aboutsummaryrefslogtreecommitdiff
path: root/wcsmbs/mbrtoc16.c
diff options
context:
space:
mode:
authorUlrich Drepper <drepper@gmail.com>2012-01-07 10:52:53 -0500
committerUlrich Drepper <drepper@gmail.com>2012-01-07 10:52:53 -0500
commit9954432e309c8fddaec2fe53e601702a5c981624 (patch)
tree3eb7513694e25391b3393afbb847dbd85ebf097a /wcsmbs/mbrtoc16.c
parentc3a87236702cb73be1dada3438bbd3c3934e83f8 (diff)
downloadglibc-9954432e309c8fddaec2fe53e601702a5c981624.zip
glibc-9954432e309c8fddaec2fe53e601702a5c981624.tar.gz
glibc-9954432e309c8fddaec2fe53e601702a5c981624.tar.bz2
More char16_t and char32_t support
It works now for UTF-8 locales
Diffstat (limited to 'wcsmbs/mbrtoc16.c')
-rw-r--r--wcsmbs/mbrtoc16.c52
1 files changed, 45 insertions, 7 deletions
diff --git a/wcsmbs/mbrtoc16.c b/wcsmbs/mbrtoc16.c
index 7b5822d..df970fb 100644
--- a/wcsmbs/mbrtoc16.c
+++ b/wcsmbs/mbrtoc16.c
@@ -1,6 +1,6 @@
/* Copyright (C) 2011, 2012 Free Software Foundation, Inc.
This file is part of the GNU C Library.
- Contributed by Ulrich Drepper <drepper@gnu.org>, 2011.
+ Contributed by Ulrich Drepper <drepper@gmail.com>, 2011.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
@@ -43,20 +43,32 @@ static mbstate_t state;
size_t
mbrtoc16 (char16_t *pc16, const char *s, size_t n, mbstate_t *ps)
{
- char16_t buf[1];
+ if (ps == NULL)
+ ps = &state;
+
+ if (ps->__count & 0x80000000)
+ {
+ /* We have to return the second word for a surrogate. */
+ ps->__count &= 0x7fffffff;
+ *pc16 = ps->__value.__wch;
+ ps->__value.__wch = L'\0';
+ return (size_t) -3;
+ }
+
+ char16_t buf[2];
struct __gconv_step_data data;
int status;
size_t result;
size_t dummy;
const unsigned char *inbuf, *endbuf;
- unsigned char *outbuf = (unsigned char *) (pc16 ?: buf);
+ unsigned char *outbuf = (unsigned char *) buf;
const struct gconv_fcts *fcts;
/* Set information for this step. */
data.__invocation_counter = 0;
data.__internal_use = 1;
data.__flags = __GCONV_IS_LAST;
- data.__statep = ps ?: &state;
+ data.__statep = ps;
data.__trans = NULL;
/* A first special case is if S is NULL. This means put PS in the
@@ -85,9 +97,22 @@ mbrtoc16 (char16_t *pc16, const char *s, size_t n, mbstate_t *ps)
if (fcts->toc16->__shlib_handle != NULL)
PTR_DEMANGLE (fct);
#endif
+
+ /* We first have to check whether the character can be represented
+ without a surrogate. If we immediately pass in a buffer large
+ enough to hold two char16_t values and the first character does
+ not require a surrogate the routine will try to convert more
+ input if N is larger then needed for the first character. */
status = DL_CALL_FCT (fct, (fcts->toc16, &data, &inbuf, endbuf,
NULL, &dummy, 0, 1));
+ if (status == __GCONV_FULL_OUTPUT && data.__outbuf == outbuf)
+ {
+ data.__outbufend = outbuf + 2 * sizeof (char16_t);
+ status = DL_CALL_FCT (fct, (fcts->toc16, &data, &inbuf, endbuf,
+ NULL, &dummy, 0, 1));
+ }
+
/* There must not be any problems with the conversion but illegal input
characters. The output buffer must be large enough, otherwise the
definition of MB_CUR_MAX is not correct. All the other possible
@@ -100,15 +125,28 @@ mbrtoc16 (char16_t *pc16, const char *s, size_t n, mbstate_t *ps)
if (status == __GCONV_OK || status == __GCONV_EMPTY_INPUT
|| status == __GCONV_FULL_OUTPUT)
{
- if (data.__outbuf != (unsigned char *) outbuf
- && *(char16_t *) outbuf == U('\0'))
+ if (pc16 != NULL)
+ *pc16 = buf[0];
+
+ if (data.__outbuf != outbuf && *(char16_t *) outbuf == U('\0'))
{
/* The converted character is the NUL character. */
assert (__mbsinit (data.__statep));
result = 0;
}
else
- result = inbuf - (const unsigned char *) s;
+ {
+ result = inbuf - (const unsigned char *) s;
+
+ if (data.__outbuf != outbuf + 2)
+ {
+ /* This is a surrogate. */
+ assert (buf[0] >= 0xd800 && buf[0] <= 0xdfff);
+ assert (buf[1] >= 0xdc00 && buf[1] <= 0xdfff);
+ ps->__count |= 0x80000000;
+ ps->__value.__wch = buf[1];
+ }
+ }
}
else if (status == __GCONV_INCOMPLETE_INPUT)
result = (size_t) -2;