|
2
|
1 |
/*
|
|
5506
|
2 |
* Copyright (c) 1996, 2001, Oracle and/or its affiliates. All rights reserved.
|
|
2
|
3 |
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
|
|
|
4 |
*
|
|
|
5 |
* This code is free software; you can redistribute it and/or modify it
|
|
|
6 |
* under the terms of the GNU General Public License version 2 only, as
|
|
5506
|
7 |
* published by the Free Software Foundation. Oracle designates this
|
|
2
|
8 |
* particular file as subject to the "Classpath" exception as provided
|
|
5506
|
9 |
* by Oracle in the LICENSE file that accompanied this code.
|
|
2
|
10 |
*
|
|
|
11 |
* This code is distributed in the hope that it will be useful, but WITHOUT
|
|
|
12 |
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
|
|
|
13 |
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
|
|
|
14 |
* version 2 for more details (a copy is included in the LICENSE file that
|
|
|
15 |
* accompanied this code).
|
|
|
16 |
*
|
|
|
17 |
* You should have received a copy of the GNU General Public License version
|
|
|
18 |
* 2 along with this work; if not, write to the Free Software Foundation,
|
|
|
19 |
* Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
|
|
|
20 |
*
|
|
5506
|
21 |
* Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
|
|
|
22 |
* or visit www.oracle.com if you need additional information or have any
|
|
|
23 |
* questions.
|
|
2
|
24 |
*/
|
|
|
25 |
|
|
|
26 |
/*
|
|
|
27 |
* (C) Copyright Taligent, Inc. 1996,1997 - All Rights Reserved
|
|
|
28 |
* (C) Copyright IBM Corp. 1996, 1997 - All Rights Reserved
|
|
|
29 |
*
|
|
|
30 |
* The original version of this source code and documentation is copyrighted
|
|
|
31 |
* and owned by Taligent, Inc., a wholly-owned subsidiary of IBM. These
|
|
|
32 |
* materials are provided under terms of a License Agreement between Taligent
|
|
|
33 |
* and Sun. This technology is protected by multiple US and International
|
|
|
34 |
* patents. This notice and attribution to Taligent may not be removed.
|
|
|
35 |
* Taligent is a registered trademark of Taligent, Inc.
|
|
|
36 |
*
|
|
|
37 |
*/
|
|
|
38 |
|
|
|
39 |
package java.text;
|
|
|
40 |
/**
|
|
|
41 |
* CollationRules contains the default en_US collation rules as a base
|
|
|
42 |
* for building other collation tables.
|
|
|
43 |
* <p>Note that decompositions are done before these rules are used,
|
|
|
44 |
* so they do not have to contain accented characters, such as A-grave.
|
|
|
45 |
* @see RuleBasedCollator
|
|
|
46 |
* @see LocaleElements
|
|
|
47 |
* @author Helena Shih, Mark Davis
|
|
|
48 |
*/
|
|
|
49 |
final class CollationRules {
|
|
|
50 |
final static String DEFAULTRULES = new String(
|
|
|
51 |
"" // no FRENCH accent order by default, add in French Delta
|
|
|
52 |
// IGNORABLES (up to first < character)
|
|
|
53 |
// COMPLETELY IGNORE format characters
|
|
|
54 |
+ "='\u200B'=\u200C=\u200D=\u200E=\u200F"
|
|
|
55 |
// Control Characters
|
|
|
56 |
+ "=\u0000 =\u0001 =\u0002 =\u0003 =\u0004" //null, .. eot
|
|
|
57 |
+ "=\u0005 =\u0006 =\u0007 =\u0008 ='\u0009'" //enq, ...
|
|
|
58 |
+ "='\u000b' =\u000e" //vt,, so
|
|
|
59 |
+ "=\u000f ='\u0010' =\u0011 =\u0012 =\u0013" //si, dle, dc1, dc2, dc3
|
|
|
60 |
+ "=\u0014 =\u0015 =\u0016 =\u0017 =\u0018" //dc4, nak, syn, etb, can
|
|
|
61 |
+ "=\u0019 =\u001a =\u001b =\u001c =\u001d" //em, sub, esc, fs, gs
|
|
|
62 |
+ "=\u001e =\u001f =\u007f" //rs, us, del
|
|
|
63 |
//....then the C1 Latin 1 reserved control codes
|
|
|
64 |
+ "=\u0080 =\u0081 =\u0082 =\u0083 =\u0084 =\u0085"
|
|
|
65 |
+ "=\u0086 =\u0087 =\u0088 =\u0089 =\u008a =\u008b"
|
|
|
66 |
+ "=\u008c =\u008d =\u008e =\u008f =\u0090 =\u0091"
|
|
|
67 |
+ "=\u0092 =\u0093 =\u0094 =\u0095 =\u0096 =\u0097"
|
|
|
68 |
+ "=\u0098 =\u0099 =\u009a =\u009b =\u009c =\u009d"
|
|
|
69 |
+ "=\u009e =\u009f"
|
|
|
70 |
// IGNORE except for secondary, tertiary difference
|
|
|
71 |
// Spaces
|
|
|
72 |
+ ";'\u0020';'\u00A0'" // spaces
|
|
|
73 |
+ ";'\u2000';'\u2001';'\u2002';'\u2003';'\u2004'" // spaces
|
|
|
74 |
+ ";'\u2005';'\u2006';'\u2007';'\u2008';'\u2009'" // spaces
|
|
|
75 |
+ ";'\u200A';'\u3000';'\uFEFF'" // spaces
|
|
|
76 |
+ ";'\r' ;'\t' ;'\n';'\f';'\u000b'" // whitespace
|
|
|
77 |
|
|
|
78 |
// Non-spacing accents
|
|
|
79 |
|
|
|
80 |
+ ";\u0301" // non-spacing acute accent
|
|
|
81 |
+ ";\u0300" // non-spacing grave accent
|
|
|
82 |
+ ";\u0306" // non-spacing breve accent
|
|
|
83 |
+ ";\u0302" // non-spacing circumflex accent
|
|
|
84 |
+ ";\u030c" // non-spacing caron/hacek accent
|
|
|
85 |
+ ";\u030a" // non-spacing ring above accent
|
|
|
86 |
+ ";\u030d" // non-spacing vertical line above
|
|
|
87 |
+ ";\u0308" // non-spacing diaeresis accent
|
|
|
88 |
+ ";\u030b" // non-spacing double acute accent
|
|
|
89 |
+ ";\u0303" // non-spacing tilde accent
|
|
|
90 |
+ ";\u0307" // non-spacing dot above/overdot accent
|
|
|
91 |
+ ";\u0304" // non-spacing macron accent
|
|
|
92 |
+ ";\u0337" // non-spacing short slash overlay (overstruck diacritic)
|
|
|
93 |
+ ";\u0327" // non-spacing cedilla accent
|
|
|
94 |
+ ";\u0328" // non-spacing ogonek accent
|
|
|
95 |
+ ";\u0323" // non-spacing dot-below/underdot accent
|
|
|
96 |
+ ";\u0332" // non-spacing underscore/underline accent
|
|
|
97 |
// with the rest of the general diacritical marks in binary order
|
|
|
98 |
+ ";\u0305" // non-spacing overscore/overline
|
|
|
99 |
+ ";\u0309" // non-spacing hook above
|
|
|
100 |
+ ";\u030e" // non-spacing double vertical line above
|
|
|
101 |
+ ";\u030f" // non-spacing double grave
|
|
|
102 |
+ ";\u0310" // non-spacing chandrabindu
|
|
|
103 |
+ ";\u0311" // non-spacing inverted breve
|
|
|
104 |
+ ";\u0312" // non-spacing turned comma above/cedilla above
|
|
|
105 |
+ ";\u0313" // non-spacing comma above
|
|
|
106 |
+ ";\u0314" // non-spacing reversed comma above
|
|
|
107 |
+ ";\u0315" // non-spacing comma above right
|
|
|
108 |
+ ";\u0316" // non-spacing grave below
|
|
|
109 |
+ ";\u0317" // non-spacing acute below
|
|
|
110 |
+ ";\u0318" // non-spacing left tack below
|
|
|
111 |
+ ";\u0319" // non-spacing tack below
|
|
|
112 |
+ ";\u031a" // non-spacing left angle above
|
|
|
113 |
+ ";\u031b" // non-spacing horn
|
|
|
114 |
+ ";\u031c" // non-spacing left half ring below
|
|
|
115 |
+ ";\u031d" // non-spacing up tack below
|
|
|
116 |
+ ";\u031e" // non-spacing down tack below
|
|
|
117 |
+ ";\u031f" // non-spacing plus sign below
|
|
|
118 |
+ ";\u0320" // non-spacing minus sign below
|
|
|
119 |
+ ";\u0321" // non-spacing palatalized hook below
|
|
|
120 |
+ ";\u0322" // non-spacing retroflex hook below
|
|
|
121 |
+ ";\u0324" // non-spacing double dot below
|
|
|
122 |
+ ";\u0325" // non-spacing ring below
|
|
|
123 |
+ ";\u0326" // non-spacing comma below
|
|
|
124 |
+ ";\u0329" // non-spacing vertical line below
|
|
|
125 |
+ ";\u032a" // non-spacing bridge below
|
|
|
126 |
+ ";\u032b" // non-spacing inverted double arch below
|
|
|
127 |
+ ";\u032c" // non-spacing hacek below
|
|
|
128 |
+ ";\u032d" // non-spacing circumflex below
|
|
|
129 |
+ ";\u032e" // non-spacing breve below
|
|
|
130 |
+ ";\u032f" // non-spacing inverted breve below
|
|
|
131 |
+ ";\u0330" // non-spacing tilde below
|
|
|
132 |
+ ";\u0331" // non-spacing macron below
|
|
|
133 |
+ ";\u0333" // non-spacing double underscore
|
|
|
134 |
+ ";\u0334" // non-spacing tilde overlay
|
|
|
135 |
+ ";\u0335" // non-spacing short bar overlay
|
|
|
136 |
+ ";\u0336" // non-spacing long bar overlay
|
|
|
137 |
+ ";\u0338" // non-spacing long slash overlay
|
|
|
138 |
+ ";\u0339" // non-spacing right half ring below
|
|
|
139 |
+ ";\u033a" // non-spacing inverted bridge below
|
|
|
140 |
+ ";\u033b" // non-spacing square below
|
|
|
141 |
+ ";\u033c" // non-spacing seagull below
|
|
|
142 |
+ ";\u033d" // non-spacing x above
|
|
|
143 |
+ ";\u033e" // non-spacing vertical tilde
|
|
|
144 |
+ ";\u033f" // non-spacing double overscore
|
|
|
145 |
//+ ";\u0340" // non-spacing grave tone mark == \u0300
|
|
|
146 |
//+ ";\u0341" // non-spacing acute tone mark == \u0301
|
|
|
147 |
+ ";\u0342;"
|
|
|
148 |
//+ "\u0343;" // == \u0313
|
|
|
149 |
+ "\u0344;\u0345;\u0360;\u0361" // newer
|
|
|
150 |
+ ";\u0483;\u0484;\u0485;\u0486" // Cyrillic accents
|
|
|
151 |
|
|
|
152 |
+ ";\u20D0;\u20D1;\u20D2" // symbol accents
|
|
|
153 |
+ ";\u20D3;\u20D4;\u20D5" // symbol accents
|
|
|
154 |
+ ";\u20D6;\u20D7;\u20D8" // symbol accents
|
|
|
155 |
+ ";\u20D9;\u20DA;\u20DB" // symbol accents
|
|
|
156 |
+ ";\u20DC;\u20DD;\u20DE" // symbol accents
|
|
|
157 |
+ ";\u20DF;\u20E0;\u20E1" // symbol accents
|
|
|
158 |
|
|
|
159 |
+ ",'\u002D';\u00AD" // dashes
|
|
|
160 |
+ ";\u2010;\u2011;\u2012" // dashes
|
|
|
161 |
+ ";\u2013;\u2014;\u2015" // dashes
|
|
|
162 |
+ ";\u2212" // dashes
|
|
|
163 |
|
|
|
164 |
// other punctuation
|
|
|
165 |
|
|
|
166 |
+ "<'\u005f'" // underline/underscore (spacing)
|
|
|
167 |
+ "<\u00af" // overline or macron (spacing)
|
|
|
168 |
+ "<'\u002c'" // comma (spacing)
|
|
|
169 |
+ "<'\u003b'" // semicolon
|
|
|
170 |
+ "<'\u003a'" // colon
|
|
|
171 |
+ "<'\u0021'" // exclamation point
|
|
|
172 |
+ "<\u00a1" // inverted exclamation point
|
|
|
173 |
+ "<'\u003f'" // question mark
|
|
|
174 |
+ "<\u00bf" // inverted question mark
|
|
|
175 |
+ "<'\u002f'" // slash
|
|
|
176 |
+ "<'\u002e'" // period/full stop
|
|
|
177 |
+ "<\u00b4" // acute accent (spacing)
|
|
|
178 |
+ "<'\u0060'" // grave accent (spacing)
|
|
|
179 |
+ "<'\u005e'" // circumflex accent (spacing)
|
|
|
180 |
+ "<\u00a8" // diaresis/umlaut accent (spacing)
|
|
|
181 |
+ "<'\u007e'" // tilde accent (spacing)
|
|
|
182 |
+ "<\u00b7" // middle dot (spacing)
|
|
|
183 |
+ "<\u00b8" // cedilla accent (spacing)
|
|
|
184 |
+ "<'\u0027'" // apostrophe
|
|
|
185 |
+ "<'\"'" // quotation marks
|
|
|
186 |
+ "<\u00ab" // left angle quotes
|
|
|
187 |
+ "<\u00bb" // right angle quotes
|
|
|
188 |
+ "<'\u0028'" // left parenthesis
|
|
|
189 |
+ "<'\u0029'" // right parenthesis
|
|
|
190 |
+ "<'\u005b'" // left bracket
|
|
|
191 |
+ "<'\u005d'" // right bracket
|
|
|
192 |
+ "<'\u007b'" // left brace
|
|
|
193 |
+ "<'\u007d'" // right brace
|
|
|
194 |
+ "<\u00a7" // section symbol
|
|
|
195 |
+ "<\u00b6" // paragraph symbol
|
|
|
196 |
+ "<\u00a9" // copyright symbol
|
|
|
197 |
+ "<\u00ae" // registered trademark symbol
|
|
|
198 |
+ "<'\u0040'" // at sign
|
|
|
199 |
+ "<\u00a4" // international currency symbol
|
|
|
200 |
+ "<\u0e3f" // baht sign
|
|
|
201 |
+ "<\u00a2" // cent sign
|
|
|
202 |
+ "<\u20a1" // colon sign
|
|
|
203 |
+ "<\u20a2" // cruzeiro sign
|
|
|
204 |
+ "<'\u0024'" // dollar sign
|
|
|
205 |
+ "<\u20ab" // dong sign
|
|
|
206 |
+ "<\u20ac" // euro sign
|
|
|
207 |
+ "<\u20a3" // franc sign
|
|
|
208 |
+ "<\u20a4" // lira sign
|
|
|
209 |
+ "<\u20a5" // mill sign
|
|
|
210 |
+ "<\u20a6" // naira sign
|
|
|
211 |
+ "<\u20a7" // peseta sign
|
|
|
212 |
+ "<\u00a3" // pound-sterling sign
|
|
|
213 |
+ "<\u20a8" // rupee sign
|
|
|
214 |
+ "<\u20aa" // new shekel sign
|
|
|
215 |
+ "<\u20a9" // won sign
|
|
|
216 |
+ "<\u00a5" // yen sign
|
|
|
217 |
+ "<'\u002a'" // asterisk
|
|
|
218 |
+ "<'\\'" // backslash
|
|
|
219 |
+ "<'\u0026'" // ampersand
|
|
|
220 |
+ "<'\u0023'" // number sign
|
|
|
221 |
+ "<'\u0025'" // percent sign
|
|
|
222 |
+ "<'\u002b'" // plus sign
|
|
|
223 |
+ "<\u00b1" // plus-or-minus sign
|
|
|
224 |
+ "<\u00f7" // divide sign
|
|
|
225 |
+ "<\u00d7" // multiply sign
|
|
|
226 |
+ "<'\u003c'" // less-than sign
|
|
|
227 |
+ "<'\u003d'" // equal sign
|
|
|
228 |
+ "<'\u003e'" // greater-than sign
|
|
|
229 |
+ "<\u00ac" // end of line symbol/logical NOT symbol
|
|
|
230 |
+ "<'\u007c'" // vertical line/logical OR symbol
|
|
|
231 |
+ "<\u00a6" // broken vertical line
|
|
|
232 |
+ "<\u00b0" // degree symbol
|
|
|
233 |
+ "<\u00b5" // micro symbol
|
|
|
234 |
|
|
|
235 |
// NUMERICS
|
|
|
236 |
|
|
|
237 |
+ "<0<1<2<3<4<5<6<7<8<9"
|
|
|
238 |
+ "<\u00bc<\u00bd<\u00be" // 1/4,1/2,3/4 fractions
|
|
|
239 |
|
|
|
240 |
// NON-IGNORABLES
|
|
|
241 |
+ "<a,A"
|
|
|
242 |
+ "<b,B"
|
|
|
243 |
+ "<c,C"
|
|
|
244 |
+ "<d,D"
|
|
|
245 |
+ "<\u00F0,\u00D0" // eth
|
|
|
246 |
+ "<e,E"
|
|
|
247 |
+ "<f,F"
|
|
|
248 |
+ "<g,G"
|
|
|
249 |
+ "<h,H"
|
|
|
250 |
+ "<i,I"
|
|
|
251 |
+ "<j,J"
|
|
|
252 |
+ "<k,K"
|
|
|
253 |
+ "<l,L"
|
|
|
254 |
+ "<m,M"
|
|
|
255 |
+ "<n,N"
|
|
|
256 |
+ "<o,O"
|
|
|
257 |
+ "<p,P"
|
|
|
258 |
+ "<q,Q"
|
|
|
259 |
+ "<r,R"
|
|
|
260 |
+ "<s, S & SS,\u00DF" // s-zet
|
|
|
261 |
+ "<t,T"
|
|
|
262 |
+ "& TH, \u00DE &TH, \u00FE " // thorn
|
|
|
263 |
+ "<u,U"
|
|
|
264 |
+ "<v,V"
|
|
|
265 |
+ "<w,W"
|
|
|
266 |
+ "<x,X"
|
|
|
267 |
+ "<y,Y"
|
|
|
268 |
+ "<z,Z"
|
|
|
269 |
+ "&AE,\u00C6" // ae & AE ligature
|
|
|
270 |
+ "&AE,\u00E6"
|
|
|
271 |
+ "&OE,\u0152" // oe & OE ligature
|
|
|
272 |
+ "&OE,\u0153"
|
|
|
273 |
);
|
|
|
274 |
}
|