Wed, 03 Jul 2019 20:42:37 +0800
Merge
aoqi@0 | 1 | /* |
aoqi@0 | 2 | * Copyright (c) 1997, 2013, 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. |
aoqi@0 | 8 | * |
aoqi@0 | 9 | * This code is distributed in the hope that it will be useful, but WITHOUT |
aoqi@0 | 10 | * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or |
aoqi@0 | 11 | * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License |
aoqi@0 | 12 | * version 2 for more details (a copy is included in the LICENSE file that |
aoqi@0 | 13 | * accompanied this code). |
aoqi@0 | 14 | * |
aoqi@0 | 15 | * You should have received a copy of the GNU General Public License version |
aoqi@0 | 16 | * 2 along with this work; if not, write to the Free Software Foundation, |
aoqi@0 | 17 | * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. |
aoqi@0 | 18 | * |
aoqi@0 | 19 | * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA |
aoqi@0 | 20 | * or visit www.oracle.com if you need additional information or have any |
aoqi@0 | 21 | * questions. |
aoqi@0 | 22 | * |
aoqi@0 | 23 | */ |
aoqi@0 | 24 | |
aoqi@0 | 25 | #ifndef SHARE_VM_UTILITIES_UTF8_HPP |
aoqi@0 | 26 | #define SHARE_VM_UTILITIES_UTF8_HPP |
aoqi@0 | 27 | |
aoqi@0 | 28 | #include "memory/allocation.hpp" |
aoqi@0 | 29 | #include "utilities/top.hpp" |
aoqi@0 | 30 | |
aoqi@0 | 31 | // Low-level interface for UTF8 strings |
aoqi@0 | 32 | |
aoqi@0 | 33 | class UTF8 : AllStatic { |
aoqi@0 | 34 | public: |
aoqi@0 | 35 | // returns the unicode length of a 0-terminated utf8 string |
aoqi@0 | 36 | static int unicode_length(const char* utf8_str); |
aoqi@0 | 37 | |
aoqi@0 | 38 | // returns the unicode length of a non-0-terminated utf8 string |
aoqi@0 | 39 | static int unicode_length(const char* utf8_str, int len); |
aoqi@0 | 40 | |
aoqi@0 | 41 | // converts a utf8 string to a unicode string |
aoqi@0 | 42 | static void convert_to_unicode(const char* utf8_str, jchar* unicode_buffer, int unicode_length); |
aoqi@0 | 43 | |
aoqi@0 | 44 | // returns the quoted ascii length of a utf8 string |
aoqi@0 | 45 | static int quoted_ascii_length(const char* utf8_str, int utf8_length); |
aoqi@0 | 46 | |
aoqi@0 | 47 | // converts a utf8 string to quoted ascii |
aoqi@0 | 48 | static void as_quoted_ascii(const char* utf8_str, int utf8_length, char* buf, int buflen); |
aoqi@0 | 49 | |
aoqi@0 | 50 | // converts a quoted ascii string to utf8 string. returns the original |
aoqi@0 | 51 | // string unchanged if nothing needs to be done. |
aoqi@0 | 52 | static const char* from_quoted_ascii(const char* quoted_ascii_string); |
aoqi@0 | 53 | |
aoqi@0 | 54 | // decodes the current utf8 character, stores the result in value, |
aoqi@0 | 55 | // and returns the end of the current utf8 chararacter. |
aoqi@0 | 56 | static char* next(const char* str, jchar* value); |
aoqi@0 | 57 | |
aoqi@0 | 58 | // decodes the current utf8 character, gets the supplementary character instead of |
aoqi@0 | 59 | // the surrogate pair when seeing a supplementary character in string, |
aoqi@0 | 60 | // stores the result in value, and returns the end of the current utf8 chararacter. |
aoqi@0 | 61 | static char* next_character(const char* str, jint* value); |
aoqi@0 | 62 | |
aoqi@0 | 63 | // Utility methods |
aoqi@0 | 64 | static const jbyte* strrchr(const jbyte* base, int length, jbyte c); |
aoqi@0 | 65 | static bool equal(const jbyte* base1, int length1, const jbyte* base2,int length2); |
aoqi@0 | 66 | static bool is_supplementary_character(const unsigned char* str); |
aoqi@0 | 67 | static jint get_supplementary_character(const unsigned char* str); |
aoqi@0 | 68 | }; |
aoqi@0 | 69 | |
aoqi@0 | 70 | |
aoqi@0 | 71 | // Low-level interface for UNICODE strings |
aoqi@0 | 72 | |
aoqi@0 | 73 | // A unicode string represents a string in the UTF-16 format in which supplementary |
aoqi@0 | 74 | // characters are represented by surrogate pairs. Index values refer to char code |
aoqi@0 | 75 | // units, so a supplementary character uses two positions in a unicode string. |
aoqi@0 | 76 | |
aoqi@0 | 77 | class UNICODE : AllStatic { |
aoqi@0 | 78 | public: |
aoqi@0 | 79 | // returns the utf8 size of a unicode character |
aoqi@0 | 80 | static int utf8_size(jchar c); |
aoqi@0 | 81 | |
aoqi@0 | 82 | // returns the utf8 length of a unicode string |
aoqi@0 | 83 | static int utf8_length(jchar* base, int length); |
aoqi@0 | 84 | |
aoqi@0 | 85 | // converts a unicode string to utf8 string |
aoqi@0 | 86 | static void convert_to_utf8(const jchar* base, int length, char* utf8_buffer); |
aoqi@0 | 87 | |
aoqi@0 | 88 | // converts a unicode string to a utf8 string; result is allocated |
aoqi@0 | 89 | // in resource area unless a buffer is provided. |
aoqi@0 | 90 | static char* as_utf8(jchar* base, int length); |
aoqi@0 | 91 | static char* as_utf8(jchar* base, int length, char* buf, int buflen); |
aoqi@0 | 92 | |
aoqi@0 | 93 | // returns the quoted ascii length of a unicode string |
aoqi@0 | 94 | static int quoted_ascii_length(jchar* base, int length); |
aoqi@0 | 95 | |
aoqi@0 | 96 | // converts a utf8 string to quoted ascii |
aoqi@0 | 97 | static void as_quoted_ascii(const jchar* base, int length, char* buf, int buflen); |
aoqi@0 | 98 | }; |
aoqi@0 | 99 | |
aoqi@0 | 100 | #endif // SHARE_VM_UTILITIES_UTF8_HPP |