Mercurial > hg > graal-compiler

annotate src/share/vm/utilities/utf8.hpp @ 13364:5a4293f24642

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
added -G:PrintCompRate option for periodically printing out the current compilation rate
author Doug Simon <doug.simon@oracle.com>
date Tue, 17 Dec 2013 16:45:02 +0100
parents 8c03fc47511d
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
a61af66fc99e Initial load
duke
parents:
diff changeset
1 /*
8851
8c03fc47511d 8011048: Possible reading from unmapped memory in UTF8::as_quoted_ascii()
iklam
parents: 6972
diff changeset
2 * Copyright (c) 1997, 2013, 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:
6972
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
35 // returns the unicode length of a 0-terminated utf8 string
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
36 static int unicode_length(const char* utf8_str);
0
a61af66fc99e Initial load
duke
parents:
diff changeset
37
6972
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
38 // returns the unicode length of a non-0-terminated utf8 string
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
39 static int unicode_length(const char* utf8_str, int len);
0
a61af66fc99e Initial load
duke
parents:
diff changeset
40
6972
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
41 // converts a utf8 string to a unicode string
0
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
6972
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
44 // returns the quoted ascii length of a utf8 string
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
45 static int quoted_ascii_length(const char* utf8_str, int utf8_length);
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
46
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
47 // converts a utf8 string to quoted ascii
8851
8c03fc47511d 8011048: Possible reading from unmapped memory in UTF8::as_quoted_ascii()
iklam
parents: 6972
diff changeset
48 static void as_quoted_ascii(const char* utf8_str, int utf8_length, char* buf, int buflen);
6972
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
49
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
50 // converts a quoted ascii string to utf8 string. returns the original
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
51 // string unchanged if nothing needs to be done.
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
52 static const char* from_quoted_ascii(const char* quoted_ascii_string);
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
53
0
a61af66fc99e Initial load
duke
parents:
diff changeset
54 // decodes the current utf8 character, stores the result in value,
6972
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
55 // and returns the end of the current utf8 chararacter.
0
a61af66fc99e Initial load
duke
parents:
diff changeset
56 static char* next(const char* str, jchar* value);
a61af66fc99e Initial load
duke
parents:
diff changeset
57
a61af66fc99e Initial load
duke
parents:
diff changeset
58 // decodes the current utf8 character, gets the supplementary character instead of
a61af66fc99e Initial load
duke
parents:
diff changeset
59 // the surrogate pair when seeing a supplementary character in string,
6972
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
60 // stores the result in value, and returns the end of the current utf8 chararacter.
0
a61af66fc99e Initial load
duke
parents:
diff changeset
61 static char* next_character(const char* str, jint* value);
a61af66fc99e Initial load
duke
parents:
diff changeset
62
a61af66fc99e Initial load
duke
parents:
diff changeset
63 // Utility methods
2177
3582bf76420e 6990754: Use native memory and reference counting to implement SymbolTable
coleenp
parents: 1972
diff changeset
64 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
65 static bool equal(const jbyte* base1, int length1, const jbyte* base2,int length2);
0
a61af66fc99e Initial load
duke
parents:
diff changeset
66 static bool is_supplementary_character(const unsigned char* str);
a61af66fc99e Initial load
duke
parents:
diff changeset
67 static jint get_supplementary_character(const unsigned char* str);
a61af66fc99e Initial load
duke
parents:
diff changeset
68 };
a61af66fc99e Initial load
duke
parents:
diff changeset
69
a61af66fc99e Initial load
duke
parents:
diff changeset
70
a61af66fc99e Initial load
duke
parents:
diff changeset
71 // Low-level interface for UNICODE strings
a61af66fc99e Initial load
duke
parents:
diff changeset
72
a61af66fc99e Initial load
duke
parents:
diff changeset
73 // A unicode string represents a string in the UTF-16 format in which supplementary
a61af66fc99e Initial load
duke
parents:
diff changeset
74 // characters are represented by surrogate pairs. Index values refer to char code
a61af66fc99e Initial load
duke
parents:
diff changeset
75 // units, so a supplementary character uses two positions in a unicode string.
a61af66fc99e Initial load
duke
parents:
diff changeset
76
a61af66fc99e Initial load
duke
parents:
diff changeset
77 class UNICODE : AllStatic {
a61af66fc99e Initial load
duke
parents:
diff changeset
78 public:
a61af66fc99e Initial load
duke
parents:
diff changeset
79 // returns the utf8 size of a unicode character
a61af66fc99e Initial load
duke
parents:
diff changeset
80 static int utf8_size(jchar c);
a61af66fc99e Initial load
duke
parents:
diff changeset
81
a61af66fc99e Initial load
duke
parents:
diff changeset
82 // returns the utf8 length of a unicode string
a61af66fc99e Initial load
duke
parents:
diff changeset
83 static int utf8_length(jchar* base, int length);
a61af66fc99e Initial load
duke
parents:
diff changeset
84
a61af66fc99e Initial load
duke
parents:
diff changeset
85 // converts a unicode string to utf8 string
a61af66fc99e Initial load
duke
parents:
diff changeset
86 static void convert_to_utf8(const jchar* base, int length, char* utf8_buffer);
a61af66fc99e Initial load
duke
parents:
diff changeset
87
a61af66fc99e Initial load
duke
parents:
diff changeset
88 // converts a unicode string to a utf8 string; result is allocated
a61af66fc99e Initial load
duke
parents:
diff changeset
89 // in resource area unless a buffer is provided.
a61af66fc99e Initial load
duke
parents:
diff changeset
90 static char* as_utf8(jchar* base, int length);
a61af66fc99e Initial load
duke
parents:
diff changeset
91 static char* as_utf8(jchar* base, int length, char* buf, int buflen);
6972
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
92
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
93 // returns the quoted ascii length of a unicode string
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
94 static int quoted_ascii_length(jchar* base, int length);
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
95
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
96 // converts a utf8 string to quoted ascii
bd7a7ce2e264 6830717: replay of compilations would help with debugging
minqi
parents: 2426
diff changeset
97 static void as_quoted_ascii(const jchar* base, int length, char* buf, int buflen);
0
a61af66fc99e Initial load
duke
parents:
diff changeset
98 };
1972
f95d63e2154a 6989984: Use standard include model for Hospot
stefank
parents: 1552
diff changeset
99
f95d63e2154a 6989984: Use standard include model for Hospot
stefank
parents: 1552
diff changeset
100 #endif // SHARE_VM_UTILITIES_UTF8_HPP