| 1 | /****************************************************************************
|
|---|
| 2 | **
|
|---|
| 3 | ** Copyright (C) 2011 Nokia Corporation and/or its subsidiary(-ies).
|
|---|
| 4 | ** All rights reserved.
|
|---|
| 5 | ** Contact: Nokia Corporation ([email protected])
|
|---|
| 6 | **
|
|---|
| 7 | ** This file is part of the QtCore module of the Qt Toolkit.
|
|---|
| 8 | **
|
|---|
| 9 | ** $QT_BEGIN_LICENSE:LGPL$
|
|---|
| 10 | ** Commercial Usage
|
|---|
| 11 | ** Licensees holding valid Qt Commercial licenses may use this file in
|
|---|
| 12 | ** accordance with the Qt Commercial License Agreement provided with the
|
|---|
| 13 | ** Software or, alternatively, in accordance with the terms contained in
|
|---|
| 14 | ** a written agreement between you and Nokia.
|
|---|
| 15 | **
|
|---|
| 16 | ** GNU Lesser General Public License Usage
|
|---|
| 17 | ** Alternatively, this file may be used under the terms of the GNU Lesser
|
|---|
| 18 | ** General Public License version 2.1 as published by the Free Software
|
|---|
| 19 | ** Foundation and appearing in the file LICENSE.LGPL included in the
|
|---|
| 20 | ** packaging of this file. Please review the following information to
|
|---|
| 21 | ** ensure the GNU Lesser General Public License version 2.1 requirements
|
|---|
| 22 | ** will be met: http://www.gnu.org/licenses/old-licenses/lgpl-2.1.html.
|
|---|
| 23 | **
|
|---|
| 24 | ** In addition, as a special exception, Nokia gives you certain additional
|
|---|
| 25 | ** rights. These rights are described in the Nokia Qt LGPL Exception
|
|---|
| 26 | ** version 1.1, included in the file LGPL_EXCEPTION.txt in this package.
|
|---|
| 27 | **
|
|---|
| 28 | ** GNU General Public License Usage
|
|---|
| 29 | ** Alternatively, this file may be used under the terms of the GNU
|
|---|
| 30 | ** General Public License version 3.0 as published by the Free Software
|
|---|
| 31 | ** Foundation and appearing in the file LICENSE.GPL included in the
|
|---|
| 32 | ** packaging of this file. Please review the following information to
|
|---|
| 33 | ** ensure the GNU General Public License version 3.0 requirements will be
|
|---|
| 34 | ** met: http://www.gnu.org/copyleft/gpl.html.
|
|---|
| 35 | **
|
|---|
| 36 | ** If you have questions regarding the use of this file, please contact
|
|---|
| 37 | ** Nokia at [email protected].
|
|---|
| 38 | ** $QT_END_LICENSE$
|
|---|
| 39 | **
|
|---|
| 40 | ****************************************************************************/
|
|---|
| 41 |
|
|---|
| 42 | #include "qiconvcodec_p.h"
|
|---|
| 43 | #include "qtextcodec_p.h"
|
|---|
| 44 | #include <qlibrary.h>
|
|---|
| 45 | #include <qdebug.h>
|
|---|
| 46 | #include <qthreadstorage.h>
|
|---|
| 47 |
|
|---|
| 48 | #include <errno.h>
|
|---|
| 49 | #include <locale.h>
|
|---|
| 50 | #include <stdio.h>
|
|---|
| 51 | #include <dlfcn.h>
|
|---|
| 52 |
|
|---|
| 53 | // unistd.h is needed for the _XOPEN_UNIX macro
|
|---|
| 54 | #include <unistd.h>
|
|---|
| 55 | #if defined(_XOPEN_UNIX) && !defined(Q_OS_QNX) && !defined(Q_OS_OSF)
|
|---|
| 56 | # include <langinfo.h>
|
|---|
| 57 | #endif
|
|---|
| 58 |
|
|---|
| 59 | #if defined(Q_OS_HPUX)
|
|---|
| 60 | # define NO_BOM
|
|---|
| 61 | # define UTF16 "ucs2"
|
|---|
| 62 | #elif defined(Q_OS_AIX)
|
|---|
| 63 | # define NO_BOM
|
|---|
| 64 | # define UTF16 "UCS-2"
|
|---|
| 65 | #elif defined(Q_OS_FREEBSD) || defined(Q_OS_MAC)
|
|---|
| 66 | # define NO_BOM
|
|---|
| 67 | # if Q_BYTE_ORDER == Q_BIG_ENDIAN
|
|---|
| 68 | # define UTF16 "UTF-16BE"
|
|---|
| 69 | # else
|
|---|
| 70 | # define UTF16 "UTF-16LE"
|
|---|
| 71 | # endif
|
|---|
| 72 | #else
|
|---|
| 73 | # define UTF16 "UTF-16"
|
|---|
| 74 | #endif
|
|---|
| 75 |
|
|---|
| 76 | #if defined(Q_OS_MAC)
|
|---|
| 77 | #ifndef GNU_LIBICONV
|
|---|
| 78 | #define GNU_LIBICONV
|
|---|
| 79 | #endif
|
|---|
| 80 | typedef iconv_t (*Ptr_iconv_open) (const char*, const char*);
|
|---|
| 81 | typedef size_t (*Ptr_iconv) (iconv_t, const char **, size_t *, char **, size_t *);
|
|---|
| 82 | typedef int (*Ptr_iconv_close) (iconv_t);
|
|---|
| 83 |
|
|---|
| 84 | static Ptr_iconv_open ptr_iconv_open = 0;
|
|---|
| 85 | static Ptr_iconv ptr_iconv = 0;
|
|---|
| 86 | static Ptr_iconv_close ptr_iconv_close = 0;
|
|---|
| 87 | #endif
|
|---|
| 88 |
|
|---|
| 89 | QT_BEGIN_NAMESPACE
|
|---|
| 90 |
|
|---|
| 91 | extern bool qt_locale_initialized;
|
|---|
| 92 |
|
|---|
| 93 | QIconvCodec::QIconvCodec()
|
|---|
| 94 | : utf16Codec(0)
|
|---|
| 95 | {
|
|---|
| 96 | utf16Codec = QTextCodec::codecForMib(1015);
|
|---|
| 97 | Q_ASSERT_X(utf16Codec != 0,
|
|---|
| 98 | "QIconvCodec::convertToUnicode",
|
|---|
| 99 | "internal error, UTF-16 codec not found");
|
|---|
| 100 | if (!utf16Codec) {
|
|---|
| 101 | fprintf(stderr, "QIconvCodec::convertToUnicode: internal error, UTF-16 codec not found\n");
|
|---|
| 102 | utf16Codec = reinterpret_cast<QTextCodec *>(~0);
|
|---|
| 103 | }
|
|---|
| 104 | #if defined(Q_OS_MAC)
|
|---|
| 105 | if (ptr_iconv_open == 0) {
|
|---|
| 106 | QLibrary libiconv(QLatin1String("/usr/lib/libiconv"));
|
|---|
| 107 | libiconv.setLoadHints(QLibrary::ExportExternalSymbolsHint);
|
|---|
| 108 |
|
|---|
| 109 | ptr_iconv_open = reinterpret_cast<Ptr_iconv_open>(libiconv.resolve("libiconv_open"));
|
|---|
| 110 | if (!ptr_iconv_open)
|
|---|
| 111 | ptr_iconv_open = reinterpret_cast<Ptr_iconv_open>(libiconv.resolve("iconv_open"));
|
|---|
| 112 | ptr_iconv = reinterpret_cast<Ptr_iconv>(libiconv.resolve("libiconv"));
|
|---|
| 113 | if (!ptr_iconv)
|
|---|
| 114 | ptr_iconv = reinterpret_cast<Ptr_iconv>(libiconv.resolve("iconv"));
|
|---|
| 115 | ptr_iconv_close = reinterpret_cast<Ptr_iconv_close>(libiconv.resolve("libiconv_close"));
|
|---|
| 116 | if (!ptr_iconv_close)
|
|---|
| 117 | ptr_iconv_close = reinterpret_cast<Ptr_iconv_close>(libiconv.resolve("iconv_close"));
|
|---|
| 118 |
|
|---|
| 119 | Q_ASSERT_X(ptr_iconv_open && ptr_iconv && ptr_iconv_close,
|
|---|
| 120 | "QIconvCodec::QIconvCodec()",
|
|---|
| 121 | "internal error, could not resolve the iconv functions");
|
|---|
| 122 |
|
|---|
| 123 | # undef iconv_open
|
|---|
| 124 | # define iconv_open ptr_iconv_open
|
|---|
| 125 | # undef iconv
|
|---|
| 126 | # define iconv ptr_iconv
|
|---|
| 127 | # undef iconv_close
|
|---|
| 128 | # define iconv_close ptr_iconv_close
|
|---|
| 129 | }
|
|---|
| 130 | #endif
|
|---|
| 131 | }
|
|---|
| 132 |
|
|---|
| 133 | QIconvCodec::~QIconvCodec()
|
|---|
| 134 | {
|
|---|
| 135 | }
|
|---|
| 136 |
|
|---|
| 137 | QIconvCodec::IconvState::IconvState(iconv_t x)
|
|---|
| 138 | : buffer(array), bufferLen(sizeof array), cd(x)
|
|---|
| 139 | {
|
|---|
| 140 | }
|
|---|
| 141 |
|
|---|
| 142 | QIconvCodec::IconvState::~IconvState()
|
|---|
| 143 | {
|
|---|
| 144 | if (cd != reinterpret_cast<iconv_t>(-1))
|
|---|
| 145 | iconv_close(cd);
|
|---|
| 146 | if (buffer != array)
|
|---|
| 147 | delete[] buffer;
|
|---|
| 148 | }
|
|---|
| 149 |
|
|---|
| 150 | void QIconvCodec::IconvState::saveChars(const char *c, int count)
|
|---|
| 151 | {
|
|---|
| 152 | if (count > bufferLen) {
|
|---|
| 153 | if (buffer != array)
|
|---|
| 154 | delete[] buffer;
|
|---|
| 155 | buffer = new char[bufferLen = count];
|
|---|
| 156 | }
|
|---|
| 157 |
|
|---|
| 158 | memcpy(buffer, c, count);
|
|---|
| 159 | }
|
|---|
| 160 |
|
|---|
| 161 | static void qIconvCodecStateFree(QTextCodec::ConverterState *state)
|
|---|
| 162 | {
|
|---|
| 163 | delete reinterpret_cast<QIconvCodec::IconvState *>(state->d);
|
|---|
| 164 | }
|
|---|
| 165 |
|
|---|
| 166 | Q_GLOBAL_STATIC(QThreadStorage<QIconvCodec::IconvState *>, toUnicodeState)
|
|---|
| 167 |
|
|---|
| 168 | QString QIconvCodec::convertToUnicode(const char* chars, int len, ConverterState *convState) const
|
|---|
| 169 | {
|
|---|
| 170 | if (utf16Codec == reinterpret_cast<QTextCodec *>(~0))
|
|---|
| 171 | return QString::fromLatin1(chars, len);
|
|---|
| 172 |
|
|---|
| 173 | int invalidCount = 0;
|
|---|
| 174 | int remainingCount = 0;
|
|---|
| 175 | char *remainingBuffer = 0;
|
|---|
| 176 | IconvState **pstate;
|
|---|
| 177 |
|
|---|
| 178 | if (convState) {
|
|---|
| 179 | // stateful conversion
|
|---|
| 180 | pstate = reinterpret_cast<IconvState **>(&convState->d);
|
|---|
| 181 | if (convState->d) {
|
|---|
| 182 | // restore state
|
|---|
| 183 | remainingCount = convState->remainingChars;
|
|---|
| 184 | remainingBuffer = (*pstate)->buffer;
|
|---|
| 185 | } else {
|
|---|
| 186 | // first time
|
|---|
| 187 | convState->flags |= FreeFunction;
|
|---|
| 188 | QTextCodecUnalignedPointer::encode(convState->state_data, qIconvCodecStateFree);
|
|---|
| 189 | }
|
|---|
| 190 | } else {
|
|---|
| 191 | QThreadStorage<QIconvCodec::IconvState *> *ts = toUnicodeState();
|
|---|
| 192 | if (!qt_locale_initialized || !ts) {
|
|---|
| 193 | // we're running after the Q_GLOBAL_STATIC has been deleted
|
|---|
| 194 | // or before the QCoreApplication initialization
|
|---|
| 195 | // bad programmer, no cookie for you
|
|---|
| 196 | return QString::fromLatin1(chars, len);
|
|---|
| 197 | }
|
|---|
| 198 |
|
|---|
| 199 | // stateless conversion -- use thread-local data
|
|---|
| 200 | pstate = &toUnicodeState()->localData();
|
|---|
| 201 | }
|
|---|
| 202 |
|
|---|
| 203 | if (!*pstate) {
|
|---|
| 204 | // first time, create the state
|
|---|
| 205 | iconv_t cd = QIconvCodec::createIconv_t(UTF16, 0);
|
|---|
| 206 | if (cd == reinterpret_cast<iconv_t>(-1)) {
|
|---|
| 207 | static int reported = 0;
|
|---|
| 208 | if (!reported++) {
|
|---|
| 209 | fprintf(stderr,
|
|---|
| 210 | "QIconvCodec::convertToUnicode: using Latin-1 for conversion, iconv_open failed\n");
|
|---|
| 211 | }
|
|---|
| 212 | return QString::fromLatin1(chars, len);
|
|---|
| 213 | }
|
|---|
| 214 |
|
|---|
| 215 | *pstate = new IconvState(cd);
|
|---|
| 216 | }
|
|---|
| 217 |
|
|---|
| 218 | IconvState *state = *pstate;
|
|---|
| 219 | size_t inBytesLeft = len;
|
|---|
| 220 | // best case assumption, each byte is converted into one UTF-16 character, plus 2 bytes for the BOM
|
|---|
| 221 | #ifdef GNU_LIBICONV
|
|---|
| 222 | // GNU doesn't disagree with POSIX :/
|
|---|
| 223 | const char *inBytes = chars;
|
|---|
| 224 | #else
|
|---|
| 225 | char *inBytes = const_cast<char *>(chars);
|
|---|
| 226 | #endif
|
|---|
| 227 |
|
|---|
| 228 | QByteArray in;
|
|---|
| 229 | if (remainingCount) {
|
|---|
| 230 | // we have to prepend the remaining bytes from the previous conversion
|
|---|
| 231 | inBytesLeft += remainingCount;
|
|---|
| 232 | in.resize(inBytesLeft);
|
|---|
| 233 | inBytes = in.data();
|
|---|
| 234 |
|
|---|
| 235 | memcpy(in.data(), remainingBuffer, remainingCount);
|
|---|
| 236 | memcpy(in.data() + remainingCount, chars, len);
|
|---|
| 237 |
|
|---|
| 238 | remainingCount = 0;
|
|---|
| 239 | }
|
|---|
| 240 |
|
|---|
| 241 | size_t outBytesLeft = len * 2 + 2;
|
|---|
| 242 | QByteArray ba(outBytesLeft, Qt::Uninitialized);
|
|---|
| 243 | char *outBytes = ba.data();
|
|---|
| 244 | do {
|
|---|
| 245 | size_t ret = iconv(state->cd, &inBytes, &inBytesLeft, &outBytes, &outBytesLeft);
|
|---|
| 246 | if (ret == (size_t) -1) {
|
|---|
| 247 | if (errno == E2BIG) {
|
|---|
| 248 | int offset = ba.size() - outBytesLeft;
|
|---|
| 249 | ba.resize(ba.size() * 2);
|
|---|
| 250 | outBytes = ba.data() + offset;
|
|---|
| 251 | outBytesLeft = ba.size() - offset;
|
|---|
| 252 |
|
|---|
| 253 | continue;
|
|---|
| 254 | }
|
|---|
| 255 |
|
|---|
| 256 | if (errno == EILSEQ) {
|
|---|
| 257 | // conversion stopped because of an invalid character in the sequence
|
|---|
| 258 | ++invalidCount;
|
|---|
| 259 | } else if (errno == EINVAL && convState) {
|
|---|
| 260 | // conversion stopped because the remaining inBytesLeft make up
|
|---|
| 261 | // an incomplete multi-byte sequence; save them for later
|
|---|
| 262 | state->saveChars(inBytes, inBytesLeft);
|
|---|
| 263 | remainingCount = inBytesLeft;
|
|---|
| 264 | break;
|
|---|
| 265 | }
|
|---|
| 266 |
|
|---|
| 267 | if (errno == EILSEQ || errno == EINVAL) {
|
|---|
| 268 | // skip the next character
|
|---|
| 269 | ++inBytes;
|
|---|
| 270 | --inBytesLeft;
|
|---|
| 271 | continue;
|
|---|
| 272 | }
|
|---|
| 273 |
|
|---|
| 274 | // some other error
|
|---|
| 275 | // note, cannot use qWarning() since we are implementing the codecForLocale :)
|
|---|
| 276 | perror("QIconvCodec::convertToUnicode: using Latin-1 for conversion, iconv failed");
|
|---|
| 277 |
|
|---|
| 278 | if (!convState) {
|
|---|
| 279 | // reset state
|
|---|
| 280 | iconv(state->cd, 0, &inBytesLeft, 0, &outBytesLeft);
|
|---|
| 281 | }
|
|---|
| 282 |
|
|---|
| 283 | return QString::fromLatin1(chars, len);
|
|---|
| 284 | }
|
|---|
| 285 | } while (inBytesLeft != 0);
|
|---|
| 286 |
|
|---|
| 287 | QString s = utf16Codec->toUnicode(ba.constData(), ba.size() - outBytesLeft);
|
|---|
| 288 |
|
|---|
| 289 | if (convState) {
|
|---|
| 290 | convState->invalidChars = invalidCount;
|
|---|
| 291 | convState->remainingChars = remainingCount;
|
|---|
| 292 | } else {
|
|---|
| 293 | // reset state
|
|---|
| 294 | iconv(state->cd, 0, &inBytesLeft, 0, &outBytesLeft);
|
|---|
| 295 | }
|
|---|
| 296 |
|
|---|
| 297 | return s;
|
|---|
| 298 | }
|
|---|
| 299 |
|
|---|
| 300 | Q_GLOBAL_STATIC(QThreadStorage<QIconvCodec::IconvState *>, fromUnicodeState)
|
|---|
| 301 |
|
|---|
| 302 | static bool setByteOrder(iconv_t cd)
|
|---|
| 303 | {
|
|---|
| 304 | #if !defined(NO_BOM)
|
|---|
| 305 | // give iconv() a BOM
|
|---|
| 306 | char buf[4];
|
|---|
| 307 | ushort bom[] = { QChar::ByteOrderMark };
|
|---|
| 308 |
|
|---|
| 309 | char *outBytes = buf;
|
|---|
| 310 | char *inBytes = reinterpret_cast<char *>(bom);
|
|---|
| 311 | size_t outBytesLeft = sizeof buf;
|
|---|
| 312 | size_t inBytesLeft = sizeof bom;
|
|---|
| 313 |
|
|---|
| 314 | #if defined(GNU_LIBICONV)
|
|---|
| 315 | const char **inBytesPtr = const_cast<const char **>(&inBytes);
|
|---|
| 316 | #else
|
|---|
| 317 | char **inBytesPtr = &inBytes;
|
|---|
| 318 | #endif
|
|---|
| 319 |
|
|---|
| 320 | if (iconv(cd, inBytesPtr, &inBytesLeft, &outBytes, &outBytesLeft) == (size_t) -1) {
|
|---|
| 321 | return false;
|
|---|
| 322 | }
|
|---|
| 323 | #endif // NO_BOM
|
|---|
| 324 |
|
|---|
| 325 | return true;
|
|---|
| 326 | }
|
|---|
| 327 |
|
|---|
| 328 | QByteArray QIconvCodec::convertFromUnicode(const QChar *uc, int len, ConverterState *convState) const
|
|---|
| 329 | {
|
|---|
| 330 | char *inBytes;
|
|---|
| 331 | char *outBytes;
|
|---|
| 332 | size_t inBytesLeft;
|
|---|
| 333 |
|
|---|
| 334 | #if defined(GNU_LIBICONV)
|
|---|
| 335 | const char **inBytesPtr = const_cast<const char **>(&inBytes);
|
|---|
| 336 | #else
|
|---|
| 337 | char **inBytesPtr = &inBytes;
|
|---|
| 338 | #endif
|
|---|
| 339 |
|
|---|
| 340 | QThreadStorage<QIconvCodec::IconvState *> *ts = fromUnicodeState();
|
|---|
| 341 | if (!qt_locale_initialized || !ts) {
|
|---|
| 342 | // we're running after the Q_GLOBAL_STATIC has been deleted
|
|---|
| 343 | // or before the QCoreApplication initialization
|
|---|
| 344 | // bad programmer, no cookie for you
|
|---|
| 345 | if (!len)
|
|---|
| 346 | // this is a special case - zero-sized string should be
|
|---|
| 347 | // translated to empty but not-null QByteArray.
|
|---|
| 348 | return QByteArray("");
|
|---|
| 349 | return QString::fromRawData(uc, len).toLatin1();
|
|---|
| 350 | }
|
|---|
| 351 | IconvState *&state = ts->localData();
|
|---|
| 352 | if (!state) {
|
|---|
| 353 | state = new IconvState(QIconvCodec::createIconv_t(0, UTF16));
|
|---|
| 354 | if (state->cd == reinterpret_cast<iconv_t>(-1)) {
|
|---|
| 355 | if (!setByteOrder(state->cd)) {
|
|---|
| 356 | perror("QIconvCodec::convertFromUnicode: using Latin-1 for conversion, iconv failed for BOM");
|
|---|
| 357 |
|
|---|
| 358 | iconv_close(state->cd);
|
|---|
| 359 | state->cd = reinterpret_cast<iconv_t>(-1);
|
|---|
| 360 |
|
|---|
| 361 | return QString(uc, len).toLatin1();
|
|---|
| 362 | }
|
|---|
| 363 | }
|
|---|
| 364 | }
|
|---|
| 365 | if (state->cd == reinterpret_cast<iconv_t>(-1)) {
|
|---|
| 366 | static int reported = 0;
|
|---|
| 367 | if (!reported++) {
|
|---|
| 368 | fprintf(stderr,
|
|---|
| 369 | "QIconvCodec::convertFromUnicode: using Latin-1 for conversion, iconv_open failed\n");
|
|---|
| 370 | }
|
|---|
| 371 | return QString(uc, len).toLatin1();
|
|---|
| 372 | }
|
|---|
| 373 |
|
|---|
| 374 | size_t outBytesLeft = len;
|
|---|
| 375 | QByteArray ba(outBytesLeft, Qt::Uninitialized);
|
|---|
| 376 | outBytes = ba.data();
|
|---|
| 377 |
|
|---|
| 378 | // now feed iconv() the real data
|
|---|
| 379 | inBytes = const_cast<char *>(reinterpret_cast<const char *>(uc));
|
|---|
| 380 | inBytesLeft = len * sizeof(QChar);
|
|---|
| 381 |
|
|---|
| 382 | QByteArray in;
|
|---|
| 383 | if (convState && convState->remainingChars) {
|
|---|
| 384 | // we have one surrogate char to be prepended
|
|---|
| 385 | in.resize(sizeof(QChar) + len);
|
|---|
| 386 | inBytes = in.data();
|
|---|
| 387 |
|
|---|
| 388 | QChar remaining = convState->state_data[0];
|
|---|
| 389 | memcpy(in.data(), &remaining, sizeof(QChar));
|
|---|
| 390 | memcpy(in.data() + sizeof(QChar), uc, inBytesLeft);
|
|---|
| 391 |
|
|---|
| 392 | inBytesLeft += sizeof(QChar);
|
|---|
| 393 | convState->remainingChars = 0;
|
|---|
| 394 | }
|
|---|
| 395 |
|
|---|
| 396 | int invalidCount = 0;
|
|---|
| 397 | while (inBytesLeft != 0) {
|
|---|
| 398 | if (iconv(state->cd, inBytesPtr, &inBytesLeft, &outBytes, &outBytesLeft) == (size_t) -1) {
|
|---|
| 399 | if (errno == EINVAL && convState) {
|
|---|
| 400 | // buffer ends in a surrogate
|
|---|
| 401 | Q_ASSERT(inBytesLeft == 2);
|
|---|
| 402 | convState->remainingChars = 1;
|
|---|
| 403 | convState->state_data[0] = uc[len - 1].unicode();
|
|---|
| 404 | break;
|
|---|
| 405 | }
|
|---|
| 406 |
|
|---|
| 407 | switch (errno) {
|
|---|
| 408 | case EILSEQ:
|
|---|
| 409 | ++invalidCount;
|
|---|
| 410 | // fall through
|
|---|
| 411 | case EINVAL:
|
|---|
| 412 | {
|
|---|
| 413 | inBytes += sizeof(QChar);
|
|---|
| 414 | inBytesLeft -= sizeof(QChar);
|
|---|
| 415 | break;
|
|---|
| 416 | }
|
|---|
| 417 | case E2BIG:
|
|---|
| 418 | {
|
|---|
| 419 | int offset = ba.size() - outBytesLeft;
|
|---|
| 420 | ba.resize(ba.size() * 2);
|
|---|
| 421 | outBytes = ba.data() + offset;
|
|---|
| 422 | outBytesLeft = ba.size() - offset;
|
|---|
| 423 | break;
|
|---|
| 424 | }
|
|---|
| 425 | default:
|
|---|
| 426 | {
|
|---|
| 427 | // note, cannot use qWarning() since we are implementing the codecForLocale :)
|
|---|
| 428 | perror("QIconvCodec::convertFromUnicode: using Latin-1 for conversion, iconv failed");
|
|---|
| 429 |
|
|---|
| 430 | // reset to initial state
|
|---|
| 431 | iconv(state->cd, 0, &inBytesLeft, 0, &outBytesLeft);
|
|---|
| 432 |
|
|---|
| 433 | return QString(uc, len).toLatin1();
|
|---|
| 434 | }
|
|---|
| 435 | }
|
|---|
| 436 | }
|
|---|
| 437 | }
|
|---|
| 438 |
|
|---|
| 439 | // reset to initial state
|
|---|
| 440 | iconv(state->cd, 0, &inBytesLeft, 0, &outBytesLeft);
|
|---|
| 441 | setByteOrder(state->cd);
|
|---|
| 442 |
|
|---|
| 443 | ba.resize(ba.size() - outBytesLeft);
|
|---|
| 444 |
|
|---|
| 445 | if (convState)
|
|---|
| 446 | convState->invalidChars = invalidCount;
|
|---|
| 447 |
|
|---|
| 448 | return ba;
|
|---|
| 449 | }
|
|---|
| 450 |
|
|---|
| 451 | QByteArray QIconvCodec::name() const
|
|---|
| 452 | {
|
|---|
| 453 | return "System";
|
|---|
| 454 | }
|
|---|
| 455 |
|
|---|
| 456 | int QIconvCodec::mibEnum() const
|
|---|
| 457 | {
|
|---|
| 458 | return 0;
|
|---|
| 459 | }
|
|---|
| 460 |
|
|---|
| 461 | iconv_t QIconvCodec::createIconv_t(const char *to, const char *from)
|
|---|
| 462 | {
|
|---|
| 463 | Q_ASSERT((to == 0 && from != 0) || (to != 0 && from == 0));
|
|---|
| 464 |
|
|---|
| 465 | iconv_t cd = (iconv_t) -1;
|
|---|
| 466 | #if defined(__GLIBC__) || defined(GNU_LIBICONV)
|
|---|
| 467 | // both GLIBC and libgnuiconv will use the locale's encoding if from or to is an empty string
|
|---|
| 468 | static const char empty_codeset[] = "";
|
|---|
| 469 | const char *codeset = empty_codeset;
|
|---|
| 470 | cd = iconv_open(to ? to : codeset, from ? from : codeset);
|
|---|
| 471 | #else
|
|---|
| 472 | char *codeset = 0;
|
|---|
| 473 | #endif
|
|---|
| 474 |
|
|---|
| 475 | #if defined(_XOPEN_UNIX) && !defined(Q_OS_QNX) && !defined(Q_OS_OSF)
|
|---|
| 476 | if (cd == (iconv_t) -1) {
|
|---|
| 477 | codeset = nl_langinfo(CODESET);
|
|---|
| 478 | if (codeset)
|
|---|
| 479 | cd = iconv_open(to ? to : codeset, from ? from : codeset);
|
|---|
| 480 | }
|
|---|
| 481 | #endif
|
|---|
| 482 |
|
|---|
| 483 | if (cd == (iconv_t) -1) {
|
|---|
| 484 | // Very poorly defined and followed standards causes lots of
|
|---|
| 485 | // code to try to get all the cases... This logic is
|
|---|
| 486 | // duplicated in QTextCodec, so if you change it here, change
|
|---|
| 487 | // it there too.
|
|---|
| 488 |
|
|---|
| 489 | // Try to determine locale codeset from locale name assigned to
|
|---|
| 490 | // LC_CTYPE category.
|
|---|
| 491 |
|
|---|
| 492 | // First part is getting that locale name. First try setlocale() which
|
|---|
| 493 | // definitely knows it, but since we cannot fully trust it, get ready
|
|---|
| 494 | // to fall back to environment variables.
|
|---|
| 495 | char * ctype = qstrdup(setlocale(LC_CTYPE, 0));
|
|---|
| 496 |
|
|---|
| 497 | // Get the first nonempty value from $LC_ALL, $LC_CTYPE, and $LANG
|
|---|
| 498 | // environment variables.
|
|---|
| 499 | char * lang = qstrdup(qgetenv("LC_ALL").constData());
|
|---|
| 500 | if (!lang || lang[0] == 0 || strcmp(lang, "C") == 0) {
|
|---|
| 501 | if (lang) delete [] lang;
|
|---|
| 502 | lang = qstrdup(qgetenv("LC_CTYPE").constData());
|
|---|
| 503 | }
|
|---|
| 504 | if (!lang || lang[0] == 0 || strcmp(lang, "C") == 0) {
|
|---|
| 505 | if (lang) delete [] lang;
|
|---|
| 506 | lang = qstrdup(qgetenv("LANG").constData());
|
|---|
| 507 | }
|
|---|
| 508 |
|
|---|
| 509 | // Now try these in order:
|
|---|
| 510 | // 1. CODESET from ctype if it contains a .CODESET part (e.g. en_US.ISO8859-15)
|
|---|
| 511 | // 2. CODESET from lang if it contains a .CODESET part
|
|---|
| 512 | // 3. ctype (maybe the locale is named "ISO-8859-1" or something)
|
|---|
| 513 | // 4. locale (ditto)
|
|---|
| 514 | // 5. check for "@euro"
|
|---|
| 515 |
|
|---|
| 516 | // 1. CODESET from ctype if it contains a .CODESET part (e.g. en_US.ISO8859-15)
|
|---|
| 517 | codeset = ctype ? strchr(ctype, '.') : 0;
|
|---|
| 518 | if (codeset && *codeset == '.') {
|
|---|
| 519 | ++codeset;
|
|---|
| 520 | cd = iconv_open(to ? to : codeset, from ? from : codeset);
|
|---|
| 521 | }
|
|---|
| 522 |
|
|---|
| 523 | // 2. CODESET from lang if it contains a .CODESET part
|
|---|
| 524 | codeset = lang ? strchr(lang, '.') : 0;
|
|---|
| 525 | if (cd == (iconv_t) -1 && codeset && *codeset == '.') {
|
|---|
| 526 | ++codeset;
|
|---|
| 527 | cd = iconv_open(to ? to : codeset, from ? from : codeset);
|
|---|
| 528 | }
|
|---|
| 529 |
|
|---|
| 530 | // 3. ctype (maybe the locale is named "ISO-8859-1" or something)
|
|---|
| 531 | if (cd == (iconv_t) -1 && ctype && *ctype != 0 && strcmp (ctype, "C") != 0)
|
|---|
| 532 | cd = iconv_open(to ? to : ctype, from ? from : ctype);
|
|---|
| 533 |
|
|---|
| 534 |
|
|---|
| 535 | // 4. locale (ditto)
|
|---|
| 536 | if (cd == (iconv_t) -1 && lang && *lang != 0)
|
|---|
| 537 | cd = iconv_open(to ? to : lang, from ? from : lang);
|
|---|
| 538 |
|
|---|
| 539 | // 5. "@euro"
|
|---|
| 540 | if ((cd == (iconv_t) -1 && ctype && strstr(ctype, "@euro")) || (lang && strstr(lang, "@euro")))
|
|---|
| 541 | cd = iconv_open(to ? to : "ISO8859-15", from ? from : "ISO8859-15");
|
|---|
| 542 |
|
|---|
| 543 | delete [] ctype;
|
|---|
| 544 | delete [] lang;
|
|---|
| 545 | }
|
|---|
| 546 |
|
|---|
| 547 | return cd;
|
|---|
| 548 | }
|
|---|
| 549 |
|
|---|
| 550 | QT_END_NAMESPACE
|
|---|