/*
* Copyright (C) 2012, 2013 Eric Biggers
*
- * This file is part of wimlib, a library for working with WIM files.
+ * This file is free software; you can redistribute it and/or modify it under
+ * the terms of the GNU Lesser General Public License as published by the Free
+ * Software Foundation; either version 3 of the License, or (at your option) any
+ * later version.
*
- * wimlib is free software; you can redistribute it and/or modify it under the
- * terms of the GNU General Public License as published by the Free
- * Software Foundation; either version 3 of the License, or (at your option)
- * any later version.
- *
- * wimlib is distributed in the hope that it will be useful, but WITHOUT ANY
- * WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
- * A PARTICULAR PURPOSE. See the GNU General Public License for more
+ * This file is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
+ * FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
* details.
*
- * You should have received a copy of the GNU General Public License
- * along with wimlib; if not, see http://www.gnu.org/licenses/.
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this file; if not, see http://www.gnu.org/licenses/.
*/
#ifdef HAVE_CONFIG_H
# include "config.h"
#endif
+#include <errno.h>
+#include <iconv.h>
+#include <pthread.h>
+#include <string.h>
+
#include "wimlib.h"
+#include "wimlib/alloca.h"
#include "wimlib/assert.h"
#include "wimlib/encoding.h"
#include "wimlib/endianness.h"
#include "wimlib/list.h"
#include "wimlib/util.h"
-#include <errno.h>
-#include <iconv.h>
-#include <pthread.h>
-#include <stdlib.h>
-#include <string.h>
bool wimlib_mbs_is_utf8 = !TCHAR_IS_UTF16LE;
struct iconv_list_head name = { \
.from_encoding = from, \
.to_encoding = to, \
- .list = LIST_HEAD_INIT(name.list), \
- .mutex = PTHREAD_MUTEX_INITIALIZER, \
}
static iconv_t *
errno = errno_save;
}
-/* Prevent printing an error message if a character conversion error occurs
- * while printing an error message. (This variable is not per-thread but it
- * doesn't matter too much since it's just the error messages.) */
-static bool error_message_being_printed = false;
-
#define DEFINE_CHAR_CONVERSION_FUNCTIONS(varname1, longname1, chartype1,\
varname2, longname2, chartype2,\
earlyreturn_on_utf8_locale, \
if (cd == NULL) \
return WIMLIB_ERR_ICONV_NOT_AVAILABLE; \
\
+ chartype2 *buf; \
+ size_t bufsize; \
+ bool buf_onheap; \
+ bufsize = (worst_case_len_expr) * sizeof(chartype2); \
/* Worst case length */ \
- chartype2 buf[worst_case_len_expr]; \
+ if (bufsize <= STACK_MAX) { \
+ buf = alloca(bufsize); \
+ buf_onheap = false; \
+ } else { \
+ buf = MALLOC(bufsize); \
+ if (!buf) \
+ return WIMLIB_ERR_NOMEM; \
+ buf_onheap = true; \
+ } \
+ \
char *inbuf = (char*)in; \
size_t inbytesleft = in_nbytes; \
char *outbuf = (char*)buf; \
- size_t outbytesleft = sizeof(buf); \
+ size_t outbytesleft = bufsize; \
size_t len; \
int ret; \
\
len = iconv(*cd, &inbuf, &inbytesleft, &outbuf, &outbytesleft); \
if (len == (size_t)-1) { \
- if (!error_message_being_printed) { \
- error_message_being_printed = true; \
- err_msg; \
- error_message_being_printed = false; \
- } \
+ err_msg; \
ret = err_return; \
} else { \
- *out_nbytes_ret = sizeof(buf) - outbytesleft; \
+ *out_nbytes_ret = bufsize - outbytesleft; \
ret = 0; \
} \
put_iconv(cd); \
+ if (buf_onheap) \
+ FREE(buf); \
return ret; \
} \
\
\
len = iconv(*cd, &inbuf, &inbytesleft, &outbuf, &outbytesleft); \
if (len == (size_t)-1) { \
- if (!error_message_being_printed) { \
- error_message_being_printed = true; \
- err_msg; \
- error_message_being_printed = false; \
- } \
+ err_msg; \
ret = err_return; \
} else { \
out[(LARGE_NUMBER-outbytesleft)/sizeof(chartype2)] = 0; \
return utf8_to_tstr(utf8str, strlen(utf8str), out, &out_nbytes);
}
+static void
+iconv_init(struct iconv_list_head *head)
+{
+ pthread_mutex_init(&head->mutex, NULL);
+ INIT_LIST_HEAD(&head->list);
+}
+
static void
iconv_cleanup(struct iconv_list_head *head)
{
}
}
+void
+iconv_global_init(void)
+{
+ iconv_init(&iconv_utf8_to_tstr);
+ iconv_init(&iconv_tstr_to_utf8);
+#if !TCHAR_IS_UTF16LE
+ iconv_init(&iconv_utf16le_to_tstr);
+ iconv_init(&iconv_tstr_to_utf16le);
+ iconv_init(&iconv_utf16le_to_utf8);
+ iconv_init(&iconv_utf8_to_utf16le);
+#endif
+}
+
void
iconv_global_cleanup(void)
{
* Copyright (c) 2008-2011 Jean-Pierre Andre
* Copyright (c) 2008 Bernhard Kaindl
*
- * License GPLv2 or later. (Or may actually be LGPL since it's in libntfs-3g.)
+ * License GPLv2 or later.
*
* The expanded table itself is the same as that used by Windows Vista for
* NTFS case insensitivity.
return (n1 < n2) ? -1 : 1;
}
-/* Duplicates a string of system-dependent encoding into a UTF-16LE string and
- * returns the string and its length, in bytes, in the pointer arguments. Frees
- * any existing string at the return location before overwriting it. */
+/* Like cmp_utf16le_strings(), but assumes the strings are null terminated. */
int
-get_utf16le_string(const tchar *name, utf16lechar **name_utf16le_ret,
- u16 *name_utf16le_nbytes_ret)
+cmp_utf16le_strings_z(const utf16lechar *s1, const utf16lechar *s2,
+ bool ignore_case)
{
- utf16lechar *name_utf16le;
- size_t name_utf16le_nbytes;
- int ret;
-#if TCHAR_IS_UTF16LE
- name_utf16le_nbytes = tstrlen(name) * sizeof(utf16lechar);
- name_utf16le = MALLOC(name_utf16le_nbytes + sizeof(utf16lechar));
- if (name_utf16le == NULL)
- return WIMLIB_ERR_NOMEM;
- memcpy(name_utf16le, name, name_utf16le_nbytes + sizeof(utf16lechar));
- ret = 0;
-#else
-
- ret = tstr_to_utf16le(name, tstrlen(name), &name_utf16le,
- &name_utf16le_nbytes);
- if (ret == 0) {
- if (name_utf16le_nbytes > 0xffff) {
- FREE(name_utf16le);
- ERROR("Multibyte string \"%"TS"\" is too long!", name);
- ret = WIMLIB_ERR_INVALID_UTF8_STRING;
+ if (ignore_case) {
+ for (;;) {
+ u16 c1 = upcase[le16_to_cpu(*s1)];
+ u16 c2 = upcase[le16_to_cpu(*s2)];
+ if (c1 != c2)
+ return (c1 < c2) ? -1 : 1;
+ if (c1 == 0)
+ return 0;
+ s1++, s2++;
}
+ } else {
+ while (*s1 && *s1 == *s2)
+ s1++, s2++;
+ if (*s1 == *s2)
+ return 0;
+ return (le16_to_cpu(*s1) < le16_to_cpu(*s2)) ? -1 : 1;
}
-#endif
- if (ret == 0) {
- FREE(*name_utf16le_ret);
- *name_utf16le_ret = name_utf16le;
- *name_utf16le_nbytes_ret = name_utf16le_nbytes;
+}
+
+/* Duplicate a UTF-16LE string. The input string might not be null terminated
+ * and might be misaligned, but the returned string is guaranteed to be null
+ * terminated and properly aligned. */
+utf16lechar *
+utf16le_dupz(const void *ustr, size_t usize)
+{
+ utf16lechar *dup = MALLOC(usize + sizeof(utf16lechar));
+ if (dup) {
+ memcpy(dup, ustr, usize);
+ dup[usize / sizeof(utf16lechar)] = 0;
}
- return ret;
+ return dup;
+}
+
+/* Duplicate a null-terminated UTF-16LE string. */
+utf16lechar *
+utf16le_dup(const utf16lechar *ustr)
+{
+ const utf16lechar *p = ustr;
+ while (*p++)
+ ;
+ return memdup(ustr, (const u8 *)p - (const u8 *)ustr);
}
+/* Return the length, in bytes, of a UTF-null terminated UTF-16 string,
+ * excluding the null terminator. */
+size_t
+utf16le_len_bytes(const utf16lechar *s)
+{
+ const utf16lechar *p = s;
+ while (*p)
+ p++;
+ return (p - s) * sizeof(utf16lechar);
+}
+
+/* Return the length, in UTF-16 coding units, of a UTF-null terminated UTF-16
+ * string, excluding the null terminator. */
+size_t
+utf16le_len_chars(const utf16lechar *s)
+{
+ return utf16le_len_bytes(s) / sizeof(utf16lechar);
+}