5c\u1e5e\u1e60\u1e62\u1e64\u1e66\u1e68\u1e6a\u1e6c\u1e6e\u1e70\u1e72\u1e74\u1e76\u1e78\u1e7a\u1e7c\u1e7e\u1e80\u1e82\u1e84\u1e86\u1e88\u1e8a\u1e8c\u1e8e\u1e90\u1e92\u1e94\u1e9e\u1ea0\u1ea2\u1ea4\u1ea6\u1ea8\u1eaa\u1eac\u1eae\u1eb0\u1eb2\u1eb4\u1eb6\u1eb8\u1eba\u1ebc\u1ebe\u1ec0\u1ec2\u1ec4\u1ec6\u1ec8\u1eca\u1ecc\u1ece\u1ed0\u1ed2\u1ed4\u1ed6\u1ed8\u1eda\u1edc\u1ede\u1ee0\u1ee2\u1ee4\u1ee6\u1ee8\u1eea\u1eec\u1eee\u1ef0\u1ef2\u1ef4\u1ef6\u1ef8\u1efa\u1efc\u1efe', | |
'InGreek_Extended' => '\u1f08-\u1f0f\u1f16-\u1f1f\u1f28-\u1f2f\u1f38-\u1f3f\u1f46-\u1f4f\u1f58-\u1f5f\u1f68-\u1f6f\u1f7e\u1f7f\u1f88-\u1f8f\u1f98-\u1f9f\u1fa8-\u1faf\u1fb5\u1fb8-\u1fbd\u1fbf-\u1fc1\u1fc5\u1fc8-\u1fcf\u1fd4\u1fd5\u1fd8-\u1fdf\u1fe8-\u1ff1\u1ff5\u1ff8-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u2109\u210b-\u210d\u2110-\u2112\u2114-\u212e\u2130-\u2133\u2135-\u2138\u213a\u213b\u213e-\u2145\u214a-\u214d\u214f', | |
'InNumber_Forms' => '\u2150-\u2183\u2185-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb07-\ufb12\ufb18-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff40\uff5b-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Lm' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02c2-\u02c5\u02d2-\u02df\u02e5-\u02eb\u02ed\u02ef-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u0373\u0375-\u0379\u037b-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u0558\u055a-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u063f\u0641-\u06e4\u06e7-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u0970\u0972-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e45\u0e47-\u0e7f', | |
'InLao' => '\u0e80-\u0ec5\u0ec7-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10fb\u10fd-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17d6\u17d8-\u17ff', | |
'InMongolian' => '\u1800-\u1842\u1844-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d2b\u1d62-\u1d77\u1d79-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070\u2072-\u207e\u2080-\u208f\u2095-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3004\u3006-\u3030\u3036-\u303a\u303c-\u303f', | |
'InHiragana' => '\u3040-\u309c\u309f', | |
'InKatakana' => '\u30a0-\u30fb\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua014\ua016-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff6f\uff71-\uff9d\uffa0-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Lo' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u01ba\u01bc-\u01bf\u01c4-\u024f', | |
'InIPA_Extensions' => '\u0250-\u0293\u0295-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05cf\u05eb-\u05ef\u05f3-\u05ff', | |
'InArabic' => '\u0600-\u0620\u0640\u064b-\u066d\u0670\u06d4\u06d6-\u06ed\u06f0-\u06f9\u06fd\u06fe', | |
'InSyriac' => '\u0700-\u070f\u0711\u0730-\u074c', | |
'InThaana' => '\u07a6-\u07b0\u07b2-\u07bf', | |
'InDevanagari' => '\u0900-\u0903\u093a-\u093c\u093e-\u094f\u0951-\u0957\u0962-\u0971\u0973-\u0978', | |
'InBengali' => '\u0980-\u0984\u098d\u098e\u0991\u0992\u09a9\u09b1\u09b3-\u09b5\u09ba-\u09bc\u09be-\u09cd\u09cf-\u09db\u09de\u09e2-\u09ef\u09f2-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a04\u0a0b-\u0a0e\u0a11\u0a12\u0a29\u0a31\u0a34\u0a37\u0a3a-\u0a58\u0a5d\u0a5f-\u0a71\u0a75-\u0a7f', | |
'InGujarati' => '\u0a80-\u0a84\u0a8e\u0a92\u0aa9\u0ab1\u0ab4\u0aba-\u0abc\u0abe-\u0acf\u0ad1-\u0adf\u0ae2-\u0aff', | |
'InOriya' => '\u0b00-\u0b04\u0b0d\u0b0e\u0b11\u0b12\u0b29\u0b31\u0b34\u0b3a-\u0b3c\u0b3e-\u0b5b\u0b5e\u0b62-\u0b70\u0b72-\u0b7f', | |
'InTamil' => '\u0b80-\u0b82\u0b84\u0b8b-\u0b8d\u0b91\u0b96-\u0b98\u0b9b\u0b9d\u0ba0-\u0ba2\u0ba5-\u0ba7\u0bab-\u0bad\u0bba-\u0bcf\u0bd1-\u0bff', | |
'InTelugu' => '\u0c00-\u0c04\u0c0d\u0c11\u0c29\u0c34\u0c3a-\u0c3c\u0c3e-\u0c57\u0c5a-\u0c5f\u0c62-\u0c7f', | |
'InKannada' => '\u0c80-\u0c84\u0c8d\u0c91\u0ca9\u0cb4\u0cba-\u0cbc\u0cbe-\u0cdd\u0cdf\u0ce2-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d04\u0d0d\u0d11\u0d29\u0d3a-\u0d3c\u0d3e-\u0d5f\u0d62-\u0d79', | |
'InSinhala' => '\u0d80-\u0d84\u0d97-\u0d99\u0db2\u0dbc\u0dbe\u0dbf\u0dc7-\u0dff', | |
'InThai' => '\u0e00\u0e31\u0e34-\u0e3f\u0e46-\u0e7f', | |
'InLao' => '\u0e80\u0e83\u0e85\u0e86\u0e89\u0e8b\u0e8c\u0e8e-\u0e93\u0e98\u0ea0\u0ea4\u0ea6\u0ea8\u0ea9\u0eac\u0eb1\u0eb4-\u0ebc\u0ebe\u0ebf\u0ec5-\u0edb\u0ede-\u0eff', | |
'InTibetan' => '\u0f01-\u0f3f\u0f48\u0f6d-\u0f87\u0f8c-\u0fff', | |
'InMyanmar' => '\u102b-\u103e\u1040-\u104f\u1056-\u1059\u105e-\u1060\u1062-\u1064\u1067-\u106d\u1071-\u1074\u1082-\u108d\u108f-\u109f', | |
'InGeorgian' => '\u10a0-\u10cf\u10fb-\u10ff', | |
'InEthiopic' => '\u1249\u124e\u124f\u1257\u1259\u125e\u125f\u1289\u128e\u128f\u12b1\u12b6\u12b7\u12bf\u12c1\u12c6\u12c7\u12d7\u1311\u1316\u1317\u135b-\u137f', | |
'InCherokee' => '\u13f5-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400\u166d\u166e', | |
'InOgham' => '\u1680\u169b-\u169f', | |
'InRunic' => '\u16eb-\u16ff', | |
'InTagalog' => '\u170d\u1712-\u171f', | |
'InHanunoo' => '\u1732-\u173f', | |
'InBuhid' => '\u1752-\u175f', | |
'InTagbanwa' => '\u176d\u1771-\u177f', | |
'InKhmer' => '\u17b4-\u17db\u17dd-\u17ff', | |
'InMongolian' => '\u1800-\u181f\u1843\u1878-\u187f\u18a9\u18ab-\u18af', | |
'InLimbu' => '\u191d-\u194f', | |
'InTai_Le' => '\u196e\u196f\u1975-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u2134\u2139-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3005\u3007-\u303b\u303d-\u303f', | |
'InHiragana' => '\u3040\u3097-\u309e', | |
'InKatakana' => '\u30a0\u30fb-\u30fe', | |
'InBopomofo' => '\u3100-\u3104\u312e\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31b8-\u31bf', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u4db6-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u9fcc-\u9fff', | |
'InYi_Syllables' => '\ua015\ua48d-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\ud7a4-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\ufa2e\ufa2f\ufa6e\ufa6f\ufada-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb1c\ufb1e\ufb29\ufb37\ufb3d\ufb3f\ufb42\ufb45', | |
'InArabic_Presentation_Forms-A' => '\ufbb2-\ufbd2\ufd3e-\ufd4f\ufd90\ufd91\ufdc8-\ufdef\ufdfc-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe75\ufefd-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff65\uff70\uff9e\uff9f\uffbf-\uffc1\uffc8\uffc9\uffd0\uffd1\uffd8\uffd9\uffdd-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Lt' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u01c4\u01c6\u01c7\u01c9\u01ca\u01cc-\u01f1\u01f3-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1f87\u1f90-\u1f97\u1fa0-\u1fa7\u1fb0-\u1fbb\u1fbd-\u1fcb\u1fcd-\u1ffb\u1ffd-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Lu' => array( | |
'InBasic_Latin' => '\u0000-\u0040\u005b-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00bf\u00d7\u00df-\u00ff', | |
'InLatin_Extended-A' => '\u0101\u0103\u0105\u0107\u0109\u010b\u010d\u010f\u0111\u0113\u0115\u0117\u0119\u011b\u011d\u011f\u0121\u0123\u0125\u0127\u0129\u012b\u012d\u012f\u0131\u0133\u0135\u0137\u0138\u013a\u013c\u013e\u0140\u0142\u0144\u0146\u0148\u0149\u014b\u014d\u014f\u0151\u0153\u0155\u0157\u0159\u015b\u015d\u015f\u0161\u0163\u0165\u0167\u0169\u016b\u016d\u016f\u0171\u0173\u0175\u0177\u017a\u017c\u017e\u017f', | |
'InLatin_Extended-B' => '\u0180\u0183\u0185\u0188\u018c\u018d\u0192\u0195\u0199-\u019b\u019e\u01a1\u01a3\u01a5\u01a8\u01aa\u01ab\u01ad\u01b0\u01b4\u01b6\u01b9-\u01bb\u01bd-\u01c3\u01c5\u01c6\u01c8\u01c9\u01cb\u01cc\u01ce\u01d0\u01d2\u01d4\u01d6\u01d8\u01da\u01dc\u01dd\u01df\u01e1\u01e3\u01e5\u01e7\u01e9\u01eb\u01ed\u01ef\u01f0\u01f2\u01f3\u01f5\u01f9\u01fb\u01fd\u01ff\u0201\u0203\u0205\u0207\u0209\u020b\u020d\u020f\u0211\u0213\u0215\u0217\u0219\u021b\u021d\u021f\u0221\u0223\u0225\u0227\u0229\u022b\u022d\u022f\u0231\u0233-\u0239\u023c\u023f\u0240\u0242\u0247\u0249\u024b\u024d\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0371\u0373-\u0375\u0377-\u0385\u0387\u038b\u038d\u0390\u03a2\u03ac-\u03ce\u03d0\u03d1\u03d5-\u03d7\u03d9\u03db\u03dd\u03df\u03e1\u03e3\u03e5\u03e7\u03e9\u03eb\u03ed\u03ef-\u03f3\u03f5\u03f6\u03f8\u03fb\u03fc', | |
'InCyrillic' => '\u0430-\u045f\u0461\u0463\u0465\u0467\u0469\u046b\u046d\u046f\u0471\u0473\u0475\u0477\u0479\u047b\u047d\u047f\u0481-\u0489\u048b\u048d\u048f\u0491\u0493\u0495\u0497\u0499\u049b\u049d\u049f\u04a1\u04a3\u04a5\u04a7\u04a9\u04ab\u04ad\u04af\u04b1\u04b3\u04b5\u04b7\u04b9\u04bb\u04bd\u04bf\u04c2\u04c4\u04c6\u04c8\u04ca\u04cc\u04ce\u04cf\u04d1\u04d3\u04d5\u04d7\u04d9\u04db\u04dd\u04df\u04e1\u04e3\u04e5\u04e7\u04e9\u04eb\u04ed\u04ef\u04f1\u04f3\u04f5\u04f7\u04f9\u04fb\u04fd\u04ff', | |
'InCyrillic_Supplementary' => '\u0501\u0503\u0505\u0507\u0509\u050b\u050d\u050f\u0511\u0513\u0515\u0517\u0519\u051b\u051d\u051f\u0521\u0523\u0525-\u052f', | |
'InArmenian' => '\u0530\u0557-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10c6-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e01\u1e03\u1e05\u1e07\u1e09\u1e0b\u1e0d\u1e0f\u1e11\u1e13\u1e15\u1e17\u1e19\u1e1b\u1e1d\u1e1f\u1e21\u1e23\u1e25\u1e27\u1e29\u1e2b\u1e2d\u1e2f\u1e31\u1e33\u1e35\u1e37\u1e39\u1e3b\u1e3d\u1e3f\u1e41\u1e43\u1e45\u1e47\u1e49\u1e4b\u1e4d\u1e4f\u1e51\u1e53\u1e55\u1e57\u1e59\u1e5b\u1e5d\u1e5f\u1e61\u1e63\u1e65\u1e67\u1e69\u1e6b\u1e6d\u1e6f\u1e71\u1e73\u1e75\u1e77\u1e79\u1e7b\u1e7d\u1e7f\u1e81\u1e83\u1e85\u1e87\u1e89\u1e8b\u1e8d\u1e8f\u1e91\u1e93\u1e95-\u1e9d\u1e9f\u1ea1\u1ea3\u1ea5\u1ea7\u1ea9\u1eab\u1ead\u1eaf\u1eb1\u1eb3\u1eb5\u1eb7\u1eb9\u1ebb\u1ebd\u1ebf\u1ec1\u1ec3\u1ec5\u1ec7\u1ec9\u1ecb\u1ecd\u1ecf\u1ed1\u1ed3\u1ed5\u1ed7\u1ed9\u1edb\u1edd\u1edf\u1ee1\u1ee3\u1ee5\u1ee7\u1ee9\u1eeb\u1eed\u1eef\u1ef1\u1ef3\u1ef5\u1ef7\u1ef9\u1efb\u1efd\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1f07\u1f10-\u1f17\u1f1e-\u1f27\u1f30-\u1f37\u1f40-\u1f47\u1f4e-\u1f58\u1f5a\u1f5c\u1f5e\u1f60-\u1f67\u1f70-\u1fb7\u1fbc-\u1fc7\u1fcc-\u1fd7\u1fdc-\u1fe7\u1fed-\u1ff7\u1ffc-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100\u2101\u2103-\u2106\u2108-\u210a\u210e\u210f\u2113\u2114\u2116-\u2118\u211e-\u2123\u2125\u2127\u2129\u212e\u212f\u2134-\u213d\u2140-\u2144\u2146-\u214f', | |
'InNumber_Forms' => '\u2150-\u2182\u2184-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff20\uff3b-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'M' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u0482\u048a-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590\u05be\u05c0\u05c3\u05c6\u05c8-\u05ff', | |
'InArabic' => '\u0600-\u060f\u061b-\u064a\u065f-\u066f\u0671-\u06d5\u06dd\u06e5\u06e6\u06e9\u06ee-\u06ff', | |
'InSyriac' => '\u0700-\u0710\u0712-\u072f\u074b-\u074f', | |
'InThaana' => '\u0780-\u07a5\u07b1-\u07bf', | |
'InDevanagari' => '\u0904-\u093b\u093d\u094f\u0950\u0956-\u0961\u0964-\u097f', | |
'InBengali' => '\u0980\u0984-\u09bb\u09bd\u09c5\u09c6\u09c9\u09ca\u09ce-\u09d6\u09d8-\u09e1\u09e4-\u09ff', | |
'InGurmukhi' => '\u0a00\u0a04-\u0a3b\u0a3d\u0a43-\u0a46\u0a49\u0a4a\u0a4e-\u0a50\u0a52-\u0a6f\u0a72-\u0a74\u0a76-\u0a7f', | |
'InGujarati' => '\u0a80\u0a84-\u0abb\u0abd\u0ac6\u0aca\u0ace-\u0ae1\u0ae4-\u0aff', | |
'InOriya' => '\u0b00\u0b04-\u0b3b\u0b3d\u0b45\u0b46\u0b49\u0b4a\u0b4e-\u0b55\u0b58-\u0b61\u0b64-\u0b7f', | |
'InTamil' => '\u0b80\u0b81\u0b83-\u0bbd\u0bc3-\u0bc5\u0bc9\u0bce-\u0bd6\u0bd8-\u0bff', | |
'InTelugu' => '\u0c00\u0c04-\u0c3d\u0c45\u0c49\u0c4e-\u0c54\u0c57-\u0c61\u0c64-\u0c7f', | |
'InKannada' => '\u0c80\u0c81\u0c84-\u0cbb\u0cbd\u0cc5\u0cc9\u0cce-\u0cd4\u0cd7-\u0ce1\u0ce4-\u0cff', | |
'InMalayalam' => '\u0d00\u0d01\u0d04-\u0d3d\u0d45\u0d49\u0d4e-\u0d56\u0d58-\u0d61\u0d64-\u0d7f', | |
'InSinhala' => '\u0d80\u0d81\u0d84-\u0dc9\u0dcb-\u0dce\u0dd5\u0dd7\u0de0-\u0df1\u0df4-\u0dff', | |
'InThai' => '\u0e00-\u0e30\u0e32\u0e33\u0e3b-\u0e46\u0e4f-\u0e7f', | |
'InLao' => '\u0e80-\u0eb0\u0eb2\u0eb3\u0eba\u0ebd-\u0ec7\u0ece-\u0eff', | |
'InTibetan' => '\u0f00-\u0f17\u0f1a-\u0f34\u0f36\u0f38\u0f3a-\u0f3d\u0f40-\u0f70\u0f85\u0f88-\u0f8f\u0f98\u0fbd-\u0fc5\u0fc7-\u0fff', | |
'InMyanmar' => '\u1000-\u102a\u103f-\u1055\u105a-\u105d\u1061\u1065\u1066\u106e-\u1070\u1075-\u1081\u108e\u1090-\u1099\u109e\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u135e\u1360-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u1711\u1715-\u171f', | |
'InHanunoo' => '\u1720-\u1731\u1735-\u173f', | |
'InBuhid' => '\u1740-\u1751\u1754-\u175f', | |
'InTagbanwa' => '\u1760-\u1771\u1774-\u177f', | |
'InKhmer' => '\u1780-\u17b5\u17d4-\u17dc\u17de-\u17ff', | |
'InMongolian' => '\u1800-\u180a\u180e-\u18a8\u18aa-\u18af', | |
'InLimbu' => '\u1900-\u191f\u192c-\u192f\u193c-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20f1-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3029\u3030-\u303f', | |
'InHiragana' => '\u3040-\u3098\u309b-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb1d\ufb1f-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InCombining_Half_Marks' => '\ufe27-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Mc' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u0902\u0904-\u093d\u0941-\u0948\u094d\u094f-\u097f', | |
'InBengali' => '\u0980\u0981\u0984-\u09bd\u09c1-\u09c6\u09c9\u09ca\u09cd-\u09d6\u09d8-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a02\u0a04-\u0a3d\u0a41-\u0a7f', | |
'InGujarati' => '\u0a80-\u0a82\u0a84-\u0abd\u0ac1-\u0ac8\u0aca\u0acd-\u0aff', | |
'InOriya' => '\u0b00\u0b01\u0b04-\u0b3d\u0b3f\u0b41-\u0b46\u0b49\u0b4a\u0b4d-\u0b56\u0b58-\u0b7f', | |
'InTamil' => '\u0b80-\u0bbd\u0bc0\u0bc3-\u0bc5\u0bc9\u0bcd-\u0bd6\u0bd8-\u0bff', | |
'InTelugu' => '\u0c00\u0c04-\u0c40\u0c45-\u0c7f', | |
'InKannada' => '\u0c80\u0c81\u0c84-\u0cbd\u0cbf\u0cc5\u0cc6\u0cc9\u0ccc-\u0cd4\u0cd7-\u0cff', | |
'InMalayalam' => '\u0d00\u0d01\u0d04-\u0d3d\u0d41-\u0d45\u0d49\u0d4d-\u0d56\u0d58-\u0d7f', | |
'InSinhala' => '\u0d80\u0d81\u0d84-\u0dce\u0dd2-\u0dd7\u0de0-\u0df1\u0df4-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0f3d\u0f40-\u0f7e\u0f80-\u0fff', | |
'InMyanmar' => '\u1000-\u102a\u102d-\u1030\u1032-\u1037\u1039\u103a\u103d-\u1055\u1058-\u1061\u1065\u1066\u106e-\u1082\u1085\u1086\u108d\u108e\u1090-\u1099\u109d-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17b5\u17b7-\u17bd\u17c6\u17c9-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u1922\u1927\u1928\u192c-\u192f\u1932\u1939-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Me' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u0487\u048a-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06dd\u06df-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20dc\u20e1\u20e5-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Mn' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u0482\u0488-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590\u05be\u05c0\u05c3\u05c6\u05c8-\u05ff', | |
'InArabic' => '\u0600-\u060f\u061b-\u064a\u065f-\u066f\u0671-\u06d5\u06dd\u06de\u06e5\u06e6\u06e9\u06ee-\u06ff', | |
'InSyriac' => '\u0700-\u0710\u0712-\u072f\u074b-\u074f', | |
'InThaana' => '\u0780-\u07a5\u07b1-\u07bf', | |
'InDevanagari' => '\u0903-\u093b\u093d-\u0940\u0949-\u094c\u094e-\u0950\u0956-\u0961\u0964-\u097f', | |
'InBengali' => '\u0980\u0982-\u09bb\u09bd-\u09c0\u09c5-\u09cc\u09ce-\u09e1\u09e4-\u09ff', | |
'InGurmukhi' => '\u0a00\u0a03-\u0a3b\u0a3d-\u0a40\u0a43-\u0a46\u0a49\u0a4a\u0a4e-\u0a50\u0a52-\u0a6f\u0a72-\u0a74\u0a76-\u0a7f', | |
'InGujarati' => '\u0a80\u0a83-\u0abb\u0abd-\u0ac0\u0ac6\u0ac9-\u0acc\u0ace-\u0ae1\u0ae4-\u0aff', | |
'InOriya' => '\u0b00\u0b02-\u0b3b\u0b3d\u0b3e\u0b40\u0b45-\u0b4c\u0b4e-\u0b55\u0b57-\u0b61\u0b64-\u0b7f', | |
'InTamil' => '\u0b80\u0b81\u0b83-\u0bbf\u0bc1-\u0bcc\u0bce-\u0bff', | |
'InTelugu' => '\u0c00-\u0c3d\u0c41-\u0c45\u0c49\u0c4e-\u0c54\u0c57-\u0c61\u0c64-\u0c7f', | |
'InKannada' => '\u0c80-\u0cbb\u0cbd\u0cbe\u0cc0-\u0cc5\u0cc7-\u0ccb\u0cce-\u0ce1\u0ce4-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d40\u0d45-\u0d4c\u0d4e-\u0d61\u0d64-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dc9\u0dcb-\u0dd1\u0dd5\u0dd7-\u0dff', | |
'InThai' => '\u0e00-\u0e30\u0e32\u0e33\u0e3b-\u0e46\u0e4f-\u0e7f', | |
'InLao' => '\u0e80-\u0eb0\u0eb2\u0eb3\u0eba\u0ebd-\u0ec7\u0ece-\u0eff', | |
'InTibetan' => '\u0f00-\u0f17\u0f1a-\u0f34\u0f36\u0f38\u0f3a-\u0f70\u0f7f\u0f85\u0f88-\u0f8f\u0f98\u0fbd-\u0fc5\u0fc7-\u0fff', | |
'InMyanmar' => '\u1000-\u102c\u1031\u1038\u103b\u103c\u103f-\u1057\u105a-\u105d\u1061-\u1070\u1075-\u1081\u1083\u1084\u1087-\u108c\u108e-\u109c\u109e\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u135e\u1360-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u1711\u1715-\u171f', | |
'InHanunoo' => '\u1720-\u1731\u1735-\u173f', | |
'InBuhid' => '\u1740-\u1751\u1754-\u175f', | |
'InTagbanwa' => '\u1760-\u1771\u1774-\u177f', | |
'InKhmer' => '\u1780-\u17b6\u17be-\u17c5\u17c7\u17c8\u17d4-\u17dc\u17de-\u17ff', | |
'InMongolian' => '\u1800-\u180a\u180e-\u18a8\u18aa-\u18af', | |
'InLimbu' => '\u1900-\u191f\u1923-\u1926\u1929-\u1931\u1933-\u1938\u193c-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20dd-\u20e0\u20e2-\u20e4\u20f1-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3029\u3030-\u303f', | |
'InHiragana' => '\u3040-\u3098\u309b-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb1d\ufb1f-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InCombining_Half_Marks' => '\ufe27-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'N' => array( | |
'InBasic_Latin' => '\u0000-\u002f\u003a-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00b1\u00b4-\u00b8\u00ba\u00bb\u00bf-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u065f\u066a-\u06ef\u06fa-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u0965\u0970-\u097f', | |
'InBengali' => '\u0980-\u09e5\u09f0-\u09f3\u09fa-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a65\u0a70-\u0a7f', | |
'InGujarati' => '\u0a80-\u0ae5\u0af0-\u0aff', | |
'InOriya' => '\u0b00-\u0b65\u0b70-\u0b7f', | |
'InTamil' => '\u0b80-\u0be5\u0bf3-\u0bff', | |
'InTelugu' => '\u0c00-\u0c65\u0c70-\u0c77\u0c7f', | |
'InKannada' => '\u0c80-\u0ce5\u0cf0-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d65\u0d76-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e4f\u0e5a-\u0e7f', | |
'InLao' => '\u0e80-\u0ecf\u0eda-\u0eff', | |
'InTibetan' => '\u0f00-\u0f1f\u0f34-\u0fff', | |
'InMyanmar' => '\u1000-\u103f\u104a-\u108f\u109a-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u1368\u137d-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ed\u16f1-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17df\u17ea-\u17ef\u17fa-\u17ff', | |
'InMongolian' => '\u1800-\u180f\u181a-\u18af', | |
'InLimbu' => '\u1900-\u1945', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2071-\u2073\u207a-\u207f\u208a-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2183\u2184\u218a-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u249c-\u24e9', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u2775\u2794-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3006\u3008-\u3020\u302a-\u3037\u303b-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190\u3191\u3196-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u321f\u322a-\u3250\u3260-\u327f\u328a-\u32b0\u32c0-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff0f\uff1a-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Nd' => array( | |
'InBasic_Latin' => '\u0000-\u002f\u003a-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u065f\u066a-\u06ef\u06fa-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u0965\u0970-\u097f', | |
'InBengali' => '\u0980-\u09e5\u09f0-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a65\u0a70-\u0a7f', | |
'InGujarati' => '\u0a80-\u0ae5\u0af0-\u0aff', | |
'InOriya' => '\u0b00-\u0b65\u0b70-\u0b7f', | |
'InTamil' => '\u0b80-\u0be5\u0bf0-\u0bff', | |
'InTelugu' => '\u0c00-\u0c65\u0c70-\u0c7f', | |
'InKannada' => '\u0c80-\u0ce5\u0cf0-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d65\u0d70-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e4f\u0e5a-\u0e7f', | |
'InLao' => '\u0e80-\u0ecf\u0eda-\u0eff', | |
'InTibetan' => '\u0f00-\u0f1f\u0f2a-\u0fff', | |
'InMyanmar' => '\u1000-\u103f\u104a-\u108f\u109a-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17df\u17ea-\u17ff', | |
'InMongolian' => '\u1800-\u180f\u181a-\u18af', | |
'InLimbu' => '\u1900-\u1945', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff0f\uff1a-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Nl' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ed\u16f1-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u215f\u2183\u2184\u2189-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3006\u3008-\u3020\u302a-\u3037\u303b-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'No' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00b1\u00b4-\u00b8\u00ba\u00bb\u00bf-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09f3\u09fa-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bef\u0bf3-\u0bff', | |
'InTelugu' => '\u0c00-\u0c77\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d6f\u0d76-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0f29\u0f34-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u1368\u137d-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ef\u17fa-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2071-\u2073\u207a-\u207f\u208a-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2160-\u2188\u218a-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u249c-\u24e9', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u2775\u2794-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190\u3191\u3196-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u321f\u322a-\u3250\u3260-\u327f\u328a-\u32b0\u32c0-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'P' => array( | |
'InBasic_Latin' => '\u0000-\u0020\u0024\u002b\u0030-\u0039\u003c-\u003e\u0041-\u005a\u005e\u0060-\u007a\u007c\u007e\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00a0\u00a2-\u00aa\u00ac-\u00b6\u00b8-\u00ba\u00bc-\u00be\u00c0-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u037d\u037f-\u0386\u0388-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u0559\u0560-\u0588\u058b-\u058f', | |
'InHebrew' => '\u0590-\u05bd\u05bf\u05c1\u05c2\u05c4\u05c5\u05c7-\u05f2\u05f5-\u05ff', | |
'InArabic' => '\u0600-\u0608\u060b\u060e-\u061a\u061c\u061d\u0620-\u0669\u066e-\u06d3\u06d5-\u06ff', | |
'InSyriac' => '\u070e-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u0963\u0966-\u096f\u0971-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0df3\u0df5-\u0dff', | |
'InThai' => '\u0e00-\u0e4e\u0e50-\u0e59\u0e5c-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0f03\u0f13-\u0f39\u0f3e-\u0f84\u0f86-\u0fcf\u0fd5-\u0fff', | |
'InMyanmar' => '\u1000-\u1049\u1050-\u109f', | |
'InGeorgian' => '\u10a0-\u10fa\u10fc-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u1360\u1369-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1401-\u166c\u166f-\u167f', | |
'InOgham' => '\u1680-\u169a\u169d-\u169f', | |
'InRunic' => '\u16a0-\u16ea\u16ee-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u1734\u1737-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17d3\u17d7\u17db-\u17ff', | |
'InMongolian' => '\u180b-\u18af', | |
'InLimbu' => '\u1900-\u1943\u1946-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u200f\u2028-\u202f\u2044\u2052\u205f-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u207c\u207f-\u208c\u208f-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u2328\u232b-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u2767\u2776-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27c4\u27c7-\u27e5', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u2982\u2999-\u29d7\u29dc-\u29fb\u29fe\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000\u3004-\u3007\u3012\u3013\u3020-\u302f\u3031-\u303c\u303e\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a1-\u30fa\u30fc-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufd3d\ufd40-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InSmall_Form_Variants' => '\ufe53\ufe62\ufe64-\ufe67\ufe69\ufe6c-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00\uff04\uff0b\uff10-\uff19\uff1c-\uff1e\uff21-\uff3a\uff3e\uff40-\uff5a\uff5c\uff5e\uff66-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Pc' => array( | |
'InBasic_Latin' => '\u0000-\u005e\u0060-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u203e\u2041-\u2053\u2055-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe32\ufe35-\ufe4c', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff3e\uff40-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Pd' => array( | |
'InBasic_Latin' => '\u0000-\u002c\u002e-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u0589\u058b-\u058f', | |
'InHebrew' => '\u0590-\u05bd\u05bf-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1401-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u1805\u1807-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u200f\u2016-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u301b\u301d-\u302f\u3031-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a1-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30\ufe33-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe57\ufe59-\ufe62\ufe64-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff0c\uff0e-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Pe' => array( | |
'InBasic_Latin' => '\u0000-\u0028\u002a-\u005c\u005e-\u007c\u007e\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0f3a\u0f3c\u0f3e-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169b\u169d-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u2045\u2047-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u207d\u207f-\u208d\u208f-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u2329\u232b-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u2768\u276a\u276c\u276e\u2770\u2772\u2774\u2776-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27c5\u27c7-\u27e6\u27e8\u27ea\u27ec\u27ee', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u2983\u2985\u2987\u2989\u298b\u298d\u298f\u2991\u2993\u2995\u2997\u2999-\u29d8\u29da\u29dc-\u29fc\u29fe\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3008\u300a\u300c\u300e\u3010\u3012-\u3014\u3016\u3018\u301a\u301c\u301d\u3020-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufd3e\ufd40-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe35\ufe37\ufe39\ufe3b\ufe3d\ufe3f\ufe41\ufe43\ufe45-\ufe47\ufe49-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe59\ufe5b\ufe5d\ufe5f-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff08\uff0a-\uff3c\uff3e-\uff5c\uff5e\uff5f\uff61\uff62\uff64-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Pf' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ba\u00bc-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u2018\u201a-\u201c\u201e-\u2039\u203b-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Pi' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00aa\u00ac-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u2017\u2019\u201a\u201d\u201e\u2020-\u2038\u203a-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Po' => array( | |
'InBasic_Latin' => '\u0000-\u0020\u0024\u0028\u0029\u002b\u002d\u0030-\u0039\u003c-\u003e\u0041-\u005b\u005d-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00a0\u00a2-\u00b6\u00b8-\u00be\u00c0-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u037d\u037f-\u0386\u0388-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u0559\u0560-\u0588\u058a-\u058f', | |
'InHebrew' => '\u0590-\u05bf\u05c1\u05c2\u05c4\u05c5\u05c7-\u05f2\u05f5-\u05ff', | |
'InArabic' => '\u0600-\u0608\u060b\u060e-\u061a\u061c\u061d\u0620-\u0669\u066e-\u06d3\u06d5-\u06ff', | |
'InSyriac' => '\u070e-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u0963\u0966-\u096f\u0971-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0df3\u0df5-\u0dff', | |
'InThai' => '\u0e00-\u0e4e\u0e50-\u0e59\u0e5c-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0f03\u0f13-\u0f84\u0f86-\u0fcf\u0fd5-\u0fff', | |
'InMyanmar' => '\u1000-\u1049\u1050-\u109f', | |
'InGeorgian' => '\u10a0-\u10fa\u10fc-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u1360\u1369-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u166c\u166f-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ea\u16ee-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u1734\u1737-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17d3\u17d7\u17db-\u17ff', | |
'InMongolian' => '\u1806\u180b-\u18af', | |
'InLimbu' => '\u1900-\u1943\u1946-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u2015\u2018-\u201f\u2028-\u202f\u2039\u203a\u203f\u2040\u2044-\u2046\u2052\u2054\u205f-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000\u3004-\u303c\u303e\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30fa\u30fc-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe31-\ufe44\ufe47\ufe48\ufe4d-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe53\ufe58-\ufe5e\ufe62-\ufe67\ufe69\ufe6c-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00\uff04\uff08\uff09\uff0b\uff0d\uff10-\uff19\uff1c-\uff1e\uff21-\uff3b\uff3d-\uff60\uff62\uff63\uff66-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Ps' => array( | |
'InBasic_Latin' => '\u0000-\u0027\u0029-\u005a\u005c-\u007a\u007c-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0f39\u0f3b\u0f3d-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169a\u169c-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u2019\u201b-\u201d\u201f-\u2044\u2046-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u207c\u207e-\u208c\u208e-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u2328\u232a-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u2767\u2769\u276b\u276d\u276f\u2771\u2773\u2775-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27c4\u27c6-\u27e5\u27e7\u27e9\u27eb\u27ed\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u2982\u2984\u2986\u2988\u298a\u298c\u298e\u2990\u2992\u2994\u2996\u2998-\u29d7\u29d9\u29db-\u29fb\u29fd-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3007\u3009\u300b\u300d\u300f\u3011-\u3013\u3015\u3017\u3019\u301b\u301c\u301e-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufd3d\ufd3f-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe34\ufe36\ufe38\ufe3a\ufe3c\ufe3e\ufe40\ufe42\ufe44-\ufe46\ufe48-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe58\ufe5a\ufe5c\ufe5e-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff07\uff09-\uff3a\uff3c-\uff5a\uff5c-\uff5e\uff60\uff61\uff63-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'S' => array( | |
'InBasic_Latin' => '\u0000-\u0023\u0025-\u002a\u002c-\u003b\u003f-\u005d\u005f\u0061-\u007b\u007d\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00a1\u00aa\u00ab\u00ad\u00b2\u00b3\u00b5\u00b7\u00b9-\u00d6\u00d8-\u00f6\u00f8-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02c1\u02c6-\u02d1\u02e0-\u02e4\u02ec\u02ee', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u0374\u0376-\u0383\u0386-\u03f5\u03f7-\u03ff', | |
'InCyrillic' => '\u0400-\u0481\u0483-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u0605\u0609\u060a\u060c\u060d\u0610-\u06e8\u06ea-\u06fc\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09f1\u09f4-\u09f9\u09fc-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0af0\u0af2-\u0aff', | |
'InOriya' => '\u0b00-\u0b6f\u0b71-\u0b7f', | |
'InTamil' => '\u0b80-\u0bf2\u0bfb-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7e', | |
'InKannada' => '\u0c80-\u0cf0\u0cf3-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d78\u0d7a-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e3e\u0e40-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00\u0f04-\u0f12\u0f18\u0f19\u0f20-\u0f33\u0f35\u0f37\u0f39-\u0fbd\u0fc6\u0fcd\u0fd0-\u0fd4\u0fd9-\u0fff', | |
'InMyanmar' => '\u1000-\u109d', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u135f\u1361-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17da\u17dc-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u193f\u1941-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fbc\u1fbe\u1fc2-\u1fcc\u1fd0-\u1fdc\u1fe0-\u1fec\u1ff0-\u1ffc\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u2043\u2045-\u2051\u2053-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u2079\u207d-\u2089\u208d-\u209f', | |
'InCurrency_Symbols' => '\u20b9-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2102\u2107\u210a-\u2113\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u212f-\u2139\u213c-\u213f\u2145-\u2149\u214e', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InMiscellaneous_Technical' => '\u2329\u232a\u23e9-\u23ff', | |
'InControl_Pictures' => '\u2427-\u243f', | |
'InOptical_Character_Recognition' => '\u244b-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u249b\u24ea-\u24ff', | |
'InMiscellaneous_Symbols' => '\u26ce\u26e2\u26e4-\u26e7', | |
'InDingbats' => '\u2700\u2705\u270a\u270b\u2728\u274c\u274e\u2753-\u2755\u275f\u2760\u2768-\u2793\u2795-\u2797\u27b0\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c5\u27c6\u27cb\u27cd-\u27cf\u27e6-\u27ef', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2983-\u2998\u29d8-\u29db\u29fc\u29fd', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b4d-\u2b4f\u2b5a-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e9a\u2ef4-\u2eff', | |
'InKangxi_Radicals' => '\u2fd6-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ffc-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3003\u3005-\u3011\u3014-\u301f\u3021-\u3035\u3038-\u303d', | |
'InHiragana' => '\u3040-\u309a\u309d-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3192-\u3195', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u321f-\u3229\u3251-\u325f\u3280-\u3289\u32b1-\u32bf\u32ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua4c7-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb28\ufb2a-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdfb\ufdfe\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe61\ufe63\ufe67\ufe68\ufe6a-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff03\uff05-\uff0a\uff0c-\uff1b\uff1f-\uff3d\uff3f\uff41-\uff5b\uff5d\uff5f-\uffdf\uffe7\uffef', | |
'InSpecials' => '\ufff0-\ufffb\ufffe\uffff', | |
), | |
'Sc' => array( | |
'InBasic_Latin' => '\u0000-\u0023\u0025-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00a1\u00a6-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u060a\u060c-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09f1\u09f4-\u09fa\u09fc-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0af0\u0af2-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bf8\u0bfa-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e3e\u0e40-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17da\u17dc-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20b9-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdfb\ufdfd-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe68\ufe6a-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff03\uff05-\uffdf\uffe2-\uffe4\uffe7-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Sk' => array( | |
'InBasic_Latin' => '\u0000-\u005d\u005f\u0061-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00a7\u00a9-\u00ae\u00b0-\u00b3\u00b5-\u00b7\u00b9-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02c1\u02c6-\u02d1\u02e0-\u02e4\u02ec\u02ee', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u0374\u0376-\u0383\u0386-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fbc\u1fbe\u1fc2-\u1fcc\u1fd0-\u1fdc\u1fe0-\u1fec\u1ff0-\u1ffc\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309a\u309d-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff3d\uff3f\uff41-\uffe2\uffe4-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Sm' => array( | |
'InBasic_Latin' => '\u0000-\u002a\u002c-\u003b\u003f-\u007b\u007d\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ab\u00ad-\u00b0\u00b2-\u00d6\u00d8-\u00f6\u00f8-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03f5\u03f7-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u0605\u0609-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u2043\u2045-\u2051\u2053-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u2079\u207d-\u2089\u208d-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u213f\u2145-\u214a\u214c-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2195-\u2199\u219c-\u219f\u21a1\u21a2\u21a4\u21a5\u21a7-\u21ad\u21af-\u21cd\u21d0\u21d1\u21d3\u21d5-\u21f3', | |
'InMiscellaneous_Technical' => '\u2300-\u2307\u230c-\u231f\u2322-\u237b\u237d-\u239a\u23b4-\u23db\u23e2-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25b6\u25b8-\u25c0\u25c2-\u25f7', | |
'InMiscellaneous_Symbols' => '\u2600-\u266e\u2670-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c5\u27c6\u27cb\u27cd-\u27cf\u27e6-\u27ef', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2983-\u2998\u29d8-\u29db\u29fc\u29fd', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2b2f\u2b45\u2b46\u2b4d-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb28\ufb2a-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe61\ufe63\ufe67-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uff0a\uff0c-\uff1b\uff1f-\uff5b\uff5d\uff5f-\uffe1\uffe3-\uffe8\uffed-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'So' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00a5\u00a8\u00aa-\u00ad\u00af\u00b1-\u00b5\u00b7-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u0481\u0483-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u060d\u0610-\u06e8\u06ea-\u06fc\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09f9\u09fb-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b6f\u0b71-\u0b7f', | |
'InTamil' => '\u0b80-\u0bf2\u0bf9\u0bfb-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7e', | |
'InKannada' => '\u0c80-\u0cf0\u0cf3-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d78\u0d7a-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00\u0f04-\u0f12\u0f18\u0f19\u0f20-\u0f33\u0f35\u0f37\u0f39-\u0fbd\u0fc6\u0fcd\u0fd0-\u0fd4\u0fd9-\u0fff', | |
'InMyanmar' => '\u1000-\u109d', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u135f\u1361-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u193f\u1941-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2102\u2107\u210a-\u2113\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u212f-\u2139\u213c-\u2149\u214b\u214e', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u2194\u219a\u219b\u21a0\u21a3\u21a6\u21ae\u21ce\u21cf\u21d2\u21d4\u21f4-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2308-\u230b\u2320\u2321\u2329\u232a\u237c\u239b-\u23b3\u23dc-\u23e1\u23e9-\u23ff', | |
'InControl_Pictures' => '\u2427-\u243f', | |
'InOptical_Character_Recognition' => '\u244b-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u249b\u24ea-\u24ff', | |
'InGeometric_Shapes' => '\u25b7\u25c1\u25f8-\u25ff', | |
'InMiscellaneous_Symbols' => '\u266f\u26ce\u26e2\u26e4-\u26e7', | |
'InDingbats' => '\u2700\u2705\u270a\u270b\u2728\u274c\u274e\u2753-\u2755\u275f\u2760\u2768-\u2793\u2795-\u2797\u27b0\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b30-\u2b44\u2b47-\u2b4f\u2b5a-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e9a\u2ef4-\u2eff', | |
'InKangxi_Radicals' => '\u2fd6-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ffc-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u3003\u3005-\u3011\u3014-\u301f\u3021-\u3035\u3038-\u303d', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3192-\u3195', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u321f-\u3229\u3251-\u325f\u3280-\u3289\u32b1-\u32bf\u32ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua4c7-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdfc\ufdfe\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffe3\uffe5-\uffe7\uffe9-\uffec\uffef', | |
'InSpecials' => '\ufff0-\ufffb\ufffe\uffff', | |
), | |
'Z' => array( | |
'InBasic_Latin' => '\u0000-\u001f\u0021-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u009f\u00a1-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1681-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u180d\u180f-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u200b-\u2027\u202a-\u202e\u2030-\u205e\u2060-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3001-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Zl' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u2027\u2029-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Zp' => array( | |
'InBasic_Latin' => '\u0000-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1680-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u2000-\u2028\u202a-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3000-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
), | |
'Zs' => array( | |
'InBasic_Latin' => '\u0000-\u001f\u0021-\u007f', | |
'InLatin-1_Supplement' => '\u0080-\u009f\u00a1-\u00ff', | |
'InLatin_Extended-A' => '\u0100-\u017f', | |
'InLatin_Extended-B' => '\u0180-\u024f', | |
'InIPA_Extensions' => '\u0250-\u02af', | |
'InSpacing_Modifier_Letters' => '\u02b0-\u02ff', | |
'InCombining_Diacritical_Marks' => '\u0300-\u036f', | |
'InGreek_and_Coptic' => '\u0370-\u03ff', | |
'InCyrillic' => '\u0400-\u04ff', | |
'InCyrillic_Supplementary' => '\u0500-\u052f', | |
'InArmenian' => '\u0530-\u058f', | |
'InHebrew' => '\u0590-\u05ff', | |
'InArabic' => '\u0600-\u06ff', | |
'InSyriac' => '\u0700-\u074f', | |
'InThaana' => '\u0780-\u07bf', | |
'InDevanagari' => '\u0900-\u097f', | |
'InBengali' => '\u0980-\u09ff', | |
'InGurmukhi' => '\u0a00-\u0a7f', | |
'InGujarati' => '\u0a80-\u0aff', | |
'InOriya' => '\u0b00-\u0b7f', | |
'InTamil' => '\u0b80-\u0bff', | |
'InTelugu' => '\u0c00-\u0c7f', | |
'InKannada' => '\u0c80-\u0cff', | |
'InMalayalam' => '\u0d00-\u0d7f', | |
'InSinhala' => '\u0d80-\u0dff', | |
'InThai' => '\u0e00-\u0e7f', | |
'InLao' => '\u0e80-\u0eff', | |
'InTibetan' => '\u0f00-\u0fff', | |
'InMyanmar' => '\u1000-\u109f', | |
'InGeorgian' => '\u10a0-\u10ff', | |
'InHangul_Jamo' => '\u1100-\u11ff', | |
'InEthiopic' => '\u1200-\u137f', | |
'InCherokee' => '\u13a0-\u13ff', | |
'InUnified_Canadian_Aboriginal_Syllabics' => '\u1400-\u167f', | |
'InOgham' => '\u1681-\u169f', | |
'InRunic' => '\u16a0-\u16ff', | |
'InTagalog' => '\u1700-\u171f', | |
'InHanunoo' => '\u1720-\u173f', | |
'InBuhid' => '\u1740-\u175f', | |
'InTagbanwa' => '\u1760-\u177f', | |
'InKhmer' => '\u1780-\u17ff', | |
'InMongolian' => '\u1800-\u180d\u180f-\u18af', | |
'InLimbu' => '\u1900-\u194f', | |
'InTai_Le' => '\u1950-\u197f', | |
'InKhmer_Symbols' => '\u19e0-\u19ff', | |
'InPhonetic_Extensions' => '\u1d00-\u1d7f', | |
'InLatin_Extended_Additional' => '\u1e00-\u1eff', | |
'InGreek_Extended' => '\u1f00-\u1fff', | |
'InGeneral_Punctuation' => '\u200b-\u202e\u2030-\u205e\u2060-\u206f', | |
'InSuperscripts_and_Subscripts' => '\u2070-\u209f', | |
'InCurrency_Symbols' => '\u20a0-\u20cf', | |
'InCombining_Diacritical_Marks_for_Symbols' => '\u20d0-\u20ff', | |
'InLetterlike_Symbols' => '\u2100-\u214f', | |
'InNumber_Forms' => '\u2150-\u218f', | |
'InArrows' => '\u2190-\u21ff', | |
'InMathematical_Operators' => '\u2200-\u22ff', | |
'InMiscellaneous_Technical' => '\u2300-\u23ff', | |
'InControl_Pictures' => '\u2400-\u243f', | |
'InOptical_Character_Recognition' => '\u2440-\u245f', | |
'InEnclosed_Alphanumerics' => '\u2460-\u24ff', | |
'InBox_Drawing' => '\u2500-\u257f', | |
'InBlock_Elements' => '\u2580-\u259f', | |
'InGeometric_Shapes' => '\u25a0-\u25ff', | |
'InMiscellaneous_Symbols' => '\u2600-\u26ff', | |
'InDingbats' => '\u2700-\u27bf', | |
'InMiscellaneous_Mathematical_Symbols-A' => '\u27c0-\u27ef', | |
'InSupplemental_Arrows-A' => '\u27f0-\u27ff', | |
'InBraille_Patterns' => '\u2800-\u28ff', | |
'InSupplemental_Arrows-B' => '\u2900-\u297f', | |
'InMiscellaneous_Mathematical_Symbols-B' => '\u2980-\u29ff', | |
'InSupplemental_Mathematical_Operators' => '\u2a00-\u2aff', | |
'InMiscellaneous_Symbols_and_Arrows' => '\u2b00-\u2bff', | |
'InCJK_Radicals_Supplement' => '\u2e80-\u2eff', | |
'InKangxi_Radicals' => '\u2f00-\u2fdf', | |
'InIdeographic_Description_Characters' => '\u2ff0-\u2fff', | |
'InCJK_Symbols_and_Punctuation' => '\u3001-\u303f', | |
'InHiragana' => '\u3040-\u309f', | |
'InKatakana' => '\u30a0-\u30ff', | |
'InBopomofo' => '\u3100-\u312f', | |
'InHangul_Compatibility_Jamo' => '\u3130-\u318f', | |
'InKanbun' => '\u3190-\u319f', | |
'InBopomofo_Extended' => '\u31a0-\u31bf', | |
'InKatakana_Phonetic_Extensions' => '\u31f0-\u31ff', | |
'InEnclosed_CJK_Letters_and_Months' => '\u3200-\u32ff', | |
'InCJK_Compatibility' => '\u3300-\u33ff', | |
'InCJK_Unified_Ideographs_Extension_A' => '\u3400-\u4dbf', | |
'InYijing_Hexagram_Symbols' => '\u4dc0-\u4dff', | |
'InCJK_Unified_Ideographs' => '\u4e00-\u9fff', | |
'InYi_Syllables' => '\ua000-\ua48f', | |
'InYi_Radicals' => '\ua490-\ua4cf', | |
'InHangul_Syllables' => '\uac00-\ud7af', | |
'InPrivate_Use_Area' => '\ue000-\uf8ff', | |
'InCJK_Compatibility_Ideographs' => '\uf900-\ufaff', | |
'InAlphabetic_Presentation_Forms' => '\ufb00-\ufb4f', | |
'InArabic_Presentation_Forms-A' => '\ufb50-\ufdff', | |
'InVariation_Selectors' => '\ufe00-\ufe0f', | |
'InCombining_Half_Marks' => '\ufe20-\ufe2f', | |
'InCJK_Compatibility_Forms' => '\ufe30-\ufe4f', | |
'InSmall_Form_Variants' => '\ufe50-\ufe6f', | |
'InArabic_Presentation_Forms-B' => '\ufe70-\ufeff', | |
'InHalfwidth_and_Fullwidth_Forms' => '\uff00-\uffef', | |
'InSpecials' => '\ufff0-\uffff', | |
) | |
); | |
/** | |
* Обработчик расширенной последовательности Unicode | |
* | |
* @param mixed $match | |
* @return string | |
*/ | |
private static function upper_X_Handler( $match ) { | |
$result = implode( "", self::$rules[ 'X' ] ); | |
return ( self::$pcre_tag_type == 0 ? '['.$result.']' : $result ).isset( $match[ 4 ] ) ? $match[ 4 ] : ''; | |
} | |
/** | |
* Обработчик свойств Unicode | |
* | |
* @param mixed $match | |
* @return string | |
*/ | |
private static function upper_lower_P_Handler( $match ) { | |
// определяем отрицание | |
$not = ( isset( $match[ 5 ] ) && $match[ 5 ] == '^' ) ? true : false; | |
// получаем нужный массив с правилами | |
$rules = ( $match[ 3 ] == '\\p' && !$not ) || ( $match[ 3 ] == '\\P' && $not ) ? self::$equalRules : self::$notEqualRules; | |
// формируем модификатор | |
$modificator = isset( $match[ 8 ] ) ? $match[ 8 ] : $match[ 6 ].( isset( $match[ 7 ] ) ? $match[ 7 ] : '' ); | |
// по умолчанию вернем все как есть не меняя | |
$result = $match[ 0 ]; | |
if ( isset( $rules[ $modificator ] ) ) { | |
$result = implode( "", $rules[ $modificator ] ); | |
// если мы не внутри класса, организуем его | |
if ( self::$pcre_tag_type == 0 ) { | |
$result = '['.$result.']'; | |
} | |
} | |
return $result; | |
} | |
/** | |
* Конвертирует регулярное выражение PHP в JavaScript совместимое | |
* | |
* @param string $pcre_string | |
* @return string | |
*/ | |
public static function toJS( $pcre_string ) { | |
return preg_replace_callback( | |
/** | |
* Рега заменяет открывающие и закрывающие символы, | |
* нормальным принятым в PCRE символом слеша '/' | |
* так же эта рега обрабатывает модификаторы | |
*/ | |
'/^(.)([\s\S]+)\\1([^\\1]*)$/', | |
/** | |
* Обработчик модификаторов | |
*/ | |
'self::fromPHPModifier', | |
/** | |
* Сканирование реги на наличие PHP-only совместимых правил | |
*/ | |
preg_replace_callback( | |
/** | |
* рега ищет специальное свойство юникод-символов | |
*/ | |
'#(?:(\x5C\x5C|\x5C(?:\[|\])|([\[\]]))|(\x5C[pPX])(?:({([\^]?)([CLMNPSZ])([cfnoslmtuedikp])?}|([CLMNPSZ])))?)#', | |
/** | |
* обработчик всего найденого | |
*/ | |
'self::processPHP_PCRE', | |
(string)$pcre_string | |
) | |
); | |
} | |
/** | |
* Внутренний метод обработки внешних модификаторов | |
* | |
* @param mixed $match | |
* @return string | |
*/ | |
private static function fromPHPModifier( $match ) { | |
if ( !empty( $match[ 3 ] ) ) { | |
$modifiers = preg_split( '//', $match[ 3 ], -1, PREG_SPLIT_NO_EMPTY ); | |
$match[ 3 ] = ''; | |
while( $modifier = array_pop( $modifiers ) ) { | |
switch( $modifier ) { | |
case 'i': | |
case 'm': | |
$match[ 3 ] .= $modifier; | |
default: | |
} | |
} | |
} | |
return '/'.$match[ 2 ].'/'.$match[ 3 ]; | |
} | |
/** | |
* Основной процессор преобразования | |
* | |
* @param mixed $match | |
* @return string | |
*/ | |
private static function processPHP_PCRE( $match ) { | |
if ( !empty( $match[ 2 ] ) ) { | |
self::$pcre_tag_type = $match[ 2 ] === '[' ? 1 : 2; | |
} | |
// проверяем наличие обработчика | |
$handler = isset( $match[ 3 ], self::$handlers[ $match[ 3 ] ] ) ? self::$handlers[ $match[ 3 ] ] : false; | |
if ( $handler ) { | |
// тут запускаем обработчик если его нашли | |
$result = call_user_func( 'self::'.$handler, $match ); | |
} else { | |
// тут просто ничего не делаем а оставляем все как есть | |
$result = $match[ 1 ].( isset( $match[ 3 ] ) ? $match[ 3 ] : '' ).( isset( $match[ 4 ] ) ? $match[ 4 ] : '' ); | |
} | |
if ( self::$pcre_tag_type === 2 ) { | |
self::$pcre_tag_type = 0; | |
} | |
return $result; | |
} | |
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
<?php | |
// преобразование PHP регулярки в JS регулярку | |
$regular = PCREConverter::toJS( "/^[a-zA-Z\s,!.'-\pL]+$/u" ); | |
?> | |
<script type="text/javascript"> | |
var match = <?php echo $regular;?>.test( "Привет Мир!!!" ); | |
alert( "Result: " + match ); | |
</script> |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment