Mercurial Repositories > eric / file comparison

comparison: ThirdParty/Pygments/pygments/lexers/jvm.py

ThirdParty/Pygments/pygments/lexers/jvm.py

changeset 1705
b0fbc9300f2b
child 2426
da76c71624de
equal deleted inserted replaced
1704:02ae6c55b35b 1705:b0fbc9300f2b
1 # -*- coding: utf-8 -*-
2 """
3 pygments.lexers.jvm
4 ~~~~~~~~~~~~~~~~~~~
5
6 Pygments lexers for JVM languages.
7
8 :copyright: Copyright 2006-2012 by the Pygments team, see AUTHORS.
9 :license: BSD, see LICENSE for details.
10 """
11
12 import re
13
14 from pygments.lexer import Lexer, RegexLexer, include, bygroups, using, \
15 this
16 from pygments.token import Text, Comment, Operator, Keyword, Name, String, \
17 Number, Punctuation
18 from pygments.util import get_choice_opt
19 from pygments import unistring as uni
20
21
22 __all__ = ['JavaLexer', 'ScalaLexer', 'GosuLexer', 'GosuTemplateLexer',
23 'GroovyLexer', 'IokeLexer', 'ClojureLexer', 'KotlinLexer']
24
25
26 class JavaLexer(RegexLexer):
27 """
28 For `Java <http://www.sun.com/java/>`_ source code.
29 """
30
31 name = 'Java'
32 aliases = ['java']
33 filenames = ['*.java']
34 mimetypes = ['text/x-java']
35
36 flags = re.MULTILINE | re.DOTALL
37
38 #: optional Comment or Whitespace
39 _ws = r'(?:\s|//.*?\n|/[*].*?[*]/)+'
40
41 tokens = {
42 'root': [
43 # method names
44 (r'^(\s*(?:[a-zA-Z_][a-zA-Z0-9_\.\[\]]*\s+)+?)' # return arguments
45 r'([a-zA-Z_][a-zA-Z0-9_]*)' # method name
46 r'(\s*)(\()', # signature start
47 bygroups(using(this), Name.Function, Text, Operator)),
48 (r'[^\S\n]+', Text),
49 (r'//.*?\n', Comment.Single),
50 (r'/\*.*?\*/', Comment.Multiline),
51 (r'@[a-zA-Z_][a-zA-Z0-9_\.]*', Name.Decorator),
52 (r'(assert|break|case|catch|continue|default|do|else|finally|for|'
53 r'if|goto|instanceof|new|return|switch|this|throw|try|while)\b',
54 Keyword),
55 (r'(abstract|const|enum|extends|final|implements|native|private|'
56 r'protected|public|static|strictfp|super|synchronized|throws|'
57 r'transient|volatile)\b', Keyword.Declaration),
58 (r'(boolean|byte|char|double|float|int|long|short|void)\b',
59 Keyword.Type),
60 (r'(package)(\s+)', bygroups(Keyword.Namespace, Text)),
61 (r'(true|false|null)\b', Keyword.Constant),
62 (r'(class|interface)(\s+)', bygroups(Keyword.Declaration, Text), 'class'),
63 (r'(import)(\s+)', bygroups(Keyword.Namespace, Text), 'import'),
64 (r'"(\\\\|\\"|[^"])*"', String),
65 (r"'\\.'|'[^\\]'|'\\u[0-9a-f]{4}'", String.Char),
66 (r'(\.)([a-zA-Z_][a-zA-Z0-9_]*)', bygroups(Operator, Name.Attribute)),
67 (r'[a-zA-Z_][a-zA-Z0-9_]*:', Name.Label),
68 (r'[a-zA-Z_\$][a-zA-Z0-9_]*', Name),
69 (r'[~\^\*!%&\[\]\(\)\{\}<>\|+=:;,./?-]', Operator),
70 (r'[0-9][0-9]*\.[0-9]+([eE][0-9]+)?[fd]?', Number.Float),
71 (r'0x[0-9a-f]+', Number.Hex),
72 (r'[0-9]+L?', Number.Integer),
73 (r'\n', Text)
74 ],
75 'class': [
76 (r'[a-zA-Z_][a-zA-Z0-9_]*', Name.Class, '#pop')
77 ],
78 'import': [
79 (r'[a-zA-Z0-9_.]+\*?', Name.Namespace, '#pop')
80 ],
81 }
82
83
84 class ScalaLexer(RegexLexer):
85 """
86 For `Scala <http://www.scala-lang.org>`_ source code.
87 """
88
89 name = 'Scala'
90 aliases = ['scala']
91 filenames = ['*.scala']
92 mimetypes = ['text/x-scala']
93
94 flags = re.MULTILINE | re.DOTALL
95
96 #: optional Comment or Whitespace
97 _ws = r'(?:\s|//.*?\n|/[*].*?[*]/)+'
98
99 # don't use raw unicode strings!
100 op = '[-~\\^\\*!%&\\\\<>\\|+=:/?@\u00a6-\u00a7\u00a9\u00ac\u00ae\u00b0-\u00b1\u00b6\u00d7\u00f7\u03f6\u0482\u0606-\u0608\u060e-\u060f\u06e9\u06fd-\u06fe\u07f6\u09fa\u0b70\u0bf3-\u0bf8\u0bfa\u0c7f\u0cf1-\u0cf2\u0d79\u0f01-\u0f03\u0f13-\u0f17\u0f1a-\u0f1f\u0f34\u0f36\u0f38\u0fbe-\u0fc5\u0fc7-\u0fcf\u109e-\u109f\u1360\u1390-\u1399\u1940\u19e0-\u19ff\u1b61-\u1b6a\u1b74-\u1b7c\u2044\u2052\u207a-\u207c\u208a-\u208c\u2100-\u2101\u2103-\u2106\u2108-\u2109\u2114\u2116-\u2118\u211e-\u2123\u2125\u2127\u2129\u212e\u213a-\u213b\u2140-\u2144\u214a-\u214d\u214f\u2190-\u2328\u232b-\u244a\u249c-\u24e9\u2500-\u2767\u2794-\u27c4\u27c7-\u27e5\u27f0-\u2982\u2999-\u29d7\u29dc-\u29fb\u29fe-\u2b54\u2ce5-\u2cea\u2e80-\u2ffb\u3004\u3012-\u3013\u3020\u3036-\u3037\u303e-\u303f\u3190-\u3191\u3196-\u319f\u31c0-\u31e3\u3200-\u321e\u322a-\u3250\u3260-\u327f\u328a-\u32b0\u32c0-\u33ff\u4dc0-\u4dff\ua490-\ua4c6\ua828-\ua82b\ufb29\ufdfd\ufe62\ufe64-\ufe66\uff0b\uff1c-\uff1e\uff5c\uff5e\uffe2\uffe4\uffe8-\uffee\ufffc-\ufffd]+'
101
102 letter = '[a-zA-Z\\$_\u00aa\u00b5\u00ba\u00c0-\u00d6\u00d8-\u00f6\u00f8-\u02af\u0370-\u0373\u0376-\u0377\u037b-\u037d\u0386\u0388-\u03f5\u03f7-\u0481\u048a-\u0556\u0561-\u0587\u05d0-\u05f2\u0621-\u063f\u0641-\u064a\u066e-\u066f\u0671-\u06d3\u06d5\u06ee-\u06ef\u06fa-\u06fc\u06ff\u0710\u0712-\u072f\u074d-\u07a5\u07b1\u07ca-\u07ea\u0904-\u0939\u093d\u0950\u0958-\u0961\u0972-\u097f\u0985-\u09b9\u09bd\u09ce\u09dc-\u09e1\u09f0-\u09f1\u0a05-\u0a39\u0a59-\u0a5e\u0a72-\u0a74\u0a85-\u0ab9\u0abd\u0ad0-\u0ae1\u0b05-\u0b39\u0b3d\u0b5c-\u0b61\u0b71\u0b83-\u0bb9\u0bd0\u0c05-\u0c3d\u0c58-\u0c61\u0c85-\u0cb9\u0cbd\u0cde-\u0ce1\u0d05-\u0d3d\u0d60-\u0d61\u0d7a-\u0d7f\u0d85-\u0dc6\u0e01-\u0e30\u0e32-\u0e33\u0e40-\u0e45\u0e81-\u0eb0\u0eb2-\u0eb3\u0ebd-\u0ec4\u0edc-\u0f00\u0f40-\u0f6c\u0f88-\u0f8b\u1000-\u102a\u103f\u1050-\u1055\u105a-\u105d\u1061\u1065-\u1066\u106e-\u1070\u1075-\u1081\u108e\u10a0-\u10fa\u1100-\u135a\u1380-\u138f\u13a0-\u166c\u166f-\u1676\u1681-\u169a\u16a0-\u16ea\u16ee-\u1711\u1720-\u1731\u1740-\u1751\u1760-\u1770\u1780-\u17b3\u17dc\u1820-\u1842\u1844-\u18a8\u18aa-\u191c\u1950-\u19a9\u19c1-\u19c7\u1a00-\u1a16\u1b05-\u1b33\u1b45-\u1b4b\u1b83-\u1ba0\u1bae-\u1baf\u1c00-\u1c23\u1c4d-\u1c4f\u1c5a-\u1c77\u1d00-\u1d2b\u1d62-\u1d77\u1d79-\u1d9a\u1e00-\u1fbc\u1fbe\u1fc2-\u1fcc\u1fd0-\u1fdb\u1fe0-\u1fec\u1ff2-\u1ffc\u2071\u207f\u2102\u2107\u210a-\u2113\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u212f-\u2139\u213c-\u213f\u2145-\u2149\u214e\u2160-\u2188\u2c00-\u2c7c\u2c80-\u2ce4\u2d00-\u2d65\u2d80-\u2dde\u3006-\u3007\u3021-\u3029\u3038-\u303a\u303c\u3041-\u3096\u309f\u30a1-\u30fa\u30ff-\u318e\u31a0-\u31b7\u31f0-\u31ff\u3400-\u4db5\u4e00-\ua014\ua016-\ua48c\ua500-\ua60b\ua610-\ua61f\ua62a-\ua66e\ua680-\ua697\ua722-\ua76f\ua771-\ua787\ua78b-\ua801\ua803-\ua805\ua807-\ua80a\ua80c-\ua822\ua840-\ua873\ua882-\ua8b3\ua90a-\ua925\ua930-\ua946\uaa00-\uaa28\uaa40-\uaa42\uaa44-\uaa4b\uac00-\ud7a3\uf900-\ufb1d\ufb1f-\ufb28\ufb2a-\ufd3d\ufd50-\ufdfb\ufe70-\ufefc\uff21-\uff3a\uff41-\uff5a\uff66-\uff6f\uff71-\uff9d\uffa0-\uffdc]'
103
104 upper = '[A-Z\\$_\u00c0-\u00d6\u00d8-\u00de\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\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\u0370\u0372\u0376\u0386\u0388-\u038f\u0391-\u03ab\u03cf\u03d2-\u03d4\u03d8\u03da\u03dc\u03de\u03e0\u03e2\u03e4\u03e6\u03e8\u03ea\u03ec\u03ee\u03f4\u03f7\u03f9-\u03fa\u03fd-\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\u0500\u0502\u0504\u0506\u0508\u050a\u050c\u050e\u0510\u0512\u0514\u0516\u0518\u051a\u051c\u051e\u0520\u0522\u0531-\u0556\u10a0-\u10c5\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\u1f08-\u1f0f\u1f18-\u1f1d\u1f28-\u1f2f\u1f38-\u1f3f\u1f48-\u1f4d\u1f59-\u1f5f\u1f68-\u1f6f\u1fb8-\u1fbb\u1fc8-\u1fcb\u1fd8-\u1fdb\u1fe8-\u1fec\u1ff8-\u1ffb\u2102\u2107\u210b-\u210d\u2110-\u2112\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u2130-\u2133\u213e-\u213f\u2145\u2183\u2c00-\u2c2e\u2c60\u2c62-\u2c64\u2c67\u2c69\u2c6b\u2c6d-\u2c6f\u2c72\u2c75\u2c80\u2c82\u2c84\u2c86\u2c88\u2c8a\u2c8c\u2c8e\u2c90\u2c92\u2c94\u2c96\u2c98\u2c9a\u2c9c\u2c9e\u2ca0\u2ca2\u2ca4\u2ca6\u2ca8\u2caa\u2cac\u2cae\u2cb0\u2cb2\u2cb4\u2cb6\u2cb8\u2cba\u2cbc\u2cbe\u2cc0\u2cc2\u2cc4\u2cc6\u2cc8\u2cca\u2ccc\u2cce\u2cd0\u2cd2\u2cd4\u2cd6\u2cd8\u2cda\u2cdc\u2cde\u2ce0\u2ce2\ua640\ua642\ua644\ua646\ua648\ua64a\ua64c\ua64e\ua650\ua652\ua654\ua656\ua658\ua65a\ua65c\ua65e\ua662\ua664\ua666\ua668\ua66a\ua66c\ua680\ua682\ua684\ua686\ua688\ua68a\ua68c\ua68e\ua690\ua692\ua694\ua696\ua722\ua724\ua726\ua728\ua72a\ua72c\ua72e\ua732\ua734\ua736\ua738\ua73a\ua73c\ua73e\ua740\ua742\ua744\ua746\ua748\ua74a\ua74c\ua74e\ua750\ua752\ua754\ua756\ua758\ua75a\ua75c\ua75e\ua760\ua762\ua764\ua766\ua768\ua76a\ua76c\ua76e\ua779\ua77b\ua77d-\ua77e\ua780\ua782\ua784\ua786\ua78b\uff21-\uff3a]'
105
106 idrest = r'%s(?:%s|[0-9])*(?:(?<=_)%s)?' % (letter, letter, op)
107
108 tokens = {
109 'root': [
110 # method names
111 (r'(class|trait|object)(\s+)', bygroups(Keyword, Text), 'class'),
112 (r"'%s" % idrest, Text.Symbol),
113 (r'[^\S\n]+', Text),
114 (r'//.*?\n', Comment.Single),
115 (r'/\*', Comment.Multiline, 'comment'),
116 (r'@%s' % idrest, Name.Decorator),
117 (r'(abstract|ca(?:se|tch)|d(?:ef|o)|e(?:lse|xtends)|'
118 r'f(?:inal(?:ly)?|or(?:Some)?)|i(?:f|mplicit)|'
119 r'lazy|match|new|override|pr(?:ivate|otected)'
120 r'|re(?:quires|turn)|s(?:ealed|uper)|'
121 r't(?:h(?:is|row)|ry)|va[lr]|w(?:hile|ith)|yield)\b|'
122 '(<[%:-]|=>|>:|[#=@_\u21D2\u2190])(\\b|(?=\\s)|$)', Keyword),
123 (r':(?!%s)' % op, Keyword, 'type'),
124 (r'%s%s\b' % (upper, idrest), Name.Class),
125 (r'(true|false|null)\b', Keyword.Constant),
126 (r'(import|package)(\s+)', bygroups(Keyword, Text), 'import'),
127 (r'(type)(\s+)', bygroups(Keyword, Text), 'type'),
128 (r'""".*?"""', String),
129 (r'"(\\\\|\\"|[^"])*"', String),
130 (r"'\\.'|'[^\\]'|'\\u[0-9a-f]{4}'", String.Char),
131 # (ur'(\.)(%s|%s|`[^`]+`)' % (idrest, op), bygroups(Operator,
132 # Name.Attribute)),
133 (idrest, Name),
134 (r'`[^`]+`', Name),
135 (r'\[', Operator, 'typeparam'),
136 (r'[\(\)\{\};,.#]', Operator),
137 (op, Operator),
138 (r'([0-9][0-9]*\.[0-9]*|\.[0-9]+)([eE][+-]?[0-9]+)?[fFdD]?',
139 Number.Float),
140 (r'0x[0-9a-f]+', Number.Hex),
141 (r'[0-9]+L?', Number.Integer),
142 (r'\n', Text)
143 ],
144 'class': [
145 (r'(%s|%s|`[^`]+`)(\s*)(\[)' % (idrest, op),
146 bygroups(Name.Class, Text, Operator), 'typeparam'),
147 (r'\s+', Text),
148 (r'{', Operator, '#pop'),
149 (r'\(', Operator, '#pop'),
150 (r'//.*?\n', Comment.Single, '#pop'),
151 (r'%s|%s|`[^`]+`' % (idrest, op), Name.Class, '#pop'),
152 ],
153 'type': [
154 (r'\s+', Text),
155 ('<[%:]|>:|[#_\u21D2]|forSome|type', Keyword),
156 (r'([,\);}]|=>|=)(\s*)', bygroups(Operator, Text), '#pop'),
157 (r'[\(\{]', Operator, '#push'),
158 (r'((?:%s|%s|`[^`]+`)(?:\.(?:%s|%s|`[^`]+`))*)(\s*)(\[)' %
159 (idrest, op, idrest, op),
160 bygroups(Keyword.Type, Text, Operator), ('#pop', 'typeparam')),
161 (r'((?:%s|%s|`[^`]+`)(?:\.(?:%s|%s|`[^`]+`))*)(\s*)$' %
162 (idrest, op, idrest, op),
163 bygroups(Keyword.Type, Text), '#pop'),
164 (r'//.*?\n', Comment.Single, '#pop'),
165 (r'\.|%s|%s|`[^`]+`' % (idrest, op), Keyword.Type)
166 ],
167 'typeparam': [
168 (r'[\s,]+', Text),
169 ('<[%:]|=>|>:|[#_\u21D2]|forSome|type', Keyword),
170 (r'([\]\)\}])', Operator, '#pop'),
171 (r'[\(\[\{]', Operator, '#push'),
172 (r'\.|%s|%s|`[^`]+`' % (idrest, op), Keyword.Type)
173 ],
174 'comment': [
175 (r'[^/\*]+', Comment.Multiline),
176 (r'/\*', Comment.Multiline, '#push'),
177 (r'\*/', Comment.Multiline, '#pop'),
178 (r'[*/]', Comment.Multiline)
179 ],
180 'import': [
181 (r'(%s|\.)+' % idrest, Name.Namespace, '#pop')
182 ],
183 }
184
185
186 class GosuLexer(RegexLexer):
187 """
188 For Gosu source code.
189
190 *New in Pygments 1.5.*
191 """
192
193 name = 'Gosu'
194 aliases = ['gosu']
195 filenames = ['*.gs', '*.gsx', '*.gsp', '*.vark']
196 mimetypes = ['text/x-gosu']
197
198 flags = re.MULTILINE | re.DOTALL
199
200 #: optional Comment or Whitespace
201 _ws = r'(?:\s|//.*?\n|/[*].*?[*]/)+'
202
203 tokens = {
204 'root': [
205 # method names
206 (r'^(\s*(?:[a-zA-Z_][a-zA-Z0-9_\.\[\]]*\s+)+?)' # modifiers etc.
207 r'([a-zA-Z_][a-zA-Z0-9_]*)' # method name
208 r'(\s*)(\()', # signature start
209 bygroups(using(this), Name.Function, Text, Operator)),
210 (r'[^\S\n]+', Text),
211 (r'//.*?\n', Comment.Single),
212 (r'/\*.*?\*/', Comment.Multiline),
213 (r'@[a-zA-Z_][a-zA-Z0-9_\.]*', Name.Decorator),
214 (r'(in|as|typeof|statictypeof|typeis|typeas|if|else|foreach|for|'
215 r'index|while|do|continue|break|return|try|catch|finally|this|'
216 r'throw|new|switch|case|default|eval|super|outer|classpath|'
217 r'using)\b', Keyword),
218 (r'(var|delegate|construct|function|private|internal|protected|'
219 r'public|abstract|override|final|static|extends|transient|'
220 r'implements|represents|readonly)\b', Keyword.Declaration),
221 (r'(property\s+)(get|set|)', Keyword.Declaration),
222 (r'(boolean|byte|char|double|float|int|long|short|void|block)\b',
223 Keyword.Type),
224 (r'(package)(\s+)', bygroups(Keyword.Namespace, Text)),
225 (r'(true|false|null|NaN|Infinity)\b', Keyword.Constant),
226 (r'(class|interface|enhancement|enum)(\s+)([a-zA-Z_][a-zA-Z0-9_]*)',
227 bygroups(Keyword.Declaration, Text, Name.Class)),
228 (r'(uses)(\s+)([a-zA-Z0-9_.]+\*?)',
229 bygroups(Keyword.Namespace, Text, Name.Namespace)),
230 (r'"', String, 'string'),
231 (r'(\??[\.#])([a-zA-Z_][a-zA-Z0-9_]*)',
232 bygroups(Operator, Name.Attribute)),
233 (r'(:)([a-zA-Z_][a-zA-Z0-9_]*)',
234 bygroups(Operator, Name.Attribute)),
235 (r'[a-zA-Z_\$][a-zA-Z0-9_]*', Name),
236 (r'and|or|not|[\\~\^\*!%&\[\]\(\)\{\}<>\|+=:;,./?-]', Operator),
237 (r'[0-9][0-9]*\.[0-9]+([eE][0-9]+)?[fd]?', Number.Float),
238 (r'[0-9]+', Number.Integer),
239 (r'\n', Text)
240 ],
241 'templateText': [
242 (r'(\\<)|(\\\$)', String),
243 (r'(<%@\s+)(extends|params)',
244 bygroups(Operator, Name.Decorator), 'stringTemplate'),
245 (r'<%!--.*?--%>', Comment.Multiline),
246 (r'(<%)|(<%=)', Operator, 'stringTemplate'),
247 (r'\$\{', Operator, 'stringTemplateShorthand'),
248 (r'.', String)
249 ],
250 'string': [
251 (r'"', String, '#pop'),
252 include('templateText')
253 ],
254 'stringTemplate': [
255 (r'"', String, 'string'),
256 (r'%>', Operator, '#pop'),
257 include('root')
258 ],
259 'stringTemplateShorthand': [
260 (r'"', String, 'string'),
261 (r'\{', Operator, 'stringTemplateShorthand'),
262 (r'\}', Operator, '#pop'),
263 include('root')
264 ],
265 }
266
267
268 class GosuTemplateLexer(Lexer):
269 """
270 For Gosu templates.
271
272 *New in Pygments 1.5.*
273 """
274
275 name = 'Gosu Template'
276 aliases = ['gst']
277 filenames = ['*.gst']
278 mimetypes = ['text/x-gosu-template']
279 lexer = GosuLexer()
280
281 def get_tokens_unprocessed(self, text):
282 stack = ['templateText']
283 for item in self.lexer.get_tokens_unprocessed(text, stack):
284 yield item
285
286
287 class GroovyLexer(RegexLexer):
288 """
289 For `Groovy <http://groovy.codehaus.org/>`_ source code.
290
291 *New in Pygments 1.5.*
292 """
293
294 name = 'Groovy'
295 aliases = ['groovy']
296 filenames = ['*.groovy']
297 mimetypes = ['text/x-groovy']
298
299 flags = re.MULTILINE | re.DOTALL
300
301 #: optional Comment or Whitespace
302 _ws = r'(?:\s|//.*?\n|/[*].*?[*]/)+'
303
304 tokens = {
305 'root': [
306 # method names
307 (r'^(\s*(?:[a-zA-Z_][a-zA-Z0-9_\.\[\]]*\s+)+?)' # return arguments
308 r'([a-zA-Z_][a-zA-Z0-9_]*)' # method name
309 r'(\s*)(\()', # signature start
310 bygroups(using(this), Name.Function, Text, Operator)),
311 (r'[^\S\n]+', Text),
312 (r'//.*?\n', Comment.Single),
313 (r'/\*.*?\*/', Comment.Multiline),
314 (r'@[a-zA-Z_][a-zA-Z0-9_\.]*', Name.Decorator),
315 (r'(assert|break|case|catch|continue|default|do|else|finally|for|'
316 r'if|goto|instanceof|new|return|switch|this|throw|try|while|in|as)\b',
317 Keyword),
318 (r'(abstract|const|enum|extends|final|implements|native|private|'
319 r'protected|public|static|strictfp|super|synchronized|throws|'
320 r'transient|volatile)\b', Keyword.Declaration),
321 (r'(def|boolean|byte|char|double|float|int|long|short|void)\b',
322 Keyword.Type),
323 (r'(package)(\s+)', bygroups(Keyword.Namespace, Text)),
324 (r'(true|false|null)\b', Keyword.Constant),
325 (r'(class|interface)(\s+)', bygroups(Keyword.Declaration, Text),
326 'class'),
327 (r'(import)(\s+)', bygroups(Keyword.Namespace, Text), 'import'),
328 (r'"(\\\\|\\"|[^"])*"', String.Double),
329 (r"'(\\\\|\\'|[^'])*'", String.Single),
330 (r'\$/((?!/\$).)*/\$', String),
331 (r'/(\\\\|\\"|[^/])*/', String),
332 (r"'\\.'|'[^\\]'|'\\u[0-9a-f]{4}'", String.Char),
333 (r'(\.)([a-zA-Z_][a-zA-Z0-9_]*)', bygroups(Operator, Name.Attribute)),
334 (r'[a-zA-Z_][a-zA-Z0-9_]*:', Name.Label),
335 (r'[a-zA-Z_\$][a-zA-Z0-9_]*', Name),
336 (r'[~\^\*!%&\[\]\(\)\{\}<>\|+=:;,./?-]', Operator),
337 (r'[0-9][0-9]*\.[0-9]+([eE][0-9]+)?[fd]?', Number.Float),
338 (r'0x[0-9a-f]+', Number.Hex),
339 (r'[0-9]+L?', Number.Integer),
340 (r'\n', Text)
341 ],
342 'class': [
343 (r'[a-zA-Z_][a-zA-Z0-9_]*', Name.Class, '#pop')
344 ],
345 'import': [
346 (r'[a-zA-Z0-9_.]+\*?', Name.Namespace, '#pop')
347 ],
348 }
349
350
351 class IokeLexer(RegexLexer):
352 """
353 For `Ioke <http://ioke.org/>`_ (a strongly typed, dynamic,
354 prototype based programming language) source.
355
356 *New in Pygments 1.4.*
357 """
358 name = 'Ioke'
359 filenames = ['*.ik']
360 aliases = ['ioke', 'ik']
361 mimetypes = ['text/x-iokesrc']
362 tokens = {
363 'interpolatableText': [
364 (r'(\\b|\\e|\\t|\\n|\\f|\\r|\\"|\\\\|\\#|\\\Z|\\u[0-9a-fA-F]{1,4}'
365 r'|\\[0-3]?[0-7]?[0-7])', String.Escape),
366 (r'#{', Punctuation, 'textInterpolationRoot')
367 ],
368
369 'text': [
370 (r'(?<!\\)"', String, '#pop'),
371 include('interpolatableText'),
372 (r'[^"]', String)
373 ],
374
375 'documentation': [
376 (r'(?<!\\)"', String.Doc, '#pop'),
377 include('interpolatableText'),
378 (r'[^"]', String.Doc)
379 ],
380
381 'textInterpolationRoot': [
382 (r'}', Punctuation, '#pop'),
383 include('root')
384 ],
385
386 'slashRegexp': [
387 (r'(?<!\\)/[oxpniums]*', String.Regex, '#pop'),
388 include('interpolatableText'),
389 (r'\\/', String.Regex),
390 (r'[^/]', String.Regex)
391 ],
392
393 'squareRegexp': [
394 (r'(?<!\\)][oxpniums]*', String.Regex, '#pop'),
395 include('interpolatableText'),
396 (r'\\]', String.Regex),
397 (r'[^\]]', String.Regex)
398 ],
399
400 'squareText': [
401 (r'(?<!\\)]', String, '#pop'),
402 include('interpolatableText'),
403 (r'[^\]]', String)
404 ],
405
406 'root': [
407 (r'\n', Text),
408 (r'\s+', Text),
409
410 # Comments
411 (r';(.*?)\n', Comment),
412 (r'\A#!(.*?)\n', Comment),
413
414 #Regexps
415 (r'#/', String.Regex, 'slashRegexp'),
416 (r'#r\[', String.Regex, 'squareRegexp'),
417
418 #Symbols
419 (r':[a-zA-Z0-9_!:?]+', String.Symbol),
420 (r'[a-zA-Z0-9_!:?]+:(?![a-zA-Z0-9_!?])', String.Other),
421 (r':"(\\\\|\\"|[^"])*"', String.Symbol),
422
423 #Documentation
424 (r'((?<=fn\()|(?<=fnx\()|(?<=method\()|(?<=macro\()|(?<=lecro\()'
425 r'|(?<=syntax\()|(?<=dmacro\()|(?<=dlecro\()|(?<=dlecrox\()'
426 r'|(?<=dsyntax\())\s*"', String.Doc, 'documentation'),
427
428 #Text
429 (r'"', String, 'text'),
430 (r'#\[', String, 'squareText'),
431
432 #Mimic
433 (r'[a-zA-Z0-9_][a-zA-Z0-9!?_:]+(?=\s*=.*mimic\s)', Name.Entity),
434
435 #Assignment
436 (r'[a-zA-Z_][a-zA-Z0-9_!:?]*(?=[\s]*[+*/-]?=[^=].*($|\.))',
437 Name.Variable),
438
439 # keywords
440 (r'(break|cond|continue|do|ensure|for|for:dict|for:set|if|let|'
441 r'loop|p:for|p:for:dict|p:for:set|return|unless|until|while|'
442 r'with)(?![a-zA-Z0-9!:_?])', Keyword.Reserved),
443
444 # Origin
445 (r'(eval|mimic|print|println)(?![a-zA-Z0-9!:_?])', Keyword),
446
447 # Base
448 (r'(cell\?|cellNames|cellOwner\?|cellOwner|cells|cell|'
449 r'documentation|hash|identity|mimic|removeCell\!|undefineCell\!)'
450 r'(?![a-zA-Z0-9!:_?])', Keyword),
451
452 # Ground
453 (r'(stackTraceAsText)(?![a-zA-Z0-9!:_?])', Keyword),
454
455 #DefaultBehaviour Literals
456 (r'(dict|list|message|set)(?![a-zA-Z0-9!:_?])', Keyword.Reserved),
457
458 #DefaultBehaviour Case
459 (r'(case|case:and|case:else|case:nand|case:nor|case:not|case:or|'
460 r'case:otherwise|case:xor)(?![a-zA-Z0-9!:_?])', Keyword.Reserved),
461
462 #DefaultBehaviour Reflection
463 (r'(asText|become\!|derive|freeze\!|frozen\?|in\?|is\?|kind\?|'
464 r'mimic\!|mimics|mimics\?|prependMimic\!|removeAllMimics\!|'
465 r'removeMimic\!|same\?|send|thaw\!|uniqueHexId)'
466 r'(?![a-zA-Z0-9!:_?])', Keyword),
467
468 #DefaultBehaviour Aspects
469 (r'(after|around|before)(?![a-zA-Z0-9!:_?])', Keyword.Reserved),
470
471 # DefaultBehaviour
472 (r'(kind|cellDescriptionDict|cellSummary|genSym|inspect|notice)'
473 r'(?![a-zA-Z0-9!:_?])', Keyword),
474 (r'(use|destructuring)', Keyword.Reserved),
475
476 #DefaultBehavior BaseBehavior
477 (r'(cell\?|cellOwner\?|cellOwner|cellNames|cells|cell|'
478 r'documentation|identity|removeCell!|undefineCell)'
479 r'(?![a-zA-Z0-9!:_?])', Keyword),
480
481 #DefaultBehavior Internal
482 (r'(internal:compositeRegexp|internal:concatenateText|'
483 r'internal:createDecimal|internal:createNumber|'
484 r'internal:createRegexp|internal:createText)'
485 r'(?![a-zA-Z0-9!:_?])', Keyword.Reserved),
486
487 #DefaultBehaviour Conditions
488 (r'(availableRestarts|bind|error\!|findRestart|handle|'
489 r'invokeRestart|rescue|restart|signal\!|warn\!)'
490 r'(?![a-zA-Z0-9!:_?])', Keyword.Reserved),
491
492 # constants
493 (r'(nil|false|true)(?![a-zA-Z0-9!:_?])', Name.Constant),
494
495 # names
496 (r'(Arity|Base|Call|Condition|DateTime|Aspects|Pointcut|'
497 r'Assignment|BaseBehavior|Boolean|Case|AndCombiner|Else|'
498 r'NAndCombiner|NOrCombiner|NotCombiner|OrCombiner|XOrCombiner|'
499 r'Conditions|Definitions|FlowControl|Internal|Literals|'
500 r'Reflection|DefaultMacro|DefaultMethod|DefaultSyntax|Dict|'
501 r'FileSystem|Ground|Handler|Hook|IO|IokeGround|Struct|'
502 r'LexicalBlock|LexicalMacro|List|Message|Method|Mixins|'
503 r'NativeMethod|Number|Origin|Pair|Range|Reflector|Regexp Match|'
504 r'Regexp|Rescue|Restart|Runtime|Sequence|Set|Symbol|'
505 r'System|Text|Tuple)(?![a-zA-Z0-9!:_?])', Name.Builtin),
506
507 # functions
508 (r'(generateMatchMethod|aliasMethod|\u03bb|\u028E|fnx|fn|method|'
509 r'dmacro|dlecro|syntax|macro|dlecrox|lecrox|lecro|syntax)'
510 r'(?![a-zA-Z0-9!:_?])', Name.Function),
511
512 # Numbers
513 (r'-?0[xX][0-9a-fA-F]+', Number.Hex),
514 (r'-?(\d+\.?\d*|\d*\.\d+)([eE][+-]?[0-9]+)?', Number.Float),
515 (r'-?\d+', Number.Integer),
516
517 (r'#\(', Punctuation),
518
519 # Operators
520 (r'(&&>>|\|\|>>|\*\*>>|:::|::|\.\.\.|===|\*\*>|\*\*=|&&>|&&=|'
521 r'\|\|>|\|\|=|\->>|\+>>|!>>|<>>>|<>>|&>>|%>>|#>>|@>>|/>>|\*>>|'
522 r'\?>>|\|>>|\^>>|~>>|\$>>|=>>|<<=|>>=|<=>|<\->|=~|!~|=>|\+\+|'
523 r'\-\-|<=|>=|==|!=|&&|\.\.|\+=|\-=|\*=|\/=|%=|&=|\^=|\|=|<\-|'
524 r'\+>|!>|<>|&>|%>|#>|\@>|\/>|\*>|\?>|\|>|\^>|~>|\$>|<\->|\->|'
525 r'<<|>>|\*\*|\?\||\?&|\|\||>|<|\*|\/|%|\+|\-|&|\^|\||=|\$|!|~|'
526 r'\?|#|\u2260|\u2218|\u2208|\u2209)', Operator),
527 (r'(and|nand|or|xor|nor|return|import)(?![a-zA-Z0-9_!?])',
528 Operator),
529
530 # Punctuation
531 (r'(\`\`|\`|\'\'|\'|\.|\,|@@|@|\[|\]|\(|\)|{|})', Punctuation),
532
533 #kinds
534 (r'[A-Z][a-zA-Z0-9_!:?]*', Name.Class),
535
536 #default cellnames
537 (r'[a-z_][a-zA-Z0-9_!:?]*', Name)
538 ]
539 }
540
541
542 class ClojureLexer(RegexLexer):
543 """
544 Lexer for `Clojure <http://clojure.org/>`_ source code.
545
546 *New in Pygments 0.11.*
547 """
548 name = 'Clojure'
549 aliases = ['clojure', 'clj']
550 filenames = ['*.clj']
551 mimetypes = ['text/x-clojure', 'application/x-clojure']
552
553 special_forms = [
554 '.', 'def', 'do', 'fn', 'if', 'let', 'new', 'quote', 'var', 'loop'
555 ]
556
557 # It's safe to consider 'ns' a declaration thing because it defines a new
558 # namespace.
559 declarations = [
560 'def-', 'defn', 'defn-', 'defmacro', 'defmulti', 'defmethod',
561 'defstruct', 'defonce', 'declare', 'definline', 'definterface',
562 'defprotocol', 'defrecord', 'deftype', 'defproject', 'ns'
563 ]
564
565 builtins = [
566 '*', '+', '-', '->', '/', '<', '<=', '=', '==', '>', '>=', '..',
567 'accessor', 'agent', 'agent-errors', 'aget', 'alength', 'all-ns',
568 'alter', 'and', 'append-child', 'apply', 'array-map', 'aset',
569 'aset-boolean', 'aset-byte', 'aset-char', 'aset-double', 'aset-float',
570 'aset-int', 'aset-long', 'aset-short', 'assert', 'assoc', 'await',
571 'await-for', 'bean', 'binding', 'bit-and', 'bit-not', 'bit-or',
572 'bit-shift-left', 'bit-shift-right', 'bit-xor', 'boolean', 'branch?',
573 'butlast', 'byte', 'cast', 'char', 'children', 'class',
574 'clear-agent-errors', 'comment', 'commute', 'comp', 'comparator',
575 'complement', 'concat', 'conj', 'cons', 'constantly', 'cond', 'if-not',
576 'construct-proxy', 'contains?', 'count', 'create-ns', 'create-struct',
577 'cycle', 'dec', 'deref', 'difference', 'disj', 'dissoc', 'distinct',
578 'doall', 'doc', 'dorun', 'doseq', 'dosync', 'dotimes', 'doto',
579 'double', 'down', 'drop', 'drop-while', 'edit', 'end?', 'ensure',
580 'eval', 'every?', 'false?', 'ffirst', 'file-seq', 'filter', 'find',
581 'find-doc', 'find-ns', 'find-var', 'first', 'float', 'flush', 'for',
582 'fnseq', 'frest', 'gensym', 'get-proxy-class', 'get',
583 'hash-map', 'hash-set', 'identical?', 'identity', 'if-let', 'import',
584 'in-ns', 'inc', 'index', 'insert-child', 'insert-left', 'insert-right',
585 'inspect-table', 'inspect-tree', 'instance?', 'int', 'interleave',
586 'intersection', 'into', 'into-array', 'iterate', 'join', 'key', 'keys',
587 'keyword', 'keyword?', 'last', 'lazy-cat', 'lazy-cons', 'left',
588 'lefts', 'line-seq', 'list*', 'list', 'load', 'load-file',
589 'locking', 'long', 'loop', 'macroexpand', 'macroexpand-1',
590 'make-array', 'make-node', 'map', 'map-invert', 'map?', 'mapcat',
591 'max', 'max-key', 'memfn', 'merge', 'merge-with', 'meta', 'min',
592 'min-key', 'name', 'namespace', 'neg?', 'new', 'newline', 'next',
593 'nil?', 'node', 'not', 'not-any?', 'not-every?', 'not=', 'ns-imports',
594 'ns-interns', 'ns-map', 'ns-name', 'ns-publics', 'ns-refers',
595 'ns-resolve', 'ns-unmap', 'nth', 'nthrest', 'or', 'parse', 'partial',
596 'path', 'peek', 'pop', 'pos?', 'pr', 'pr-str', 'print', 'print-str',
597 'println', 'println-str', 'prn', 'prn-str', 'project', 'proxy',
598 'proxy-mappings', 'quot', 'rand', 'rand-int', 'range', 're-find',
599 're-groups', 're-matcher', 're-matches', 're-pattern', 're-seq',
600 'read', 'read-line', 'reduce', 'ref', 'ref-set', 'refer', 'rem',
601 'remove', 'remove-method', 'remove-ns', 'rename', 'rename-keys',
602 'repeat', 'replace', 'replicate', 'resolve', 'rest', 'resultset-seq',
603 'reverse', 'rfirst', 'right', 'rights', 'root', 'rrest', 'rseq',
604 'second', 'select', 'select-keys', 'send', 'send-off', 'seq',
605 'seq-zip', 'seq?', 'set', 'short', 'slurp', 'some', 'sort',
606 'sort-by', 'sorted-map', 'sorted-map-by', 'sorted-set',
607 'special-symbol?', 'split-at', 'split-with', 'str', 'string?',
608 'struct', 'struct-map', 'subs', 'subvec', 'symbol', 'symbol?',
609 'sync', 'take', 'take-nth', 'take-while', 'test', 'time', 'to-array',
610 'to-array-2d', 'tree-seq', 'true?', 'union', 'up', 'update-proxy',
611 'val', 'vals', 'var-get', 'var-set', 'var?', 'vector', 'vector-zip',
612 'vector?', 'when', 'when-first', 'when-let', 'when-not',
613 'with-local-vars', 'with-meta', 'with-open', 'with-out-str',
614 'xml-seq', 'xml-zip', 'zero?', 'zipmap', 'zipper']
615
616 # valid names for identifiers
617 # well, names can only not consist fully of numbers
618 # but this should be good enough for now
619
620 # TODO / should divide keywords/symbols into namespace/rest
621 # but that's hard, so just pretend / is part of the name
622 valid_name = r'(?!#)[\w!$%*+<=>?/.#-]+'
623
624 def _multi_escape(entries):
625 return '(%s)' % ('|'.join(re.escape(entry) + ' ' for entry in entries))
626
627 tokens = {
628 'root': [
629 # the comments - always starting with semicolon
630 # and going to the end of the line
631 (r';.*$', Comment.Single),
632
633 # whitespaces - usually not relevant
634 (r'[,\s]+', Text),
635
636 # numbers
637 (r'-?\d+\.\d+', Number.Float),
638 (r'-?\d+', Number.Integer),
639 (r'0x-?[abcdef\d]+', Number.Hex),
640
641 # strings, symbols and characters
642 (r'"(\\\\|\\"|[^"])*"', String),
643 (r"'" + valid_name, String.Symbol),
644 (r"\\(.|[a-z]+)", String.Char),
645
646 # keywords
647 (r':' + valid_name, String.Symbol),
648
649 # special operators
650 (r'~@|[`\'#^~&]', Operator),
651
652 # highlight the special forms
653 (_multi_escape(special_forms), Keyword),
654
655 # Technically, only the special forms are 'keywords'. The problem
656 # is that only treating them as keywords means that things like
657 # 'defn' and 'ns' need to be highlighted as builtins. This is ugly
658 # and weird for most styles. So, as a compromise we're going to
659 # highlight them as Keyword.Declarations.
660 (_multi_escape(declarations), Keyword.Declaration),
661
662 # highlight the builtins
663 (_multi_escape(builtins), Name.Builtin),
664
665 # the remaining functions
666 (r'(?<=\()' + valid_name, Name.Function),
667
668 # find the remaining variables
669 (valid_name, Name.Variable),
670
671 # Clojure accepts vector notation
672 (r'(\[|\])', Punctuation),
673
674 # Clojure accepts map notation
675 (r'(\{|\})', Punctuation),
676
677 # the famous parentheses!
678 (r'(\(|\))', Punctuation),
679 ],
680 }
681
682
683 class TeaLangLexer(RegexLexer):
684 """
685 For `Tea <http://teatrove.org/>`_ source code. Only used within a
686 TeaTemplateLexer.
687
688 *New in Pygments 1.5.*
689 """
690
691 flags = re.MULTILINE | re.DOTALL
692
693 #: optional Comment or Whitespace
694 _ws = r'(?:\s|//.*?\n|/[*].*?[*]/)+'
695
696 tokens = {
697 'root': [
698 # method names
699 (r'^(\s*(?:[a-zA-Z_][a-zA-Z0-9_\.\[\]]*\s+)+?)' # return arguments
700 r'([a-zA-Z_][a-zA-Z0-9_]*)' # method name
701 r'(\s*)(\()', # signature start
702 bygroups(using(this), Name.Function, Text, Operator)),
703 (r'[^\S\n]+', Text),
704 (r'//.*?\n', Comment.Single),
705 (r'/\*.*?\*/', Comment.Multiline),
706 (r'@[a-zA-Z_][a-zA-Z0-9_\.]*', Name.Decorator),
707 (r'(and|break|else|foreach|if|in|not|or|reverse)\b',
708 Keyword),
709 (r'(as|call|define)\b', Keyword.Declaration),
710 (r'(true|false|null)\b', Keyword.Constant),
711 (r'(template)(\s+)', bygroups(Keyword.Declaration, Text), 'template'),
712 (r'(import)(\s+)', bygroups(Keyword.Namespace, Text), 'import'),
713 (r'"(\\\\|\\"|[^"])*"', String),
714 (r'\'(\\\\|\\\'|[^\'])*\'', String),
715 (r'(\.)([a-zA-Z_][a-zA-Z0-9_]*)', bygroups(Operator, Name.Attribute)),
716 (r'[a-zA-Z_][a-zA-Z0-9_]*:', Name.Label),
717 (r'[a-zA-Z_\$][a-zA-Z0-9_]*', Name),
718 (r'(isa|[.]{3}|[.]{2}|[=#!<>+-/%&;,.\*\\\(\)\[\]\{\}])', Operator),
719 (r'[0-9][0-9]*\.[0-9]+([eE][0-9]+)?[fd]?', Number.Float),
720 (r'0x[0-9a-f]+', Number.Hex),
721 (r'[0-9]+L?', Number.Integer),
722 (r'\n', Text)
723 ],
724 'template': [
725 (r'[a-zA-Z_][a-zA-Z0-9_]*', Name.Class, '#pop')
726 ],
727 'import': [
728 (r'[a-zA-Z0-9_.]+\*?', Name.Namespace, '#pop')
729 ],
730 }
731
732
733 class KotlinLexer(RegexLexer):
734 """
735 For `Kotlin <http://confluence.jetbrains.net/display/Kotlin/>`_
736 source code.
737
738 Additional options accepted:
739
740 `unicodelevel`
741 Determines which Unicode characters this lexer allows for identifiers.
742 The possible values are:
743
744 * ``none`` -- only the ASCII letters and numbers are allowed. This
745 is the fastest selection.
746 * ``basic`` -- all Unicode characters from the specification except
747 category ``Lo`` are allowed.
748 * ``full`` -- all Unicode characters as specified in the C# specs
749 are allowed. Note that this means a considerable slowdown since the
750 ``Lo`` category has more than 40,000 characters in it!
751
752 The default value is ``basic``.
753
754 *New in Pygments 1.5.*
755 """
756
757 name = 'Kotlin'
758 aliases = ['kotlin']
759 filenames = ['*.kt']
760 mimetypes = ['text/x-kotlin'] # inferred
761
762 flags = re.MULTILINE | re.DOTALL | re.UNICODE
763
764 # for the range of allowed unicode characters in identifiers,
765 # see http://www.ecma-international.org/publications/files/ECMA-ST/Ecma-334.pdf
766
767 def _escape(st):
768 return st.replace('\\', r'\\').replace('-', r'\-').\
769 replace('[', r'\[').replace(']', r'\]')
770
771 levels = {
772 'none': '@?[_a-zA-Z][a-zA-Z0-9_]*',
773 'basic': ('@?[_' + uni.Lu + uni.Ll + uni.Lt + uni.Lm + uni.Nl + ']' +
774 '[' + uni.Lu + uni.Ll + uni.Lt + uni.Lm + uni.Nl +
775 uni.Nd + uni.Pc + uni.Cf + uni.Mn + uni.Mc + ']*'),
776 'full': ('@?(?:_|[^' +
777 _escape(uni.allexcept('Lu', 'Ll', 'Lt', 'Lm', 'Lo', 'Nl')) + '])'
778 + '[^' + _escape(uni.allexcept('Lu', 'Ll', 'Lt', 'Lm', 'Lo',
779 'Nl', 'Nd', 'Pc', 'Cf', 'Mn',
780 'Mc')) + ']*'),
781 }
782
783 tokens = {}
784 token_variants = True
785
786 for levelname, cs_ident in list(levels.items()):
787 tokens[levelname] = {
788 'root': [
789 # method names
790 (r'^([ \t]*(?:' + cs_ident + r'(?:\[\])?\s+)+?)' # return type
791 r'(' + cs_ident + ')' # method name
792 r'(\s*)(\()', # signature start
793 bygroups(using(this), Name.Function, Text, Punctuation)),
794 (r'^\s*\[.*?\]', Name.Attribute),
795 (r'[^\S\n]+', Text),
796 (r'\\\n', Text), # line continuation
797 (r'//.*?\n', Comment.Single),
798 (r'/[*](.|\n)*?[*]/', Comment.Multiline),
799 (r'\n', Text),
800 (r'[~!%^&*()+=|\[\]:;,.<>/?-]', Punctuation),
801 (r'[{}]', Punctuation),
802 (r'@"(""|[^"])*"', String),
803 (r'"(\\\\|\\"|[^"\n])*["\n]', String),
804 (r"'\\.'|'[^\\]'", String.Char),
805 (r"[0-9](\.[0-9]*)?([eE][+-][0-9]+)?"
806 r"[flFLdD]?|0[xX][0-9a-fA-F]+[Ll]?", Number),
807 (r'#[ \t]*(if|endif|else|elif|define|undef|'
808 r'line|error|warning|region|endregion|pragma)\b.*?\n',
809 Comment.Preproc),
810 (r'\b(extern)(\s+)(alias)\b', bygroups(Keyword, Text,
811 Keyword)),
812 (r'(abstract|as|break|catch|'
813 r'fun|continue|default|delegate|'
814 r'do|else|enum|extern|false|finally|'
815 r'fixed|for|goto|if|implicit|in|interface|'
816 r'internal|is|lock|null|'
817 r'out|override|private|protected|public|readonly|'
818 r'ref|return|sealed|sizeof|'
819 r'when|this|throw|true|try|typeof|'
820 r'unchecked|unsafe|virtual|void|while|'
821 r'get|set|new|partial|yield|val|var)\b', Keyword),
822 (r'(global)(::)', bygroups(Keyword, Punctuation)),
823 (r'(bool|byte|char|decimal|double|dynamic|float|int|long|'
824 r'short)\b\??', Keyword.Type),
825 (r'(class|struct)(\s+)', bygroups(Keyword, Text), 'class'),
826 (r'(package|using)(\s+)', bygroups(Keyword, Text), 'package'),
827 (cs_ident, Name),
828 ],
829 'class': [
830 (cs_ident, Name.Class, '#pop')
831 ],
832 'package': [
833 (r'(?=\()', Text, '#pop'), # using (resource)
834 ('(' + cs_ident + r'|\.)+', Name.Namespace, '#pop')
835 ]
836 }
837
838 def __init__(self, **options):
839 level = get_choice_opt(options, 'unicodelevel', list(self.tokens.keys()),
840 'basic')
841 if level not in self._all_tokens:
842 # compile the regexes now
843 self._tokens = self.__class__.process_tokendef(level)
844 else:
845 self._tokens = self._all_tokens[level]
846
847 RegexLexer.__init__(self, **options)

Mercurial Repository: eric

eric ide

mercurial