• file
• revisions
• annotate
• diff
• comparison
• raw

src/share/jaxws_classes/com/sun/xml/internal/dtdparser/XmlChars.java@9c07ef4934dd (annotated)

src/share/jaxws_classes/com/sun/xml/internal/dtdparser/XmlChars.java

Thu, 31 Aug 2017 15:18:52 +0800

author

aoqi

date

Thu, 31 Aug 2017 15:18:52 +0800

changeset 637

9c07ef4934dd

parent 397

b99d7e355d4b

parent 0

373ffda63c9a

permissions

-rw-r--r--

merge

aoqi@0	1	/*
aoqi@0	2	* Copyright (c) 2009, Oracle and/or its affiliates. All rights reserved.
aoqi@0	3	* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
aoqi@0	4	*
aoqi@0	5	* This code is free software; you can redistribute it and/or modify it
aoqi@0	6	* under the terms of the GNU General Public License version 2 only, as
aoqi@0	7	* published by the Free Software Foundation. Oracle designates this
aoqi@0	8	* particular file as subject to the "Classpath" exception as provided
aoqi@0	9	* by Oracle in the LICENSE file that accompanied this code.
aoqi@0	10	*
aoqi@0	11	* This code is distributed in the hope that it will be useful, but WITHOUT
aoqi@0	12	* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
aoqi@0	13	* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
aoqi@0	14	* version 2 for more details (a copy is included in the LICENSE file that
aoqi@0	15	* accompanied this code).
aoqi@0	16	*
aoqi@0	17	* You should have received a copy of the GNU General Public License version
aoqi@0	18	* 2 along with this work; if not, write to the Free Software Foundation,
aoqi@0	19	* Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
aoqi@0	20	*
aoqi@0	21	* Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
aoqi@0	22	* or visit www.oracle.com if you need additional information or have any
aoqi@0	23	* questions.
aoqi@0	24	*/
aoqi@0	25
aoqi@0	26	package com.sun.xml.internal.dtdparser;
aoqi@0	27
aoqi@0	28
aoqi@0	29	/**
aoqi@0	30	* Methods in this class are used to determine whether characters may
aoqi@0	31	* appear in certain roles in XML documents. Such methods are used
aoqi@0	32	* both to parse and to create such documents.
aoqi@0	33	*
aoqi@0	34	* @author David Brownell
aoqi@0	35	* @version 1.1, 00/08/05
aoqi@0	36	*/
aoqi@0	37	public class XmlChars {
aoqi@0	38	// can't construct instances
aoqi@0	39	private XmlChars() {
aoqi@0	40	}
aoqi@0	41
aoqi@0	42	/**
aoqi@0	43	* Returns true if the argument, a UCS-4 character code, is valid in
aoqi@0	44	* XML documents. Unicode characters fit into the low sixteen
aoqi@0	45	* bits of a UCS-4 character, and pairs of Unicode <em>surrogate
aoqi@0	46	* characters</em> can be combined to encode UCS-4 characters in
aoqi@0	47	* documents containing only Unicode. (The <code>char</code> datatype
aoqi@0	48	* in the Java Programming Language represents Unicode characters,
aoqi@0	49	* including unpaired surrogates.)
aoqi@0	50	* <p/>
aoqi@0	51	* <P> In XML, UCS-4 characters can also be encoded by the use of
aoqi@0	52	* <em>character references</em> such as <b>&amp;#x12345678;</b>, which
aoqi@0	53	* happens to refer to a character that is disallowed in XML documents.
aoqi@0	54	* UCS-4 characters allowed in XML documents can be expressed with
aoqi@0	55	* one or two Unicode characters.
aoqi@0	56	*
aoqi@0	57	* @param ucs4char The 32-bit UCS-4 character being tested.
aoqi@0	58	*/
aoqi@0	59	static public boolean isChar(int ucs4char) {
aoqi@0	60	// [2] Char ::= #x0009 | #x000A | #x000D
aoqi@0	61	// | [#x0020-#xD7FF]
aoqi@0	62	// ... surrogates excluded!
aoqi@0	63	// | [#xE000-#xFFFD]
aoqi@0	64	// | [#x10000-#x10ffff]
aoqi@0	65	return ((ucs4char >= 0x0020 && ucs4char <= 0xD7FF)
aoqi@0	66	|| ucs4char == 0x000A || ucs4char == 0x0009
aoqi@0	67	|| ucs4char == 0x000D
aoqi@0	68	|| (ucs4char >= 0xE000 && ucs4char <= 0xFFFD)
aoqi@0	69	|| (ucs4char >= 0x10000 && ucs4char <= 0x10ffff));
aoqi@0	70	}
aoqi@0	71
aoqi@0	72	/**
aoqi@0	73	* Returns true if the character is allowed to be a non-initial
aoqi@0	74	* character in names according to the XML recommendation.
aoqi@0	75	*
aoqi@0	76	* @see #isNCNameChar(char)
aoqi@0	77	* @see #isLetter(char)
aoqi@0	78	*/
aoqi@0	79	public static boolean isNameChar(char c) {
aoqi@0	80	// [4] NameChar ::= Letter | Digit | '.' | '_' | ':'
aoqi@0	81	// | CombiningChar | Extender
aoqi@0	82
aoqi@0	83	if (isLetter2(c))
aoqi@0	84	return true;
aoqi@0	85	else if (c == '>')
aoqi@0	86	return false;
aoqi@0	87	else if (c == '.' || c == '-' || c == '_' || c == ':'
aoqi@0	88	|| isExtender(c))
aoqi@0	89	return true;
aoqi@0	90	else
aoqi@0	91	return false;
aoqi@0	92	}
aoqi@0	93
aoqi@0	94	/**
aoqi@0	95	* Returns true if the character is allowed to be a non-initial
aoqi@0	96	* character in unscoped names according to the rules of the XML
aoqi@0	97	* Namespaces proposed recommendation. Except for precluding
aoqi@0	98	* the colon (used to separate names from their scopes) these
aoqi@0	99	* characters are just as allowed by the XML recommendation.
aoqi@0	100	*
aoqi@0	101	* @see #isNameChar(char)
aoqi@0	102	* @see #isLetter(char)
aoqi@0	103	*/
aoqi@0	104	public static boolean isNCNameChar(char c) {
aoqi@0	105	// [NC 5] NCNameChar ::= Letter | Digit | '.' | '_'
aoqi@0	106	// | CombiningChar | Extender
aoqi@0	107	return c != ':' && isNameChar(c);
aoqi@0	108	}
aoqi@0	109
aoqi@0	110	/**
aoqi@0	111	* Returns true if the character is allowed where XML supports
aoqi@0	112	* whitespace characters, false otherwise.
aoqi@0	113	*/
aoqi@0	114	public static boolean isSpace(char c) {
aoqi@0	115	return c == ' ' || c == '\t' || c == '\n' || c == '\r';
aoqi@0	116	}
aoqi@0	117
aoqi@0	118
aoqi@0	119	/*
aoqi@0	120	* NOTE: java.lang.Character.getType() values are:
aoqi@0	121	*
aoqi@0	122	* UNASSIGNED = 0,
aoqi@0	123	*
aoqi@0	124	* UPPERCASE_LETTER = 1, // Lu
aoqi@0	125	* LOWERCASE_LETTER = 2, // Ll
aoqi@0	126	* TITLECASE_LETTER = 3, // Lt
aoqi@0	127	* MODIFIER_LETTER = 4, // Lm
aoqi@0	128	* OTHER_LETTER = 5, // Lo
aoqi@0	129	* NON_SPACING_MARK = 6, // Mn
aoqi@0	130	* ENCLOSING_MARK = 7, // Me
aoqi@0	131	* COMBINING_SPACING_MARK = 8, // Mc
aoqi@0	132	* DECIMAL_DIGIT_NUMBER = 9, // Nd
aoqi@0	133	* LETTER_NUMBER = 10, // Nl
aoqi@0	134	* OTHER_NUMBER = 11, // No
aoqi@0	135	* SPACE_SEPARATOR = 12, // Zs
aoqi@0	136	* LINE_SEPARATOR = 13, // Zl
aoqi@0	137	* PARAGRAPH_SEPARATOR = 14, // Zp
aoqi@0	138	* CONTROL = 15, // Cc
aoqi@0	139	* FORMAT = 16, // Cf
aoqi@0	140	* // 17 reserved for proposed Ci category
aoqi@0	141	* PRIVATE_USE = 18, // Co
aoqi@0	142	* SURROGATE = 19, // Cs
aoqi@0	143	* DASH_PUNCTUATION = 20, // Pd
aoqi@0	144	* START_PUNCTUATION = 21, // Ps
aoqi@0	145	* END_PUNCTUATION = 22, // Pe
aoqi@0	146	* CONNECTOR_PUNCTUATION = 23, // Pc
aoqi@0	147	* OTHER_PUNCTUATION = 24, // Po
aoqi@0	148	* MATH_SYMBOL = 25, // Sm
aoqi@0	149	* CURRENCY_SYMBOL = 26, // Sc
aoqi@0	150	* MODIFIER_SYMBOL = 27, // Sk
aoqi@0	151	* OTHER_SYMBOL = 28; // So
aoqi@0	152	*/
aoqi@0	153
aoqi@0	154	/**
aoqi@0	155	* Returns true if the character is an XML "letter". XML Names must
aoqi@0	156	* start with Letters or a few other characters, but other characters
aoqi@0	157	* in names must only satisfy the <em>isNameChar</em> predicate.
aoqi@0	158	*
aoqi@0	159	* @see #isNameChar(char)
aoqi@0	160	* @see #isNCNameChar(char)
aoqi@0	161	*/
aoqi@0	162	public static boolean isLetter(char c) {
aoqi@0	163	// [84] Letter ::= BaseChar | Ideographic
aoqi@0	164	// [85] BaseChar ::= ... too much to repeat
aoqi@0	165	// [86] Ideographic ::= ... too much to repeat
aoqi@0	166
aoqi@0	167	//
aoqi@0	168	// Optimize the typical case.
aoqi@0	169	//
aoqi@0	170	if (c >= 'a' && c <= 'z')
aoqi@0	171	return true;
aoqi@0	172	if (c == '/')
aoqi@0	173	return false;
aoqi@0	174	if (c >= 'A' && c <= 'Z')
aoqi@0	175	return true;
aoqi@0	176
aoqi@0	177	//
aoqi@0	178	// Since the tables are too ridiculous to use in code,
aoqi@0	179	// we're using the footnotes here to drive this test.
aoqi@0	180	//
aoqi@0	181	switch (Character.getType(c)) {
aoqi@0	182	// app. B footnote says these are 'name start'
aoqi@0	183	// chars' ...
aoqi@0	184	case Character.LOWERCASE_LETTER: // Ll
aoqi@0	185	case Character.UPPERCASE_LETTER: // Lu
aoqi@0	186	case Character.OTHER_LETTER: // Lo
aoqi@0	187	case Character.TITLECASE_LETTER: // Lt
aoqi@0	188	case Character.LETTER_NUMBER: // Nl
aoqi@0	189
aoqi@0	190	// OK, here we just have some exceptions to check...
aoqi@0	191	return !isCompatibilityChar(c)
aoqi@0	192	// per "5.14 of Unicode", rule out some combiners
aoqi@0	193	&& !(c >= 0x20dd && c <= 0x20e0);
aoqi@0	194
aoqi@0	195	default:
aoqi@0	196	// check for some exceptions: these are "alphabetic"
aoqi@0	197	return ((c >= 0x02bb && c <= 0x02c1)
aoqi@0	198	|| c == 0x0559 || c == 0x06e5 || c == 0x06e6);
aoqi@0	199	}
aoqi@0	200	}
aoqi@0	201
aoqi@0	202	//
aoqi@0	203	// XML 1.0 discourages "compatibility" characters in names; these
aoqi@0	204	// were defined to permit passing through some information stored in
aoqi@0	205	// older non-Unicode character sets. These always have alternative
aoqi@0	206	// representations in Unicode, e.g. using combining chars.
aoqi@0	207	//
aoqi@0	208	private static boolean isCompatibilityChar(char c) {
aoqi@0	209	// the numerous comparisions here seem unavoidable,
aoqi@0	210	// but the switch can reduce the number which must
aoqi@0	211	// actually be executed.
aoqi@0	212
aoqi@0	213	switch ((c >> 8) & 0x0ff) {
aoqi@0	214	case 0x00:
aoqi@0	215	// ISO Latin/1 has a few compatibility characters
aoqi@0	216	return c == 0x00aa || c == 0x00b5 || c == 0x00ba;
aoqi@0	217
aoqi@0	218	case 0x01:
aoqi@0	219	// as do Latin Extended A and (parts of) B
aoqi@0	220	return (c >= 0x0132 && c <= 0x0133)
aoqi@0	221	|| (c >= 0x013f && c <= 0x0140)
aoqi@0	222	|| c == 0x0149
aoqi@0	223	|| c == 0x017f
aoqi@0	224	|| (c >= 0x01c4 && c <= 0x01cc)
aoqi@0	225	|| (c >= 0x01f1 && c <= 0x01f3);
aoqi@0	226
aoqi@0	227	case 0x02:
aoqi@0	228	// some spacing modifiers
aoqi@0	229	return (c >= 0x02b0 && c <= 0x02b8)
aoqi@0	230	|| (c >= 0x02e0 && c <= 0x02e4);
aoqi@0	231
aoqi@0	232	case 0x03:
aoqi@0	233	return c == 0x037a; // Greek
aoqi@0	234
aoqi@0	235	case 0x05:
aoqi@0	236	return c == 0x0587; // Armenian
aoqi@0	237
aoqi@0	238	case 0x0e:
aoqi@0	239	return c >= 0x0edc && c <= 0x0edd; // Laotian
aoqi@0	240
aoqi@0	241	case 0x11:
aoqi@0	242	// big chunks of Hangul Jamo are all "compatibility"
aoqi@0	243	return c == 0x1101
aoqi@0	244	|| c == 0x1104
aoqi@0	245	|| c == 0x1108
aoqi@0	246	|| c == 0x110a
aoqi@0	247	|| c == 0x110d
aoqi@0	248	|| (c >= 0x1113 && c <= 0x113b)
aoqi@0	249	|| c == 0x113d
aoqi@0	250	|| c == 0x113f
aoqi@0	251	|| (c >= 0x1141 && c <= 0x114b)
aoqi@0	252	|| c == 0x114d
aoqi@0	253	|| c == 0x114f
aoqi@0	254	|| (c >= 0x1151 && c <= 0x1153)
aoqi@0	255	|| (c >= 0x1156 && c <= 0x1158)
aoqi@0	256	|| c == 0x1162
aoqi@0	257	|| c == 0x1164
aoqi@0	258	|| c == 0x1166
aoqi@0	259	|| c == 0x1168
aoqi@0	260	|| (c >= 0x116a && c <= 0x116c)
aoqi@0	261	|| (c >= 0x116f && c <= 0x1171)
aoqi@0	262	|| c == 0x1174
aoqi@0	263	|| (c >= 0x1176 && c <= 0x119d)
aoqi@0	264	|| (c >= 0x119f && c <= 0x11a2)
aoqi@0	265	|| (c >= 0x11a9 && c <= 0x11aa)
aoqi@0	266	|| (c >= 0x11ac && c <= 0x11ad)
aoqi@0	267	|| (c >= 0x11b0 && c <= 0x11b6)
aoqi@0	268	|| c == 0x11b9
aoqi@0	269	|| c == 0x11bb
aoqi@0	270	|| (c >= 0x11c3 && c <= 0x11ea)
aoqi@0	271	|| (c >= 0x11ec && c <= 0x11ef)
aoqi@0	272	|| (c >= 0x11f1 && c <= 0x11f8)
aoqi@0	273	;
aoqi@0	274
aoqi@0	275	case 0x20:
aoqi@0	276	return c == 0x207f; // superscript
aoqi@0	277
aoqi@0	278	case 0x21:
aoqi@0	279	return
aoqi@0	280	// various letterlike symbols
aoqi@0	281	c == 0x2102
aoqi@0	282	|| c == 0x2107
aoqi@0	283	|| (c >= 0x210a && c <= 0x2113)
aoqi@0	284	|| c == 0x2115
aoqi@0	285	|| (c >= 0x2118 && c <= 0x211d)
aoqi@0	286	|| c == 0x2124
aoqi@0	287	|| c == 0x2128
aoqi@0	288	|| (c >= 0x212c && c <= 0x212d)
aoqi@0	289	|| (c >= 0x212f && c <= 0x2138)
aoqi@0	290
aoqi@0	291	// most Roman numerals (less 1K, 5K, 10K)
aoqi@0	292	|| (c >= 0x2160 && c <= 0x217f)
aoqi@0	293	;
aoqi@0	294
aoqi@0	295	case 0x30:
aoqi@0	296	// some Hiragana
aoqi@0	297	return c >= 0x309b && c <= 0x309c;
aoqi@0	298
aoqi@0	299	case 0x31:
aoqi@0	300	// all Hangul Compatibility Jamo
aoqi@0	301	return c >= 0x3131 && c <= 0x318e;
aoqi@0	302
aoqi@0	303	case 0xf9:
aoqi@0	304	case 0xfa:
aoqi@0	305	case 0xfb:
aoqi@0	306	case 0xfc:
aoqi@0	307	case 0xfd:
aoqi@0	308	case 0xfe:
aoqi@0	309	case 0xff:
aoqi@0	310	// the whole "compatibility" area is for that purpose!
aoqi@0	311	return true;
aoqi@0	312
aoqi@0	313	default:
aoqi@0	314	// most of Unicode isn't flagged as being for compatibility
aoqi@0	315	return false;
aoqi@0	316	}
aoqi@0	317	}
aoqi@0	318
aoqi@0	319	// guts of isNameChar/isNCNameChar
aoqi@0	320	private static boolean isLetter2(char c) {
aoqi@0	321	// [84] Letter ::= BaseChar | Ideographic
aoqi@0	322	// [85] BaseChar ::= ... too much to repeat
aoqi@0	323	// [86] Ideographic ::= ... too much to repeat
aoqi@0	324	// [87] CombiningChar ::= ... too much to repeat
aoqi@0	325
aoqi@0	326	//
aoqi@0	327	// Optimize the typical case.
aoqi@0	328	//
aoqi@0	329	if (c >= 'a' && c <= 'z')
aoqi@0	330	return true;
aoqi@0	331	if (c == '>')
aoqi@0	332	return false;
aoqi@0	333	if (c >= 'A' && c <= 'Z')
aoqi@0	334	return true;
aoqi@0	335
aoqi@0	336	//
aoqi@0	337	// Since the tables are too ridiculous to use in code,
aoqi@0	338	// we're using the footnotes here to drive this test.
aoqi@0	339	//
aoqi@0	340	switch (Character.getType(c)) {
aoqi@0	341	// app. B footnote says these are 'name start'
aoqi@0	342	// chars' ...
aoqi@0	343	case Character.LOWERCASE_LETTER: // Ll
aoqi@0	344	case Character.UPPERCASE_LETTER: // Lu
aoqi@0	345	case Character.OTHER_LETTER: // Lo
aoqi@0	346	case Character.TITLECASE_LETTER: // Lt
aoqi@0	347	case Character.LETTER_NUMBER: // Nl
aoqi@0	348	// ... and these are name characters 'other
aoqi@0	349	// than name start characters'
aoqi@0	350	case Character.COMBINING_SPACING_MARK: // Mc
aoqi@0	351	case Character.ENCLOSING_MARK: // Me
aoqi@0	352	case Character.NON_SPACING_MARK: // Mn
aoqi@0	353	case Character.MODIFIER_LETTER: // Lm
aoqi@0	354	case Character.DECIMAL_DIGIT_NUMBER: // Nd
aoqi@0	355
aoqi@0	356	// OK, here we just have some exceptions to check...
aoqi@0	357	return !isCompatibilityChar(c)
aoqi@0	358	// per "5.14 of Unicode", rule out some combiners
aoqi@0	359	&& !(c >= 0x20dd && c <= 0x20e0);
aoqi@0	360
aoqi@0	361	default:
aoqi@0	362	// added a character ...
aoqi@0	363	return c == 0x0387;
aoqi@0	364	}
aoqi@0	365	}
aoqi@0	366
aoqi@0	367	private static boolean isDigit(char c) {
aoqi@0	368	// [88] Digit ::= ...
aoqi@0	369
aoqi@0	370	//
aoqi@0	371	// java.lang.Character.isDigit is correct from the XML point
aoqi@0	372	// of view except that it allows "fullwidth" digits.
aoqi@0	373	//
aoqi@0	374	return Character.isDigit(c)
aoqi@0	375	&& !((c >= 0xff10) && (c <= 0xff19));
aoqi@0	376	}
aoqi@0	377
aoqi@0	378	private static boolean isExtender(char c) {
aoqi@0	379	// [89] Extender ::= ...
aoqi@0	380	return c == 0x00b7 || c == 0x02d0 || c == 0x02d1 || c == 0x0387
aoqi@0	381	|| c == 0x0640 || c == 0x0e46 || c == 0x0ec6
aoqi@0	382	|| c == 0x3005 || (c >= 0x3031 && c <= 0x3035)
aoqi@0	383	|| (c >= 0x309d && c <= 0x309e)
aoqi@0	384	|| (c >= 0x30fc && c <= 0x30fe)
aoqi@0	385	;
aoqi@0	386	}
aoqi@0	387	}