| 1 | /* $Id: utf8-posix.cpp 26344 2010-02-09 03:39:45Z vboxsync $ */
|
|---|
| 2 | /** @file
|
|---|
| 3 | * IPRT - UTF-8 helpers, POSIX.
|
|---|
| 4 | */
|
|---|
| 5 |
|
|---|
| 6 | /*
|
|---|
| 7 | * Copyright (C) 2006-2007 Sun Microsystems, Inc.
|
|---|
| 8 | *
|
|---|
| 9 | * This file is part of VirtualBox Open Source Edition (OSE), as
|
|---|
| 10 | * available from http://www.virtualbox.org. This file is free software;
|
|---|
| 11 | * you can redistribute it and/or modify it under the terms of the GNU
|
|---|
| 12 | * General Public License (GPL) as published by the Free Software
|
|---|
| 13 | * Foundation, in version 2 as it comes in the "COPYING" file of the
|
|---|
| 14 | * VirtualBox OSE distribution. VirtualBox OSE is distributed in the
|
|---|
| 15 | * hope that it will be useful, but WITHOUT ANY WARRANTY of any kind.
|
|---|
| 16 | *
|
|---|
| 17 | * The contents of this file may alternatively be used under the terms
|
|---|
| 18 | * of the Common Development and Distribution License Version 1.0
|
|---|
| 19 | * (CDDL) only, as it comes in the "COPYING.CDDL" file of the
|
|---|
| 20 | * VirtualBox OSE distribution, in which case the provisions of the
|
|---|
| 21 | * CDDL are applicable instead of those of the GPL.
|
|---|
| 22 | *
|
|---|
| 23 | * You may elect to license modified versions of this file under the
|
|---|
| 24 | * terms and conditions of either the GPL or the CDDL or both.
|
|---|
| 25 | *
|
|---|
| 26 | * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa
|
|---|
| 27 | * Clara, CA 95054 USA or visit http://www.sun.com if you need
|
|---|
| 28 | * additional information or have any questions.
|
|---|
| 29 | */
|
|---|
| 30 |
|
|---|
| 31 |
|
|---|
| 32 | /*******************************************************************************
|
|---|
| 33 | * Header Files *
|
|---|
| 34 | *******************************************************************************/
|
|---|
| 35 | #include <iprt/string.h>
|
|---|
| 36 | #include <iprt/alloc.h>
|
|---|
| 37 | #include <iprt/assert.h>
|
|---|
| 38 | #include <iprt/err.h>
|
|---|
| 39 | #include <iprt/string.h>
|
|---|
| 40 |
|
|---|
| 41 | #include <errno.h>
|
|---|
| 42 | #include <locale.h>
|
|---|
| 43 | #include <iconv.h>
|
|---|
| 44 | #include <wctype.h>
|
|---|
| 45 |
|
|---|
| 46 | #ifdef RT_OS_SOLARIS
|
|---|
| 47 | # include <langinfo.h>
|
|---|
| 48 | #endif
|
|---|
| 49 |
|
|---|
| 50 | #include "internal/alignmentchecks.h"
|
|---|
| 51 |
|
|---|
| 52 |
|
|---|
| 53 | /*******************************************************************************
|
|---|
| 54 | * Internal Functions *
|
|---|
| 55 | *******************************************************************************/
|
|---|
| 56 | static int rtstrConvert(const void *pvInput, size_t cbInput, const char *pszInputCS, void **ppvOutput, size_t cbOutput, const char *pszOutputCS, unsigned cFactor);
|
|---|
| 57 |
|
|---|
| 58 |
|
|---|
| 59 | /**
|
|---|
| 60 | * Converts a string from one charset to another.
|
|---|
| 61 | *
|
|---|
| 62 | * @returns iprt status code.
|
|---|
| 63 | * @param pvInput Pointer to intput string.
|
|---|
| 64 | * @param cbInput Size (in bytes) of input string. Excludes any terminators.
|
|---|
| 65 | * @param pszInputCS Codeset of the input string.
|
|---|
| 66 | * @param ppvOutput Pointer to pointer to output buffer if cbOutput > 0.
|
|---|
| 67 | * If cbOutput is 0 this is where the pointer to the allocated
|
|---|
| 68 | * buffer is stored.
|
|---|
| 69 | * @param cbOutput Size of the passed in buffer.
|
|---|
| 70 | * @param pszOutputCS Codeset of the input string.
|
|---|
| 71 | * @param cFactor Input vs. output size factor.
|
|---|
| 72 | */
|
|---|
| 73 | static int rtstrConvert(const void *pvInput, size_t cbInput, const char *pszInputCS, void **ppvOutput, size_t cbOutput, const char *pszOutputCS, unsigned cFactor)
|
|---|
| 74 | {
|
|---|
| 75 | /*
|
|---|
| 76 | * Allocate buffer
|
|---|
| 77 | */
|
|---|
| 78 | bool fUcs2Term;
|
|---|
| 79 | void *pvOutput;
|
|---|
| 80 | size_t cbOutput2;
|
|---|
| 81 | if (!cbOutput)
|
|---|
| 82 | {
|
|---|
| 83 | cbOutput2 = cbInput * cFactor;
|
|---|
| 84 | pvOutput = RTMemTmpAlloc(cbOutput2 + sizeof(RTUTF16));
|
|---|
| 85 | if (!pvOutput)
|
|---|
| 86 | return VERR_NO_TMP_MEMORY;
|
|---|
| 87 | fUcs2Term = true;
|
|---|
| 88 | }
|
|---|
| 89 | else
|
|---|
| 90 | {
|
|---|
| 91 | pvOutput = *ppvOutput;
|
|---|
| 92 | fUcs2Term = !strcmp(pszOutputCS, "UCS-2");
|
|---|
| 93 | cbOutput2 = cbOutput - (fUcs2Term ? sizeof(RTUTF16) : 1);
|
|---|
| 94 | if (cbOutput2 > cbOutput)
|
|---|
| 95 | return VERR_BUFFER_OVERFLOW;
|
|---|
| 96 | }
|
|---|
| 97 |
|
|---|
| 98 | /*
|
|---|
| 99 | * Use a loop here to retry with bigger buffers.
|
|---|
| 100 | */
|
|---|
| 101 | for (unsigned cTries = 10; cTries > 0; cTries--)
|
|---|
| 102 | {
|
|---|
| 103 | /*
|
|---|
| 104 | * Create conversion object.
|
|---|
| 105 | */
|
|---|
| 106 | #ifdef RT_OS_SOLARIS
|
|---|
| 107 | /* Solaris doesn't grok empty codeset strings, so help it find the current codeset. */
|
|---|
| 108 | if (!*pszInputCS)
|
|---|
| 109 | pszInputCS = nl_langinfo(CODESET);
|
|---|
| 110 | if (!*pszOutputCS)
|
|---|
| 111 | pszOutputCS = nl_langinfo(CODESET);
|
|---|
| 112 | #endif
|
|---|
| 113 | IPRT_ALIGNMENT_CHECKS_DISABLE(); /* glibc causes trouble */
|
|---|
| 114 | iconv_t icHandle = iconv_open(pszOutputCS, pszInputCS);
|
|---|
| 115 | IPRT_ALIGNMENT_CHECKS_ENABLE();
|
|---|
| 116 | if (icHandle != (iconv_t)-1)
|
|---|
| 117 | {
|
|---|
| 118 | /*
|
|---|
| 119 | * Do the conversion.
|
|---|
| 120 | */
|
|---|
| 121 | size_t cbInLeft = cbInput;
|
|---|
| 122 | size_t cbOutLeft = cbOutput2;
|
|---|
| 123 | const void *pvInputLeft = pvInput;
|
|---|
| 124 | void *pvOutputLeft = pvOutput;
|
|---|
| 125 | #if defined(RT_OS_LINUX) || (defined(RT_OS_DARWIN) && defined(_DARWIN_FEATURE_UNIX_CONFORMANCE)) /* there are different opinions about the constness of the input buffer. */
|
|---|
| 126 | if (iconv(icHandle, (char **)&pvInputLeft, &cbInLeft, (char **)&pvOutputLeft, &cbOutLeft) != (size_t)-1)
|
|---|
| 127 | #else
|
|---|
| 128 | if (iconv(icHandle, (const char **)&pvInputLeft, &cbInLeft, (char **)&pvOutputLeft, &cbOutLeft) != (size_t)-1)
|
|---|
| 129 | #endif
|
|---|
| 130 | {
|
|---|
| 131 | if (!cbInLeft)
|
|---|
| 132 | {
|
|---|
| 133 | /*
|
|---|
| 134 | * We're done, just add the terminator and return.
|
|---|
| 135 | * (Two terminators to support UCS-2 output, too.)
|
|---|
| 136 | */
|
|---|
| 137 | iconv_close(icHandle);
|
|---|
| 138 | ((char *)pvOutputLeft)[0] = '\0';
|
|---|
| 139 | if (fUcs2Term)
|
|---|
| 140 | ((char *)pvOutputLeft)[1] = '\0';
|
|---|
| 141 | *ppvOutput = pvOutput;
|
|---|
| 142 | return VINF_SUCCESS;
|
|---|
| 143 | }
|
|---|
| 144 | errno = E2BIG;
|
|---|
| 145 | }
|
|---|
| 146 | iconv_close(icHandle);
|
|---|
| 147 |
|
|---|
| 148 | /*
|
|---|
| 149 | * If we failed because of output buffer space we'll
|
|---|
| 150 | * increase the output buffer size and retry.
|
|---|
| 151 | */
|
|---|
| 152 | if (errno == E2BIG)
|
|---|
| 153 | {
|
|---|
| 154 | if (!cbOutput)
|
|---|
| 155 | {
|
|---|
| 156 | RTMemTmpFree(pvOutput);
|
|---|
| 157 | cbOutput2 *= 2;
|
|---|
| 158 | pvOutput = RTMemTmpAlloc(cbOutput2 + sizeof(RTUTF16));
|
|---|
| 159 | if (!pvOutput)
|
|---|
| 160 | return VERR_NO_TMP_MEMORY;
|
|---|
| 161 | continue;
|
|---|
| 162 | }
|
|---|
| 163 | return VERR_BUFFER_OVERFLOW;
|
|---|
| 164 | }
|
|---|
| 165 | }
|
|---|
| 166 | break;
|
|---|
| 167 | }
|
|---|
| 168 |
|
|---|
| 169 | /* failure */
|
|---|
| 170 | if (!cbOutput)
|
|---|
| 171 | RTMemTmpFree(pvOutput);
|
|---|
| 172 | return VERR_NO_TRANSLATION;
|
|---|
| 173 | }
|
|---|
| 174 |
|
|---|
| 175 |
|
|---|
| 176 | /**
|
|---|
| 177 | * Allocates tmp buffer, translates pszString from UTF8 to current codepage.
|
|---|
| 178 | *
|
|---|
| 179 | * @returns iprt status code.
|
|---|
| 180 | * @param ppszString Receives pointer of allocated native CP string.
|
|---|
| 181 | * The returned pointer must be freed using RTStrFree().
|
|---|
| 182 | * @param pszString UTF-8 string to convert.
|
|---|
| 183 | */
|
|---|
| 184 | RTR3DECL(int) RTStrUtf8ToCurrentCP(char **ppszString, const char *pszString)
|
|---|
| 185 | {
|
|---|
| 186 | Assert(ppszString);
|
|---|
| 187 | Assert(pszString);
|
|---|
| 188 | *ppszString = NULL;
|
|---|
| 189 |
|
|---|
| 190 | /*
|
|---|
| 191 | * Assume result string length is not longer than UTF-8 string.
|
|---|
| 192 | */
|
|---|
| 193 | size_t cch = strlen(pszString);
|
|---|
| 194 | if (cch <= 0)
|
|---|
| 195 | {
|
|---|
| 196 | /* zero length string passed. */
|
|---|
| 197 | *ppszString = (char *)RTMemTmpAllocZ(sizeof(char));
|
|---|
| 198 | if (*ppszString)
|
|---|
| 199 | return VINF_SUCCESS;
|
|---|
| 200 | return VERR_NO_TMP_MEMORY;
|
|---|
| 201 | }
|
|---|
| 202 | return rtstrConvert(pszString, cch, "UTF-8", (void **)ppszString, 0, "", 1);
|
|---|
| 203 | }
|
|---|
| 204 |
|
|---|
| 205 |
|
|---|
| 206 | /**
|
|---|
| 207 | * Allocates tmp buffer, translates pszString from current codepage to UTF-8.
|
|---|
| 208 | *
|
|---|
| 209 | * @returns iprt status code.
|
|---|
| 210 | * @param ppszString Receives pointer of allocated UTF-8 string.
|
|---|
| 211 | * The returned pointer must be freed using RTStrFree().
|
|---|
| 212 | * @param pszString Native string to convert.
|
|---|
| 213 | */
|
|---|
| 214 | RTR3DECL(int) RTStrCurrentCPToUtf8(char **ppszString, const char *pszString)
|
|---|
| 215 | {
|
|---|
| 216 | Assert(ppszString);
|
|---|
| 217 | Assert(pszString);
|
|---|
| 218 | *ppszString = NULL;
|
|---|
| 219 |
|
|---|
| 220 | /*
|
|---|
| 221 | * Attempt with UTF-8 length of 2x the native lenght.
|
|---|
| 222 | */
|
|---|
| 223 | size_t cch = strlen(pszString);
|
|---|
| 224 | if (cch <= 0)
|
|---|
| 225 | {
|
|---|
| 226 | /* zero length string passed. */
|
|---|
| 227 | *ppszString = (char *)RTMemTmpAllocZ(sizeof(char));
|
|---|
| 228 | if (*ppszString)
|
|---|
| 229 | return VINF_SUCCESS;
|
|---|
| 230 | return VERR_NO_TMP_MEMORY;
|
|---|
| 231 | }
|
|---|
| 232 | return rtstrConvert(pszString, cch, "", (void **)ppszString, 0, "UTF-8", 2);
|
|---|
| 233 | }
|
|---|
| 234 |
|
|---|
