#include <string.h>
#include <stdlib.h>
-#ifdef G_OS_WIN32
+#include "glib.h"
+#include "config.h"
+
+#ifdef G_PLATFORM_WIN32
+#define STRICT
#include <windows.h>
+#undef STRICT
#endif
-#include "glib.h"
-
-#define _(s) (s)
+#include "glibintl.h"
GQuark
g_convert_error_quark()
static GQuark quark;
if (!quark)
quark = g_quark_from_static_string ("g_convert_error");
+
return quark;
}
-static iconv_t
+#if defined(USE_LIBICONV) && !defined (_LIBICONV_H)
+#error libiconv in use but included iconv.h not from libiconv
+#endif
+#if !defined(USE_LIBICONV) && defined (_LIBICONV_H)
+#error libiconv not in use but included iconv.h is from libiconv
+#endif
+
+/**
+ * g_iconv_open:
+ * @to_codeset: destination codeset
+ * @from_codeset: source codeset
+ *
+ * Same as the standard UNIX routine iconv_open(), but
+ * may be implemented via libiconv on UNIX flavors that lack
+ * a native implementation.
+ *
+ * GLib provides g_convert() and g_locale_to_utf8() which are likely
+ * more convenient than the raw iconv wrappers.
+ *
+ * Return value: a "conversion descriptor"
+ **/
+GIConv
+g_iconv_open (const gchar *to_codeset,
+ const gchar *from_codeset)
+{
+ iconv_t cd = iconv_open (to_codeset, from_codeset);
+
+ return (GIConv)cd;
+}
+
+/**
+ * g_iconv:
+ * @converter: conversion descriptor from g_iconv_open()
+ * @inbuf: bytes to convert
+ * @inbytes_left: inout parameter, bytes remaining to convert in @inbuf
+ * @outbuf: converted output bytes
+ * @outbytes_left: inout parameter, bytes available to fill in @outbuf
+ *
+ * Same as the standard UNIX routine iconv(), but
+ * may be implemented via libiconv on UNIX flavors that lack
+ * a native implementation.
+ *
+ * GLib provides g_convert() and g_locale_to_utf8() which are likely
+ * more convenient than the raw iconv wrappers.
+ *
+ * Return value: count of non-reversible conversions, or -1 on error
+ **/
+size_t
+g_iconv (GIConv converter,
+ gchar **inbuf,
+ size_t *inbytes_left,
+ gchar **outbuf,
+ size_t *outbytes_left)
+{
+ iconv_t cd = (iconv_t)converter;
+
+ return iconv (cd, inbuf, inbytes_left, outbuf, outbytes_left);
+}
+
+/**
+ * g_iconv_close:
+ * @converter: a conversion descriptor from g_iconv_open()
+ *
+ * Same as the standard UNIX routine iconv_close(), but
+ * may be implemented via libiconv on UNIX flavors that lack
+ * a native implementation. Should be called to clean up
+ * the conversion descriptor from iconv_open() when
+ * you are done converting things.
+ *
+ * GLib provides g_convert() and g_locale_to_utf8() which are likely
+ * more convenient than the raw iconv wrappers.
+ *
+ * Return value: -1 on error, 0 on success
+ **/
+gint
+g_iconv_close (GIConv converter)
+{
+ iconv_t cd = (iconv_t)converter;
+
+ return iconv_close (cd);
+}
+
+static GIConv
open_converter (const gchar *to_codeset,
- const gchar *from_codeset,
+ const gchar *from_codeset,
GError **error)
{
- iconv_t cd = iconv_open (to_codeset, from_codeset);
+ GIConv cd = g_iconv_open (to_codeset, from_codeset);
if (cd == (iconv_t) -1)
{
/* Something went wrong. */
if (errno == EINVAL)
- g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_NO_CONVERSION,
- _("Conversion from character set `%s' to `%s' is not supported"),
- from_codeset, to_codeset);
+ g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_NO_CONVERSION,
+ _("Conversion from character set `%s' to `%s' is not supported"),
+ from_codeset, to_codeset);
else
g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_FAILED,
- _("Could not open converter from `%s' to `%s': %s"),
- from_codeset, to_codeset, strerror (errno));
+ _("Could not open converter from `%s' to `%s': %s"),
+ from_codeset, to_codeset, strerror (errno));
}
return cd;
size_t inbytes_remaining;
size_t outbytes_remaining;
size_t err;
- iconv_t cd;
+ GIConv cd;
size_t outbuf_size;
gboolean have_error = FALSE;
cd = open_converter (to_codeset, from_codeset, error);
- if (cd == (iconv_t) -1)
+ if (cd == (GIConv) -1)
{
if (bytes_read)
*bytes_read = 0;
p = str;
inbytes_remaining = len;
- outbuf_size = len + 1; /* + 1 for nul in case len == 1 */
+
+ /* Due to a GLIBC bug, round outbuf_size up to a multiple of 4 */
+ /* + 1 for nul in case len == 1 */
+ outbuf_size = ((len + 3) & ~3) + 1;
+
outbytes_remaining = outbuf_size - 1; /* -1 for nul */
outp = dest = g_malloc (outbuf_size);
again:
- err = iconv (cd, &p, &inbytes_remaining, &outp, &outbytes_remaining);
+ err = g_iconv (cd, (char **)&p, &inbytes_remaining, &outp, &outbytes_remaining);
if (err == (size_t) -1)
{
case E2BIG:
{
size_t used = outp - dest;
- outbuf_size *= 2;
- dest = g_realloc (dest, outbuf_size);
- outp = dest + used;
- outbytes_remaining = outbuf_size - used - 1; /* -1 for nul */
+ /* glibc's iconv can return E2BIG even if there is space
+ * remaining if an internal buffer is exhausted. The
+ * folllowing is a heuristic to catch this. The 16 is
+ * pretty arbitrary.
+ */
+ if (used + 16 > outbuf_size)
+ {
+ outbuf_size = (outbuf_size - 1) * 2 + 1;
+ dest = g_realloc (dest, outbuf_size);
+
+ outp = dest + used;
+ outbytes_remaining = outbuf_size - used - 1; /* -1 for nul */
+ }
goto again;
}
*outp = '\0';
- iconv_close (cd);
+ g_iconv_close (cd);
if (bytes_read)
*bytes_read = p - str;
{
if ((p - str) != len)
{
- g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_PARTIAL_INPUT,
- _("Partial character sequence at end of input"));
- have_error = TRUE;
+ if (!have_error)
+ {
+ g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_PARTIAL_INPUT,
+ _("Partial character sequence at end of input"));
+ have_error = TRUE;
+ }
}
}
size_t save_inbytes = 0;
size_t outbytes_remaining;
size_t err;
- iconv_t cd;
+ GIConv cd;
size_t outbuf_size;
gboolean have_error = FALSE;
gboolean done = FALSE;
else
g_error_free (local_error);
+ local_error = NULL;
+
/* No go; to proceed, we need a converter from "UTF-8" to
* to_codeset, and the string as UTF-8.
*/
cd = open_converter (to_codeset, "UTF-8", error);
- if (cd == (iconv_t) -1)
+ if (cd == (GIConv) -1)
{
if (bytes_read)
*bytes_read = 0;
* for the original string while we are converting the fallback
*/
p = utf8;
- outbuf_size = len + 1; /* + 1 for nul in case len == 1 */
+ /* Due to a GLIBC bug, round outbuf_size up to a multiple of 4 */
+ /* + 1 for nul in case len == 1 */
+ outbuf_size = ((len + 3) & ~3) + 1;
outbytes_remaining = outbuf_size - 1; /* -1 for nul */
outp = dest = g_malloc (outbuf_size);
while (!done && !have_error)
{
size_t inbytes_tmp = inbytes_remaining;
- err = iconv (cd, &p, &inbytes_tmp, &outp, &outbytes_remaining);
+ err = g_iconv (cd, (char **)&p, &inbytes_tmp, &outp, &outbytes_remaining);
inbytes_remaining = inbytes_tmp;
if (err == (size_t) -1)
case E2BIG:
{
size_t used = outp - dest;
- outbuf_size *= 2;
- dest = g_realloc (dest, outbuf_size);
-
- outp = dest + used;
- outbytes_remaining = outbuf_size - used - 1; /* -1 for nul */
+
+ /* glibc's iconv can return E2BIG even if there is space
+ * remaining if an internal buffer is exhausted. The
+ * folllowing is a heuristic to catch this. The 16 is
+ * pretty arbitrary.
+ */
+ if (used + 16 > outbuf_size)
+ {
+ outbuf_size = (outbuf_size - 1) * 2 + 1;
+ dest = g_realloc (dest, outbuf_size);
+
+ outp = dest + used;
+ outbytes_remaining = outbuf_size - used - 1; /* -1 for nul */
+ }
break;
}
*/
*outp = '\0';
- iconv_close (cd);
+ g_iconv_close (cd);
if (bytes_written)
*bytes_written = outp - str; /* Doesn't include '\0' */
/*
* g_locale_to_utf8
*
+ *
+ */
+
+/**
+ * g_locale_to_utf8:
+ * @opsysstring: a string in the encoding of the current locale
+ * @len: the length of the string, or -1 if the string is
+ * NULL-terminated.
+ * @bytes_read: location to store the number of bytes in the
+ * input string that were successfully converted, or %NULL.
+ * Even if the conversion was succesful, this may be
+ * less than len if there were partial characters
+ * at the end of the input. If the error
+ * G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
+ * stored will the byte fofset after the last valid
+ * input sequence.
+ * @bytes_written: the stored in the output buffer (not including the
+ * terminating nul.
+ * @error: location to store the error occuring, or %NULL to ignore
+ * errors. Any of the errors in #GConvertError may occur.
+ *
* Converts a string which is in the encoding used for strings by
* the C runtime (usually the same as that used by the operating
* system) in the current locale into a UTF-8 string.
- */
-
+ *
+ * Return value: The converted string, or %NULL on an error.
+ **/
gchar *
-g_locale_to_utf8 (const gchar *opsysstring, GError **error)
+g_locale_to_utf8 (const gchar *opsysstring,
+ gint len,
+ gint *bytes_read,
+ gint *bytes_written,
+ GError **error)
{
-#ifdef G_OS_WIN32
+#ifdef G_PLATFORM_WIN32
- gint i, clen, wclen, first;
- const gint len = strlen (opsysstring);
+ gint i, clen, total_len, wclen, first;
wchar_t *wcs, wc;
gchar *result, *bp;
const wchar_t *wcp;
+ if (len == -1)
+ len = strlen (opsysstring);
+
wcs = g_new (wchar_t, len);
wclen = MultiByteToWideChar (CP_ACP, 0, opsysstring, len, wcs, len);
wcp = wcs;
- clen = 0;
+ total_len = 0;
for (i = 0; i < wclen; i++)
{
wc = *wcp++;
if (wc < 0x80)
- clen += 1;
+ total_len += 1;
else if (wc < 0x800)
- clen += 2;
+ total_len += 2;
else if (wc < 0x10000)
- clen += 3;
+ total_len += 3;
else if (wc < 0x200000)
- clen += 4;
+ total_len += 4;
else if (wc < 0x4000000)
- clen += 5;
+ total_len += 5;
else
- clen += 6;
+ total_len += 6;
}
- result = g_malloc (clen + 1);
+ result = g_malloc (total_len + 1);
wcp = wcs;
bp = result;
g_free (wcs);
+ if (bytes_read)
+ *bytes_read = len;
+ if (bytes_written)
+ *bytes_written = total_len;
+
return result;
-#else
+#else /* !G_PLATFORM_WIN32 */
char *charset, *str;
if (g_get_charset (&charset))
return g_strdup (opsysstring);
- str = g_convert (opsysstring, strlen (opsysstring),
- "UTF-8", charset, NULL, NULL, error);
+ str = g_convert (opsysstring, len,
+ "UTF-8", charset, bytes_read, bytes_written, error);
return str;
-#endif
+#endif /* !G_PLATFORM_WIN32 */
}
-/*
- * g_locale_from_utf8
- *
- * The reverse of g_locale_to_utf8.
- */
-
+/**
+ * g_locale_from_utf8:
+ * @utf8string: a UTF-8 encoded string
+ * @len: the length of the string, or -1 if the string is
+ * NULL-terminated.
+ * @bytes_read: location to store the number of bytes in the
+ * input string that were successfully converted, or %NULL.
+ * Even if the conversion was succesful, this may be
+ * less than len if there were partial characters
+ * at the end of the input. If the error
+ * G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
+ * stored will the byte fofset after the last valid
+ * input sequence.
+ * @bytes_written: the stored in the output buffer (not including the
+ * terminating nul.
+ * @error: location to store the error occuring, or %NULL to ignore
+ * errors. Any of the errors in #GConvertError may occur.
+ *
+ * Converts a string from UTF-8 to the encoding used for strings by
+ * the C runtime (usually the same as that used by the operating
+ * system) in the current locale.
+ *
+ * Return value: The converted string, or %NULL on an error.
+ **/
gchar *
-g_locale_from_utf8 (const gchar *utf8string, GError **error)
+g_locale_from_utf8 (const gchar *utf8string,
+ gint len,
+ gint *bytes_read,
+ gint *bytes_written,
+ GError **error)
{
-#ifdef G_OS_WIN32
+#ifdef G_PLATFORM_WIN32
gint i, mask, clen, mblen;
- const gint len = strlen (utf8string);
wchar_t *wcs, *wcp;
gchar *result;
guchar *cp, *end, c;
gint n;
+ if (len == -1)
+ len = strlen (utf8string);
+
/* First convert to wide chars */
cp = (guchar *) utf8string;
end = cp + len;
result[mblen] = 0;
g_free (wcs);
+ if (bytes_read)
+ *bytes_read = len;
+ if (bytes_written)
+ *bytes_written = mblen;
+
return result;
-#else
+#else /* !G_PLATFORM_WIN32 */
gchar *charset, *str;
return g_strdup (utf8string);
str = g_convert (utf8string, strlen (utf8string),
- charset, "UTF-8", NULL, NULL, error);
+ charset, "UTF-8", bytes_read, bytes_written, error);
return str;
-#endif
+#endif /* !G_PLATFORM_WIN32 */
}
-/* Filenames are in UTF-8 unless specificially requested otherwise */
-
+/**
+ * g_filename_to_utf8:
+ * @opsysstring: a string in the encoding for filenames
+ * @len: the length of the string, or -1 if the string is
+ * NULL-terminated.
+ * @bytes_read: location to store the number of bytes in the
+ * input string that were successfully converted, or %NULL.
+ * Even if the conversion was succesful, this may be
+ * less than len if there were partial characters
+ * at the end of the input. If the error
+ * G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
+ * stored will the byte fofset after the last valid
+ * input sequence.
+ * @bytes_written: the stored in the output buffer (not including the
+ * terminating nul.
+ * @error: location to store the error occuring, or %NULL to ignore
+ * errors. Any of the errors in #GConvertError may occur.
+ *
+ * Converts a string which is in the encoding used for filenames
+ * into a UTF-8 string.
+ *
+ * Return value: The converted string, or %NULL on an error.
+ **/
gchar*
-g_filename_to_utf8 (const gchar *string, GError **error)
+g_filename_to_utf8 (const gchar *opsysstring,
+ gint len,
+ gint *bytes_read,
+ gint *bytes_written,
+ GError **error)
{
-#ifdef G_OS_WIN32
- return g_locale_to_utf8 (string, error);
-#else
+#ifdef G_PLATFORM_WIN32
+ return g_locale_to_utf8 (opsysstring, len,
+ bytes_read, bytes_written,
+ error);
+#else /* !G_PLATFORM_WIN32 */
if (getenv ("G_BROKEN_FILENAMES"))
- return g_locale_to_utf8 (string, error);
+ return g_locale_to_utf8 (opsysstring, len,
+ bytes_read, bytes_written,
+ error);
- return g_strdup (string);
-#endif
+ if (bytes_read || bytes_written)
+ {
+ gint len = strlen (opsysstring);
+
+ if (bytes_read)
+ *bytes_read = len;
+ if (bytes_written)
+ *bytes_written = len;
+ }
+
+ if (len < 0)
+ return g_strdup (opsysstring);
+ else
+ return g_strndup (opsysstring, len);
+#endif /* !G_PLATFORM_WIN32 */
}
+/**
+ * g_filename_from_utf8:
+ * @utf8string: a UTF-8 encoded string
+ * @len: the length of the string, or -1 if the string is
+ * NULL-terminated.
+ * @bytes_read: location to store the number of bytes in the
+ * input string that were successfully converted, or %NULL.
+ * Even if the conversion was succesful, this may be
+ * less than len if there were partial characters
+ * at the end of the input. If the error
+ * G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
+ * stored will the byte fofset after the last valid
+ * input sequence.
+ * @bytes_written: the stored in the output buffer (not including the
+ * terminating nul.
+ * @error: location to store the error occuring, or %NULL to ignore
+ * errors. Any of the errors in #GConvertError may occur.
+ *
+ * Converts a string from UTF-8 to the encoding used for filenames.
+ *
+ * Return value: The converted string, or %NULL on an error.
+ **/
gchar*
-g_filename_from_utf8 (const gchar *string, GError **error)
+g_filename_from_utf8 (const gchar *utf8string,
+ gint len,
+ gint *bytes_read,
+ gint *bytes_written,
+ GError **error)
{
-#ifdef G_OS_WIN32
- return g_locale_from_utf8 (string, error);
-#else
+#ifdef G_PLATFORM_WIN32
+ return g_locale_from_utf8 (utf8string, len,
+ bytes_read, bytes_written,
+ error);
+#else /* !G_PLATFORM_WIN32 */
if (getenv ("G_BROKEN_FILENAMES"))
- return g_locale_from_utf8 (string, error);
+ return g_locale_from_utf8 (utf8string, len,
+ bytes_read, bytes_written,
+ error);
- return g_strdup (string);
-#endif
-}
+ if (bytes_read || bytes_written)
+ {
+ gint len = strlen (utf8string);
+ if (bytes_read)
+ *bytes_read = len;
+ if (bytes_written)
+ *bytes_written = len;
+ }
+
+ if (len < 0)
+ return g_strdup (utf8string);
+ else
+ return g_strndup (utf8string, len);
+#endif /* !G_PLATFORM_WIN32 */
+}