aoqi@0: /*
aoqi@0:  * Copyright (c) 1997, 2013, Oracle and/or its affiliates. All rights reserved.
aoqi@0:  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
aoqi@0:  *
aoqi@0:  * This code is free software; you can redistribute it and/or modify it
aoqi@0:  * under the terms of the GNU General Public License version 2 only, as
aoqi@0:  * published by the Free Software Foundation.
aoqi@0:  *
aoqi@0:  * This code is distributed in the hope that it will be useful, but WITHOUT
aoqi@0:  * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
aoqi@0:  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
aoqi@0:  * version 2 for more details (a copy is included in the LICENSE file that
aoqi@0:  * accompanied this code).
aoqi@0:  *
aoqi@0:  * You should have received a copy of the GNU General Public License version
aoqi@0:  * 2 along with this work; if not, write to the Free Software Foundation,
aoqi@0:  * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
aoqi@0:  *
aoqi@0:  * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
aoqi@0:  * or visit www.oracle.com if you need additional information or have any
aoqi@0:  * questions.
aoqi@0:  *
aoqi@0:  */
aoqi@0: 
aoqi@0: #ifndef SHARE_VM_UTILITIES_UTF8_HPP
aoqi@0: #define SHARE_VM_UTILITIES_UTF8_HPP
aoqi@0: 
aoqi@0: #include "memory/allocation.hpp"
aoqi@0: #include "utilities/top.hpp"
aoqi@0: 
aoqi@0: // Low-level interface for UTF8 strings
aoqi@0: 
aoqi@0: class UTF8 : AllStatic {
aoqi@0:  public:
aoqi@0:   // returns the unicode length of a 0-terminated utf8 string
aoqi@0:   static int unicode_length(const char* utf8_str);
aoqi@0: 
aoqi@0:   // returns the unicode length of a non-0-terminated utf8 string
aoqi@0:   static int unicode_length(const char* utf8_str, int len);
aoqi@0: 
aoqi@0:   // converts a utf8 string to a unicode string
aoqi@0:   static void convert_to_unicode(const char* utf8_str, jchar* unicode_buffer, int unicode_length);
aoqi@0: 
aoqi@0:   // returns the quoted ascii length of a utf8 string
aoqi@0:   static int quoted_ascii_length(const char* utf8_str, int utf8_length);
aoqi@0: 
aoqi@0:   // converts a utf8 string to quoted ascii
aoqi@0:   static void as_quoted_ascii(const char* utf8_str, int utf8_length, char* buf, int buflen);
aoqi@0: 
aoqi@0:   // converts a quoted ascii string to utf8 string.  returns the original
aoqi@0:   // string unchanged if nothing needs to be done.
aoqi@0:   static const char* from_quoted_ascii(const char* quoted_ascii_string);
aoqi@0: 
aoqi@0:   // decodes the current utf8 character, stores the result in value,
aoqi@0:   // and returns the end of the current utf8 chararacter.
aoqi@0:   static char* next(const char* str, jchar* value);
aoqi@0: 
aoqi@0:   // decodes the current utf8 character, gets the supplementary character instead of
aoqi@0:   // the surrogate pair when seeing a supplementary character in string,
aoqi@0:   // stores the result in value, and returns the end of the current utf8 chararacter.
aoqi@0:   static char* next_character(const char* str, jint* value);
aoqi@0: 
aoqi@0:   // Utility methods
aoqi@0:   static const jbyte* strrchr(const jbyte* base, int length, jbyte c);
aoqi@0:   static bool   equal(const jbyte* base1, int length1, const jbyte* base2,int length2);
aoqi@0:   static bool   is_supplementary_character(const unsigned char* str);
aoqi@0:   static jint   get_supplementary_character(const unsigned char* str);
aoqi@0: };
aoqi@0: 
aoqi@0: 
aoqi@0: // Low-level interface for UNICODE strings
aoqi@0: 
aoqi@0: // A unicode string represents a string in the UTF-16 format in which supplementary
aoqi@0: // characters are represented by surrogate pairs. Index values refer to char code
aoqi@0: // units, so a supplementary character uses two positions in a unicode string.
aoqi@0: 
aoqi@0: class UNICODE : AllStatic {
aoqi@0:  public:
aoqi@0:   // returns the utf8 size of a unicode character
aoqi@0:   static int utf8_size(jchar c);
aoqi@0: 
aoqi@0:   // returns the utf8 length of a unicode string
aoqi@0:   static int utf8_length(jchar* base, int length);
aoqi@0: 
aoqi@0:   // converts a unicode string to utf8 string
aoqi@0:   static void convert_to_utf8(const jchar* base, int length, char* utf8_buffer);
aoqi@0: 
aoqi@0:   // converts a unicode string to a utf8 string; result is allocated
aoqi@0:   // in resource area unless a buffer is provided.
aoqi@0:   static char* as_utf8(jchar* base, int length);
aoqi@0:   static char* as_utf8(jchar* base, int length, char* buf, int buflen);
aoqi@0: 
aoqi@0:   // returns the quoted ascii length of a unicode string
aoqi@0:   static int quoted_ascii_length(jchar* base, int length);
aoqi@0: 
aoqi@0:   // converts a utf8 string to quoted ascii
aoqi@0:   static void as_quoted_ascii(const jchar* base, int length, char* buf, int buflen);
aoqi@0: };
aoqi@0: 
aoqi@0: #endif // SHARE_VM_UTILITIES_UTF8_HPP