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 <i18nlangtag/languagetag.hxx>
26 #include <i18nlangtag/languagetagicu.hxx>
27 #include <collator_unicode.hxx>
28 #include <localedata.hxx>
29 #include <com/sun/star/i18n/CollatorOptions.hpp>
31 using namespace ::com::sun::star
;
32 using namespace ::com::sun::star::lang
;
33 using namespace ::com::sun::star::uno
;
34 using namespace ::rtl
;
36 namespace com
{ namespace sun
{ namespace star
{ namespace i18n
{
38 Collator_Unicode::Collator_Unicode()
40 implementationName
= "com.sun.star.i18n.Collator_Unicode";
43 #ifndef DISABLE_DYNLOADING
48 Collator_Unicode::~Collator_Unicode()
50 if (collator
) delete collator
;
51 if (uca_base
) delete uca_base
;
52 #ifndef DISABLE_DYNLOADING
53 if (hModule
) osl_unloadModule(hModule
);
57 #ifdef DISABLE_DYNLOADING
61 // For DISABLE_DYNLOADING the generated functions have names that
62 // start with get_collator_data_ to avoid clashing with a few
63 // functions in the generated libindex_data that are called just
64 // get_zh_pinyin for instance.
66 const sal_uInt8
* get_collator_data_ca_charset();
67 const sal_uInt8
* get_collator_data_dz_charset();
68 const sal_uInt8
* get_collator_data_hu_charset();
69 const sal_uInt8
* get_collator_data_ja_charset();
70 const sal_uInt8
* get_collator_data_ja_phonetic_alphanumeric_first();
71 const sal_uInt8
* get_collator_data_ja_phonetic_alphanumeric_last();
72 const sal_uInt8
* get_collator_data_ko_charset();
73 const sal_uInt8
* get_collator_data_ku_alphanumeric();
74 const sal_uInt8
* get_collator_data_ln_charset();
75 const sal_uInt8
* get_collator_data_my_dictionary();
76 const sal_uInt8
* get_collator_data_ne_charset();
77 const sal_uInt8
* get_collator_data_sid_charset();
78 const sal_uInt8
* get_collator_data_zh_TW_charset();
79 const sal_uInt8
* get_collator_data_zh_TW_radical();
80 const sal_uInt8
* get_collator_data_zh_TW_stroke();
81 const sal_uInt8
* get_collator_data_zh_charset();
82 const sal_uInt8
* get_collator_data_zh_pinyin();
83 const sal_uInt8
* get_collator_data_zh_radical();
84 const sal_uInt8
* get_collator_data_zh_stroke();
85 const sal_uInt8
* get_collator_data_zh_zhuyin();
92 Collator_Unicode::compareSubstring( const OUString
& str1
, sal_Int32 off1
, sal_Int32 len1
,
93 const OUString
& str2
, sal_Int32 off2
, sal_Int32 len2
) throw(RuntimeException
)
95 return collator
->compare(reinterpret_cast<const UChar
*>(str1
.getStr()) + off1
, len1
, reinterpret_cast<const UChar
*>(str2
.getStr()) + off2
, len2
); // UChar != sal_Unicode in MinGW
99 Collator_Unicode::compareString( const OUString
& str1
, const OUString
& str2
) throw(RuntimeException
)
101 return collator
->compare(reinterpret_cast<const UChar
*>(str1
.getStr()), reinterpret_cast<const UChar
*>(str2
.getStr())); // UChar != sal_Unicode in MinGW
104 #ifndef DISABLE_DYNLOADING
106 extern "C" { static void SAL_CALL
thisModule() {} }
111 Collator_Unicode::loadCollatorAlgorithm(const OUString
& rAlgorithm
, const lang::Locale
& rLocale
, sal_Int32 options
)
112 throw(RuntimeException
)
115 UErrorCode status
= U_ZERO_ERROR
;
116 OUString rule
= LocaleData().getCollatorRuleByAlgorithm(rLocale
, rAlgorithm
);
117 if (!rule
.isEmpty()) {
118 collator
= new RuleBasedCollator(reinterpret_cast<const UChar
*>(rule
.getStr()), status
); // UChar != sal_Unicode in MinGW
119 if (! U_SUCCESS(status
)) throw RuntimeException();
121 if (!collator
&& OUString::createFromAscii(LOCAL_RULE_LANGS
).indexOf(rLocale
.Language
) >= 0) {
122 const sal_uInt8
* (*func
)() = NULL
;
124 #ifndef DISABLE_DYNLOADING
127 aBuf
.appendAscii(SAL_DLLPREFIX
);
129 aBuf
.appendAscii( "collator_data" ).appendAscii( SAL_DLLEXTENSION
);
130 hModule
= osl_loadModuleRelative( &thisModule
, aBuf
.makeStringAndClear().pData
, SAL_LOADMODULE_DEFAULT
);
132 aBuf
.appendAscii("get_").append(rLocale
.Language
).appendAscii("_");
133 if ( rLocale
.Language
== "zh" ) {
134 OUString func_base
= aBuf
.makeStringAndClear();
135 if (OUString("TW HK MO").indexOf(rLocale
.Country
) >= 0)
136 func
=(const sal_uInt8
* (*)()) osl_getFunctionSymbol(hModule
,
137 OUString(func_base
+ "TW_" + rAlgorithm
).pData
);
139 func
=(const sal_uInt8
* (*)()) osl_getFunctionSymbol(hModule
, OUString(func_base
+ rAlgorithm
).pData
);
141 if ( rLocale
.Language
== "ja" ) {
142 // replace algorithm name to implementation name.
143 if (rAlgorithm
.equalsAsciiL(RTL_CONSTASCII_STRINGPARAM("phonetic (alphanumeric first)")) )
144 aBuf
.appendAscii("phonetic_alphanumeric_first");
145 else if (rAlgorithm
.equalsAsciiL(RTL_CONSTASCII_STRINGPARAM("phonetic (alphanumeric last)")))
146 aBuf
.appendAscii("phonetic_alphanumeric_last");
148 aBuf
.append(rAlgorithm
);
150 aBuf
.append(rAlgorithm
);
152 func
=(const sal_uInt8
* (*)()) osl_getFunctionSymbol(hModule
, aBuf
.makeStringAndClear().pData
);
156 if ( rLocale
.Language
== "ca" ) {
157 if ( rAlgorithm
== "charset" )
158 func
= get_collator_data_ca_charset
;
159 } else if ( rLocale
.Language
== "dz" ) {
160 if ( rAlgorithm
== "charset" )
161 func
= get_collator_data_dz_charset
;
162 } else if ( rLocale
.Language
== "hu" ) {
163 if ( rAlgorithm
== "charset" )
164 func
= get_collator_data_hu_charset
;
165 } else if ( rLocale
.Language
== "ja" ) {
166 if ( rAlgorithm
== "charset" )
167 func
= get_collator_data_ja_charset
;
168 else if ( rAlgorithm
== "phonetic (alphanumeric first)" )
169 func
= get_collator_data_ja_phonetic_alphanumeric_first
;
170 else if ( rAlgorithm
== "phonetic (alphanumeric last)" )
171 func
= get_collator_data_ja_phonetic_alphanumeric_last
;
172 } else if ( rLocale
.Language
== "ko" ) {
173 if ( rAlgorithm
== "charset" )
174 func
= get_collator_data_ko_charset
;
175 } else if ( rLocale
.Language
== "ku" ) {
176 if ( rAlgorithm
== "alphanumeric" )
177 func
= get_collator_data_ku_alphanumeric
;
178 } else if ( rLocale
.Language
== "ln" ) {
179 if ( rAlgorithm
== "charset" )
180 func
= get_collator_data_ln_charset
;
181 } else if ( rLocale
.Language
== "my" ) {
182 if ( rAlgorithm
== "dictionary" )
183 func
= get_collator_data_my_dictionary
;
184 } else if ( rLocale
.Language
== "ne" ) {
185 if ( rAlgorithm
== "charset" )
186 func
= get_collator_data_ne_charset
;
187 } else if ( rLocale
.Language
== "sid" ) {
188 if ( rAlgorithm
== "charset" )
189 func
= get_collator_data_sid_charset
;
190 } else if ( rLocale
.Language
== "zh" && (rLocale
.Country
== "TW" || rLocale
.Country
== "HK" || rLocale
.Country
== "MO") ) {
191 if ( rAlgorithm
== "charset" )
192 func
= get_collator_data_zh_TW_charset
;
193 else if ( rAlgorithm
== "radical" )
194 func
= get_collator_data_zh_TW_radical
;
195 else if ( rAlgorithm
== "stroke" )
196 func
= get_collator_data_zh_TW_stroke
;
197 } else if ( rLocale
.Language
== "zh" ) {
198 if ( rAlgorithm
== "charset" )
199 func
= get_collator_data_zh_charset
;
200 else if ( rAlgorithm
== "pinyin" )
201 func
= get_collator_data_zh_pinyin
;
202 else if ( rAlgorithm
== "radical" )
203 func
= get_collator_data_zh_radical
;
204 else if ( rAlgorithm
== "stroke" )
205 func
= get_collator_data_zh_stroke
;
206 else if ( rAlgorithm
== "zhuyin" )
207 func
= get_collator_data_zh_zhuyin
;
211 const sal_uInt8
* ruleImage
=func();
212 uca_base
= new RuleBasedCollator(static_cast<UChar
*>(NULL
), status
);
213 if (! U_SUCCESS(status
)) throw RuntimeException();
214 collator
= new RuleBasedCollator(reinterpret_cast<const uint8_t*>(ruleImage
), -1, uca_base
, status
);
215 if (! U_SUCCESS(status
)) throw RuntimeException();
219 /** ICU collators are loaded using a locale only.
220 ICU uses Variant as collation algorithm name (like de__PHONEBOOK
221 locale), note the empty territory (Country) designator in this special
222 case here. The icu::Locale constructor changes the algorithm name to
223 uppercase itself, so we don't have to bother with that.
225 icu::Locale
icuLocale( LanguageTagIcu::getIcuLocale( LanguageTag( rLocale
), rAlgorithm
));
227 collator
= (RuleBasedCollator
*) icu::Collator::createInstance(icuLocale
, status
);
228 if (! U_SUCCESS(status
)) throw RuntimeException();
232 if (options
& CollatorOptions::CollatorOptions_IGNORE_CASE_ACCENT
)
233 collator
->setStrength(Collator::PRIMARY
);
234 else if (options
& CollatorOptions::CollatorOptions_IGNORE_CASE
)
235 collator
->setStrength(Collator::SECONDARY
);
237 collator
->setStrength(Collator::TERTIARY
);
244 Collator_Unicode::getImplementationName() throw( RuntimeException
)
246 return OUString::createFromAscii(implementationName
);
250 Collator_Unicode::supportsService(const OUString
& rServiceName
) throw( RuntimeException
)
252 return !rServiceName
.compareToAscii(implementationName
);
255 Sequence
< OUString
> SAL_CALL
256 Collator_Unicode::getSupportedServiceNames() throw( RuntimeException
)
258 Sequence
< OUString
> aRet(1);
259 aRet
[0] = OUString::createFromAscii(implementationName
);
265 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */