annotate src/share/vm/utilities/utf8.hpp @ 2177:3582bf76420e

6990754: Use native memory and reference counting to implement SymbolTable Summary: move symbols from permgen into C heap and reference count them Reviewed-by: never, acorn, jmasa, stefank
author coleenp
date Thu, 27 Jan 2011 16:11:27 -0800
parents f95d63e2154a
children 1d1603768966
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
a61af66fc99e Initial load
duke
parents:
diff changeset
1 /*
1972
f95d63e2154a 6989984: Use standard include model for Hospot
stefank
parents: 1552
diff changeset
2 * Copyright (c) 1997, 2010, Oracle and/or its affiliates. All rights reserved.
0
a61af66fc99e Initial load
duke
parents:
diff changeset
3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
a61af66fc99e Initial load
duke
parents:
diff changeset
4 *
a61af66fc99e Initial load
duke
parents:
diff changeset
5 * This code is free software; you can redistribute it and/or modify it
a61af66fc99e Initial load
duke
parents:
diff changeset
6 * under the terms of the GNU General Public License version 2 only, as
a61af66fc99e Initial load
duke
parents:
diff changeset
7 * published by the Free Software Foundation.
a61af66fc99e Initial load
duke
parents:
diff changeset
8 *
a61af66fc99e Initial load
duke
parents:
diff changeset
9 * This code is distributed in the hope that it will be useful, but WITHOUT
a61af66fc99e Initial load
duke
parents:
diff changeset
10 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
a61af66fc99e Initial load
duke
parents:
diff changeset
11 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
a61af66fc99e Initial load
duke
parents:
diff changeset
12 * version 2 for more details (a copy is included in the LICENSE file that
a61af66fc99e Initial load
duke
parents:
diff changeset
13 * accompanied this code).
a61af66fc99e Initial load
duke
parents:
diff changeset
14 *
a61af66fc99e Initial load
duke
parents:
diff changeset
15 * You should have received a copy of the GNU General Public License version
a61af66fc99e Initial load
duke
parents:
diff changeset
16 * 2 along with this work; if not, write to the Free Software Foundation,
a61af66fc99e Initial load
duke
parents:
diff changeset
17 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
a61af66fc99e Initial load
duke
parents:
diff changeset
18 *
1552
c18cbe5936b8 6941466: Oracle rebranding changes for Hotspot repositories
trims
parents: 0
diff changeset
19 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
c18cbe5936b8 6941466: Oracle rebranding changes for Hotspot repositories
trims
parents: 0
diff changeset
20 * or visit www.oracle.com if you need additional information or have any
c18cbe5936b8 6941466: Oracle rebranding changes for Hotspot repositories
trims
parents: 0
diff changeset
21 * questions.
0
a61af66fc99e Initial load
duke
parents:
diff changeset
22 *
a61af66fc99e Initial load
duke
parents:
diff changeset
23 */
a61af66fc99e Initial load
duke
parents:
diff changeset
24
1972
f95d63e2154a 6989984: Use standard include model for Hospot
stefank
parents: 1552
diff changeset
25 #ifndef SHARE_VM_UTILITIES_UTF8_HPP
f95d63e2154a 6989984: Use standard include model for Hospot
stefank
parents: 1552
diff changeset
26 #define SHARE_VM_UTILITIES_UTF8_HPP
f95d63e2154a 6989984: Use standard include model for Hospot
stefank
parents: 1552
diff changeset
27
f95d63e2154a 6989984: Use standard include model for Hospot
stefank
parents: 1552
diff changeset
28 #include "memory/allocation.hpp"
f95d63e2154a 6989984: Use standard include model for Hospot
stefank
parents: 1552
diff changeset
29 #include "utilities/top.hpp"
f95d63e2154a 6989984: Use standard include model for Hospot
stefank
parents: 1552
diff changeset
30
0
a61af66fc99e Initial load
duke
parents:
diff changeset
31 // Low-level interface for UTF8 strings
a61af66fc99e Initial load
duke
parents:
diff changeset
32
a61af66fc99e Initial load
duke
parents:
diff changeset
33 class UTF8 : AllStatic {
a61af66fc99e Initial load
duke
parents:
diff changeset
34 public:
a61af66fc99e Initial load
duke
parents:
diff changeset
35 // returns the unicode length of a 0-terminated uft8 string
a61af66fc99e Initial load
duke
parents:
diff changeset
36 static int unicode_length(const char* uft8_str);
a61af66fc99e Initial load
duke
parents:
diff changeset
37
a61af66fc99e Initial load
duke
parents:
diff changeset
38 // returns the unicode length of a non-0-terminated uft8 string
a61af66fc99e Initial load
duke
parents:
diff changeset
39 static int unicode_length(const char* uft8_str, int len);
a61af66fc99e Initial load
duke
parents:
diff changeset
40
a61af66fc99e Initial load
duke
parents:
diff changeset
41 // converts a uft8 string to a unicode string
a61af66fc99e Initial load
duke
parents:
diff changeset
42 static void convert_to_unicode(const char* utf8_str, jchar* unicode_buffer, int unicode_length);
a61af66fc99e Initial load
duke
parents:
diff changeset
43
a61af66fc99e Initial load
duke
parents:
diff changeset
44 // decodes the current utf8 character, stores the result in value,
a61af66fc99e Initial load
duke
parents:
diff changeset
45 // and returns the end of the current uft8 chararacter.
a61af66fc99e Initial load
duke
parents:
diff changeset
46 static char* next(const char* str, jchar* value);
a61af66fc99e Initial load
duke
parents:
diff changeset
47
a61af66fc99e Initial load
duke
parents:
diff changeset
48 // decodes the current utf8 character, gets the supplementary character instead of
a61af66fc99e Initial load
duke
parents:
diff changeset
49 // the surrogate pair when seeing a supplementary character in string,
a61af66fc99e Initial load
duke
parents:
diff changeset
50 // stores the result in value, and returns the end of the current uft8 chararacter.
a61af66fc99e Initial load
duke
parents:
diff changeset
51 static char* next_character(const char* str, jint* value);
a61af66fc99e Initial load
duke
parents:
diff changeset
52
a61af66fc99e Initial load
duke
parents:
diff changeset
53 // Utility methods
2177
3582bf76420e 6990754: Use native memory and reference counting to implement SymbolTable
coleenp
parents: 1972
diff changeset
54 static const jbyte* strrchr(const jbyte* base, int length, jbyte c);
3582bf76420e 6990754: Use native memory and reference counting to implement SymbolTable
coleenp
parents: 1972
diff changeset
55 static bool equal(const jbyte* base1, int length1, const jbyte* base2,int length2);
0
a61af66fc99e Initial load
duke
parents:
diff changeset
56 static bool is_supplementary_character(const unsigned char* str);
a61af66fc99e Initial load
duke
parents:
diff changeset
57 static jint get_supplementary_character(const unsigned char* str);
a61af66fc99e Initial load
duke
parents:
diff changeset
58 };
a61af66fc99e Initial load
duke
parents:
diff changeset
59
a61af66fc99e Initial load
duke
parents:
diff changeset
60
a61af66fc99e Initial load
duke
parents:
diff changeset
61 // Low-level interface for UNICODE strings
a61af66fc99e Initial load
duke
parents:
diff changeset
62
a61af66fc99e Initial load
duke
parents:
diff changeset
63 // A unicode string represents a string in the UTF-16 format in which supplementary
a61af66fc99e Initial load
duke
parents:
diff changeset
64 // characters are represented by surrogate pairs. Index values refer to char code
a61af66fc99e Initial load
duke
parents:
diff changeset
65 // units, so a supplementary character uses two positions in a unicode string.
a61af66fc99e Initial load
duke
parents:
diff changeset
66
a61af66fc99e Initial load
duke
parents:
diff changeset
67 class UNICODE : AllStatic {
a61af66fc99e Initial load
duke
parents:
diff changeset
68 public:
a61af66fc99e Initial load
duke
parents:
diff changeset
69 // returns the utf8 size of a unicode character
a61af66fc99e Initial load
duke
parents:
diff changeset
70 static int utf8_size(jchar c);
a61af66fc99e Initial load
duke
parents:
diff changeset
71
a61af66fc99e Initial load
duke
parents:
diff changeset
72 // returns the utf8 length of a unicode string
a61af66fc99e Initial load
duke
parents:
diff changeset
73 static int utf8_length(jchar* base, int length);
a61af66fc99e Initial load
duke
parents:
diff changeset
74
a61af66fc99e Initial load
duke
parents:
diff changeset
75 // converts a unicode string to utf8 string
a61af66fc99e Initial load
duke
parents:
diff changeset
76 static void convert_to_utf8(const jchar* base, int length, char* utf8_buffer);
a61af66fc99e Initial load
duke
parents:
diff changeset
77
a61af66fc99e Initial load
duke
parents:
diff changeset
78 // converts a unicode string to a utf8 string; result is allocated
a61af66fc99e Initial load
duke
parents:
diff changeset
79 // in resource area unless a buffer is provided.
a61af66fc99e Initial load
duke
parents:
diff changeset
80 static char* as_utf8(jchar* base, int length);
a61af66fc99e Initial load
duke
parents:
diff changeset
81 static char* as_utf8(jchar* base, int length, char* buf, int buflen);
a61af66fc99e Initial load
duke
parents:
diff changeset
82 };
1972
f95d63e2154a 6989984: Use standard include model for Hospot
stefank
parents: 1552
diff changeset
83
f95d63e2154a 6989984: Use standard include model for Hospot
stefank
parents: 1552
diff changeset
84 #endif // SHARE_VM_UTILITIES_UTF8_HPP