1 /* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
3 * This file is part of the LibreOffice project.
5 * This Source Code Form is subject to the terms of the Mozilla Public
6 * License, v. 2.0. If a copy of the MPL was not distributed with this
7 * file, You can obtain one at http://mozilla.org/MPL/2.0/.
9 * This file incorporates work covered by the following license notice:
11 * Licensed to the Apache Software Foundation (ASF) under one or more
12 * contributor license agreements. See the NOTICE file distributed
13 * with this work for additional information regarding copyright
14 * ownership. The ASF licenses this file to you under the Apache
15 * License, Version 2.0 (the "License"); you may not use this file
16 * except in compliance with the License. You may obtain a copy of
17 * the License at http://www.apache.org/licenses/LICENSE-2.0 .
21 // generated list of languages
22 #include "lrl_include.hxx"
24 #include <rtl/ustrbuf.hxx>
25 #include <collator_unicode.hxx>
26 #include <localedata.hxx>
27 #include <com/sun/star/i18n/CollatorOptions.hpp>
29 using namespace ::com::sun::star
;
30 using namespace ::com::sun::star::lang
;
31 using namespace ::com::sun::star::uno
;
32 using namespace ::rtl
;
34 namespace com
{ namespace sun
{ namespace star
{ namespace i18n
{
36 Collator_Unicode::Collator_Unicode()
38 implementationName
= "com.sun.star.i18n.Collator_Unicode";
41 #ifndef DISABLE_DYNLOADING
46 Collator_Unicode::~Collator_Unicode()
48 if (collator
) delete collator
;
49 if (uca_base
) delete uca_base
;
50 #ifndef DISABLE_DYNLOADING
51 if (hModule
) osl_unloadModule(hModule
);
55 #ifdef DISABLE_DYNLOADING
59 // For DISABLE_DYNLOADING the generated functions have names that
60 // start with get_collator_data_ to avoid clashing with a few
61 // functions in the generated libindex_data that are called just
62 // get_zh_pinyin for instance.
64 const sal_uInt8
* get_collator_data_ca_charset();
65 const sal_uInt8
* get_collator_data_dz_charset();
66 const sal_uInt8
* get_collator_data_hu_charset();
67 const sal_uInt8
* get_collator_data_ja_charset();
68 const sal_uInt8
* get_collator_data_ja_phonetic_alphanumeric_first();
69 const sal_uInt8
* get_collator_data_ja_phonetic_alphanumeric_last();
70 const sal_uInt8
* get_collator_data_ko_charset();
71 const sal_uInt8
* get_collator_data_ku_alphanumeric();
72 const sal_uInt8
* get_collator_data_ln_charset();
73 const sal_uInt8
* get_collator_data_my_dictionary();
74 const sal_uInt8
* get_collator_data_ne_charset();
75 const sal_uInt8
* get_collator_data_zh_TW_charset();
76 const sal_uInt8
* get_collator_data_zh_TW_radical();
77 const sal_uInt8
* get_collator_data_zh_TW_stroke();
78 const sal_uInt8
* get_collator_data_zh_charset();
79 const sal_uInt8
* get_collator_data_zh_pinyin();
80 const sal_uInt8
* get_collator_data_zh_radical();
81 const sal_uInt8
* get_collator_data_zh_stroke();
82 const sal_uInt8
* get_collator_data_zh_zhuyin();
89 Collator_Unicode::compareSubstring( const OUString
& str1
, sal_Int32 off1
, sal_Int32 len1
,
90 const OUString
& str2
, sal_Int32 off2
, sal_Int32 len2
) throw(RuntimeException
)
92 return collator
->compare(reinterpret_cast<const UChar
*>(str1
.getStr()) + off1
, len1
, reinterpret_cast<const UChar
*>(str2
.getStr()) + off2
, len2
); // UChar != sal_Unicode in MinGW
96 Collator_Unicode::compareString( const OUString
& str1
, const OUString
& str2
) throw(RuntimeException
)
98 return collator
->compare(reinterpret_cast<const UChar
*>(str1
.getStr()), reinterpret_cast<const UChar
*>(str2
.getStr())); // UChar != sal_Unicode in MinGW
101 #ifndef DISABLE_DYNLOADING
103 extern "C" { static void SAL_CALL
thisModule() {} }
108 Collator_Unicode::loadCollatorAlgorithm(const OUString
& rAlgorithm
, const lang::Locale
& rLocale
, sal_Int32 options
)
109 throw(RuntimeException
)
112 UErrorCode status
= U_ZERO_ERROR
;
113 OUString rule
= LocaleData().getCollatorRuleByAlgorithm(rLocale
, rAlgorithm
);
114 if (!rule
.isEmpty()) {
115 collator
= new RuleBasedCollator(reinterpret_cast<const UChar
*>(rule
.getStr()), status
); // UChar != sal_Unicode in MinGW
116 if (! U_SUCCESS(status
)) throw RuntimeException();
118 if (!collator
&& OUString::createFromAscii(LOCAL_RULE_LANGS
).indexOf(rLocale
.Language
) >= 0) {
119 const sal_uInt8
* (*func
)() = NULL
;
121 #ifndef DISABLE_DYNLOADING
124 aBuf
.appendAscii(SAL_DLLPREFIX
);
126 aBuf
.appendAscii( "collator_data" ).appendAscii( SAL_DLLEXTENSION
);
127 hModule
= osl_loadModuleRelative( &thisModule
, aBuf
.makeStringAndClear().pData
, SAL_LOADMODULE_DEFAULT
);
129 aBuf
.appendAscii("get_").append(rLocale
.Language
).appendAscii("_");
130 if ( rLocale
.Language
== "zh" ) {
131 OUString func_base
= aBuf
.makeStringAndClear();
132 if (OUString("TW HK MO").indexOf(rLocale
.Country
) >= 0)
133 func
=(const sal_uInt8
* (*)()) osl_getFunctionSymbol(hModule
,
134 OUString(func_base
+ "TW_" + rAlgorithm
).pData
);
136 func
=(const sal_uInt8
* (*)()) osl_getFunctionSymbol(hModule
, OUString(func_base
+ rAlgorithm
).pData
);
138 if ( rLocale
.Language
== "ja" ) {
139 // replace algrithm name to implementation name.
140 if (rAlgorithm
.equalsAsciiL(RTL_CONSTASCII_STRINGPARAM("phonetic (alphanumeric first)")) )
141 aBuf
.appendAscii("phonetic_alphanumeric_first");
142 else if (rAlgorithm
.equalsAsciiL(RTL_CONSTASCII_STRINGPARAM("phonetic (alphanumeric last)")))
143 aBuf
.appendAscii("phonetic_alphanumeric_last");
145 aBuf
.append(rAlgorithm
);
147 aBuf
.append(rAlgorithm
);
149 func
=(const sal_uInt8
* (*)()) osl_getFunctionSymbol(hModule
, aBuf
.makeStringAndClear().pData
);
153 if ( rLocale
.Language
== "ca" ) {
154 if ( rAlgorithm
== "charset" )
155 func
= get_collator_data_ca_charset
;
156 } else if ( rLocale
.Language
== "dz" ) {
157 if ( rAlgorithm
== "charset" )
158 func
= get_collator_data_dz_charset
;
159 } else if ( rLocale
.Language
== "hu" ) {
160 if ( rAlgorithm
== "charset" )
161 func
= get_collator_data_hu_charset
;
162 } else if ( rLocale
.Language
== "ja" ) {
163 if ( rAlgorithm
== "charset" )
164 func
= get_collator_data_hu_charset
;
165 else if ( rAlgorithm
== "phonetic (alphanumeric first)" )
166 func
= get_collator_data_ja_phonetic_alphanumeric_first
;
167 else if ( rAlgorithm
== "phonetic (alphanumeric last)" )
168 func
= get_collator_data_ja_phonetic_alphanumeric_last
;
169 } else if ( rLocale
.Language
== "ko" ) {
170 if ( rAlgorithm
== "charset" )
171 func
= get_collator_data_ko_charset
;
172 } else if ( rLocale
.Language
== "ku" ) {
173 if ( rAlgorithm
== "alphanumeric" )
174 func
= get_collator_data_ku_alphanumeric
;
175 } else if ( rLocale
.Language
== "ln" ) {
176 if ( rAlgorithm
== "charset" )
177 func
= get_collator_data_ln_charset
;
178 } else if ( rLocale
.Language
== "my" ) {
179 if ( rAlgorithm
== "dictionary" )
180 func
= get_collator_data_my_dictionary
;
181 } else if ( rLocale
.Language
== "ne" ) {
182 if ( rAlgorithm
== "charset" )
183 func
= get_collator_data_ne_charset
;
184 } else if ( rLocale
.Language
== "zh" && (rLocale
.Country
== "TW" || rLocale
.Country
== "HK" || rLocale
.Country
== "MO") ) {
185 if ( rAlgorithm
== "charset" )
186 func
= get_collator_data_zh_TW_charset
;
187 else if ( rAlgorithm
== "radical" )
188 func
= get_collator_data_zh_TW_radical
;
189 else if ( rAlgorithm
== "stroke" )
190 func
= get_collator_data_zh_TW_stroke
;
191 } else if ( rLocale
.Language
== "zh" ) {
192 if ( rAlgorithm
== "charset" )
193 func
= get_collator_data_zh_charset
;
194 else if ( rAlgorithm
== "pinyin" )
195 func
= get_collator_data_zh_pinyin
;
196 else if ( rAlgorithm
== "radical" )
197 func
= get_collator_data_zh_radical
;
198 else if ( rAlgorithm
== "stroke" )
199 func
= get_collator_data_zh_stroke
;
200 else if ( rAlgorithm
== "zhuyin" )
201 func
= get_collator_data_zh_zhuyin
;
205 const sal_uInt8
* ruleImage
=func();
206 uca_base
= new RuleBasedCollator(static_cast<UChar
*>(NULL
), status
);
207 if (! U_SUCCESS(status
)) throw RuntimeException();
208 collator
= new RuleBasedCollator(reinterpret_cast<const uint8_t*>(ruleImage
), -1, uca_base
, status
);
209 if (! U_SUCCESS(status
)) throw RuntimeException();
213 /** ICU collators are loaded using a locale only.
214 ICU uses Variant as collation algorithm name (like de__PHONEBOOK
215 locale), note the empty territory (Country) designator in this special
216 case here. The icu::Locale contructor changes the algorithm name to
217 uppercase itself, so we don't have to bother with that.
219 icu::Locale
icuLocale(
220 OUStringToOString(rLocale
.Language
, RTL_TEXTENCODING_ASCII_US
).getStr(),
221 OUStringToOString(rLocale
.Country
, RTL_TEXTENCODING_ASCII_US
).getStr(),
222 OUStringToOString(rAlgorithm
, RTL_TEXTENCODING_ASCII_US
).getStr());
224 collator
= (RuleBasedCollator
*) icu::Collator::createInstance(icuLocale
, status
);
225 if (! U_SUCCESS(status
)) throw RuntimeException();
229 if (options
& CollatorOptions::CollatorOptions_IGNORE_CASE_ACCENT
)
230 collator
->setStrength(Collator::PRIMARY
);
231 else if (options
& CollatorOptions::CollatorOptions_IGNORE_CASE
)
232 collator
->setStrength(Collator::SECONDARY
);
234 collator
->setStrength(Collator::TERTIARY
);
241 Collator_Unicode::getImplementationName() throw( RuntimeException
)
243 return OUString::createFromAscii(implementationName
);
247 Collator_Unicode::supportsService(const rtl::OUString
& rServiceName
) throw( RuntimeException
)
249 return !rServiceName
.compareToAscii(implementationName
);
252 Sequence
< OUString
> SAL_CALL
253 Collator_Unicode::getSupportedServiceNames() throw( RuntimeException
)
255 Sequence
< OUString
> aRet(1);
256 aRet
[0] = OUString::createFromAscii(implementationName
);
262 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */