mirror of
https://sourceware.org/git/glibc.git
synced 2024-11-18 02:50:09 +00:00
fd1b5c0fb6
2000-04-09 Ulrich Drepper <drepper@redhat.com> Implement handling of restartable conversion functions according to ISO C. * iconv/gconv.h (__gconv_fct): Add additional parameter. * iconv/gconv_int.h (__BUILTIN_TRANS): Likewise. * iconv/gconv.c: Pass additional parameter to conversion function. * iconv/gconv_simple.c (internal_ucs4_loop_single): New function. (internal_ucs4le_loop_single): New function. (__gconv_transform_ascii_internal): Define ONE_DIRECTION. (__gconv_transform_internal_ascii): Likewise. (__gconv_transform_internal_utf8): Likewise. (__gconv_transform_utf8_internal): Likewise. (__gconv_transform_ucs2_internal): Likewise. (__gconv_transform_internal_ucs2): Likewise. (__gconv_transform_ucs2reverse_internal): Likewise. (__gconv_transform_internal_ucs2reverse): Likewise. (internal_ucs4le_loop_unaligned): Before return __GCONV_INCOMPLETE_INPUT check that the remaining bytes really form a valid character. Otherwise return __GCONV_ILLEGAL_INPUT. (__gconv_transform_utf8_internal): Define STORE_REST and UNPACK_BYTES. * iconv/loop.c: Fit in definition of function to convert one character for processing of left-over bytes from the state object. * iconv/skeleton.c (gconv): Rename inbuf to inptrp and inbufend to inend to match names in loop functions. (RESET_INPUT_BUFFER): Change apprpriately. (gconv): If needed, call function to process bytes from the state object. Similar at the end: store left over bytes if input is incomplete. Take extra argument and add new argument to all calls of the conversion function. * iconvdata/iso-2022-cn.c: Adjust numeric values used to store information in the state object to not conflict with length count. * iconvdata/iso-2022-jp.c: Likewise. * iconvdata/iso-2022-kr.c: Likewise. * iconvdata/unicode.c: Adjust for change change in parameters of skeleton function. * iconvdata/utf-16.c: Likewise. * libio/iofwide.c: Add new parameter to all calls of conversion function. * wcsmbs/btowc.c: Likewise. * wcsmbs/mbrtowc.c: Likewise. * wcsmbs/mbsnrtowcs.c: Likewise. * wcsmbs/mbsrtowcs.c: Likewise. * wcsmbs/wcrtomb.c: Likewise. * wcsmbs/wcsnrtombs.c: Likewise. * wcsmbs/wcsrtombs.c: Likewise. * wcsmbs/wctob.c: Likewise. * iconvdata/gbgbk.c: Always define MAX_NEEDED_OUTPUT and MAX_NEEDED_INPUT.
348 lines
9.9 KiB
C
348 lines
9.9 KiB
C
/* Conversion module for UTF-16.
|
|
Copyright (C) 1999, 2000 Free Software Foundation, Inc.
|
|
This file is part of the GNU C Library.
|
|
Contributed by Ulrich Drepper <drepper@cygnus.com>, 1999.
|
|
|
|
The GNU C Library is free software; you can redistribute it and/or
|
|
modify it under the terms of the GNU Library General Public License as
|
|
published by the Free Software Foundation; either version 2 of the
|
|
License, or (at your option) any later version.
|
|
|
|
The GNU C Library 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
|
|
Library General Public License for more details.
|
|
|
|
You should have received a copy of the GNU Library General Public
|
|
License along with the GNU C Library; see the file COPYING.LIB. If not,
|
|
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
|
|
Boston, MA 02111-1307, USA. */
|
|
|
|
#include <byteswap.h>
|
|
#include <gconv.h>
|
|
#include <stddef.h>
|
|
#include <stdint.h>
|
|
#include <stdlib.h>
|
|
#include <string.h>
|
|
|
|
/* This is the Byte Order Mark character (BOM). */
|
|
#define BOM 0xfeff
|
|
/* And in the other byte order. */
|
|
#define BOM_OE 0xfffe
|
|
|
|
|
|
/* Definitions used in the body of the `gconv' function. */
|
|
#define FROM_LOOP from_utf16_loop
|
|
#define TO_LOOP to_utf16_loop
|
|
#define DEFINE_INIT 0
|
|
#define DEFINE_FINI 0
|
|
#define MIN_NEEDED_FROM 2
|
|
#define MAX_NEEDED_FROM 4
|
|
#define MIN_NEEDED_TO 4
|
|
#define FROM_DIRECTION (dir == from_utf16)
|
|
#define PREPARE_LOOP \
|
|
enum direction dir = ((struct utf16_data *) step->__data)->dir; \
|
|
enum variant var = ((struct utf16_data *) step->__data)->var; \
|
|
int swap = ((struct utf16_data *) step->__data)->swap; \
|
|
if (FROM_DIRECTION || var == UTF_16) \
|
|
{ \
|
|
if (data->__invocation_counter == 0) \
|
|
{ \
|
|
/* We have to find out which byte order the file is encoded in. */ \
|
|
if (inptr + 2 > inend) \
|
|
return __GCONV_EMPTY_INPUT; \
|
|
\
|
|
if (get16u (inptr) == BOM) \
|
|
/* Simply ignore the BOM character. */ \
|
|
inptr += 2; \
|
|
else if (get16u (inptr) == BOM_OE) \
|
|
{ \
|
|
((struct utf16_data *) step->__data)->swap = 1; \
|
|
inptr += 2; \
|
|
} \
|
|
} \
|
|
} \
|
|
else if (!FROM_DIRECTION && var == UTF_16 && !data->__internal_use \
|
|
&& data->__invocation_counter == 0) \
|
|
{ \
|
|
/* Emit the Byte Order Mark. */ \
|
|
if (outbuf + 2 > outend) \
|
|
return __GCONV_FULL_OUTPUT; \
|
|
\
|
|
put16u (outbuf, BOM); \
|
|
outbuf += 2; \
|
|
}
|
|
#define EXTRA_LOOP_ARGS , var, data, swap
|
|
|
|
|
|
/* Direction of the transformation. */
|
|
enum direction
|
|
{
|
|
illegal_dir,
|
|
to_utf16,
|
|
from_utf16
|
|
};
|
|
|
|
enum variant
|
|
{
|
|
illegal_var,
|
|
UTF_16,
|
|
UTF_16LE,
|
|
UTF_16BE
|
|
};
|
|
|
|
struct utf16_data
|
|
{
|
|
enum direction dir;
|
|
enum variant var;
|
|
int swap;
|
|
};
|
|
|
|
|
|
int
|
|
gconv_init (struct __gconv_step *step)
|
|
{
|
|
/* Determine which direction. */
|
|
struct utf16_data *new_data;
|
|
enum direction dir = illegal_dir;
|
|
enum variant var = illegal_var;
|
|
int result;
|
|
|
|
if (__strcasecmp (step->__from_name, "UTF-16") == 0)
|
|
{
|
|
dir = from_utf16;
|
|
var = UTF_16;
|
|
}
|
|
else if (__strcasecmp (step->__to_name, "UTF-16") == 0)
|
|
{
|
|
dir = to_utf16;
|
|
var = UTF_16;
|
|
}
|
|
else if (__strcasecmp (step->__from_name, "UTF-16BE") == 0)
|
|
{
|
|
dir = from_utf16;
|
|
var = UTF_16BE;
|
|
}
|
|
else if (__strcasecmp (step->__to_name, "UTF-16BE") == 0)
|
|
{
|
|
dir = to_utf16;
|
|
var = UTF_16BE;
|
|
}
|
|
else if (__strcasecmp (step->__from_name, "UTF-16LE") == 0)
|
|
{
|
|
dir = from_utf16;
|
|
var = UTF_16LE;
|
|
}
|
|
else if (__strcasecmp (step->__to_name, "UTF-16LE") == 0)
|
|
{
|
|
dir = to_utf16;
|
|
var = UTF_16LE;
|
|
}
|
|
|
|
result = __GCONV_NOCONV;
|
|
if (dir != illegal_dir)
|
|
{
|
|
new_data = (struct utf16_data *) malloc (sizeof (struct utf16_data));
|
|
|
|
result = __GCONV_NOMEM;
|
|
if (new_data != NULL)
|
|
{
|
|
new_data->dir = dir;
|
|
new_data->var = var;
|
|
new_data->swap = ((var == UTF_16LE && BYTE_ORDER == BIG_ENDIAN)
|
|
|| (var == UTF_16BE
|
|
&& BYTE_ORDER == LITTLE_ENDIAN));
|
|
step->__data = new_data;
|
|
|
|
if (dir == from_utf16)
|
|
{
|
|
step->__min_needed_from = MIN_NEEDED_FROM;
|
|
step->__max_needed_from = MIN_NEEDED_FROM;
|
|
step->__min_needed_to = MIN_NEEDED_TO;
|
|
step->__max_needed_to = MIN_NEEDED_TO;
|
|
}
|
|
else
|
|
{
|
|
step->__min_needed_from = MIN_NEEDED_TO;
|
|
step->__max_needed_from = MIN_NEEDED_TO;
|
|
step->__min_needed_to = MIN_NEEDED_FROM;
|
|
step->__max_needed_to = MIN_NEEDED_FROM;
|
|
}
|
|
|
|
step->__stateful = 0;
|
|
|
|
result = __GCONV_OK;
|
|
}
|
|
}
|
|
|
|
return result;
|
|
}
|
|
|
|
|
|
void
|
|
gconv_end (struct __gconv_step *data)
|
|
{
|
|
free (data->__data);
|
|
}
|
|
|
|
|
|
/* Convert from the internal (UCS4-like) format to UTF-16. */
|
|
#define MIN_NEEDED_INPUT MIN_NEEDED_TO
|
|
#define MIN_NEEDED_OUTPUT MIN_NEEDED_FROM
|
|
#define MAX_NEEDED_OUTPUT MAX_NEEDED_FROM
|
|
#define LOOPFCT TO_LOOP
|
|
#define BODY \
|
|
{ \
|
|
uint32_t c = get32 (inptr); \
|
|
\
|
|
if (swap) \
|
|
{ \
|
|
if (c >= 0x10000) \
|
|
{ \
|
|
if (c >= 0x110000) \
|
|
{ \
|
|
result = __GCONV_ILLEGAL_INPUT; \
|
|
break; \
|
|
} \
|
|
\
|
|
/* Generate a surrogate character. */ \
|
|
if (NEED_LENGTH_TEST && outptr + 4 > outend) \
|
|
{ \
|
|
/* Overflow in the output buffer. */ \
|
|
result = __GCONV_FULL_OUTPUT; \
|
|
break; \
|
|
} \
|
|
\
|
|
put16 (outptr, bswap_16 (0xd7c0 + (c >> 10))); \
|
|
outptr += 2; \
|
|
put16 (outptr, bswap_16 (0xdc00 + (c & 0x3ff))); \
|
|
} \
|
|
else \
|
|
put16 (outptr, bswap_16 (c)); \
|
|
} \
|
|
else \
|
|
{ \
|
|
if (c >= 0x10000) \
|
|
{ \
|
|
if (c >= 0x110000) \
|
|
{ \
|
|
result = __GCONV_ILLEGAL_INPUT; \
|
|
break; \
|
|
} \
|
|
\
|
|
/* Generate a surrogate character. */ \
|
|
if (NEED_LENGTH_TEST && outptr + 4 > outend) \
|
|
{ \
|
|
/* Overflow in the output buffer. */ \
|
|
result = __GCONV_FULL_OUTPUT; \
|
|
break; \
|
|
} \
|
|
\
|
|
put16 (outptr, 0xd7c0 + (c >> 10)); \
|
|
outptr += 2; \
|
|
put16 (outptr, 0xdc00 + (c & 0x3ff)); \
|
|
} \
|
|
else \
|
|
put16 (outptr, c); \
|
|
} \
|
|
outptr += 2; \
|
|
inptr += 4; \
|
|
}
|
|
#define EXTRA_LOOP_DECLS \
|
|
, enum variant var, struct __gconv_step_data *step_data, int swap
|
|
#include <iconv/loop.c>
|
|
|
|
|
|
/* Convert from UTF-16 to the internal (UCS4-like) format. */
|
|
#define MIN_NEEDED_INPUT MIN_NEEDED_FROM
|
|
#define MAX_NEEDED_INPUT MAX_NEEDED_FROM
|
|
#define MIN_NEEDED_OUTPUT MIN_NEEDED_TO
|
|
#define LOOPFCT FROM_LOOP
|
|
#define BODY \
|
|
{ \
|
|
uint16_t u1 = get16 (inptr); \
|
|
\
|
|
if (swap) \
|
|
{ \
|
|
u1 = bswap_16 (u1); \
|
|
\
|
|
if (u1 < 0xd800 || u1 > 0xdfff) \
|
|
{ \
|
|
/* No surrogate. */ \
|
|
put32 (outptr, u1); \
|
|
inptr += 2; \
|
|
} \
|
|
else \
|
|
{ \
|
|
uint16_t u2; \
|
|
\
|
|
/* It's a surrogate character. At least the first word says \
|
|
it is. */ \
|
|
if (NEED_LENGTH_TEST && inptr + 4 > inend) \
|
|
{ \
|
|
/* We don't have enough input for another complete input \
|
|
character. */ \
|
|
result = __GCONV_INCOMPLETE_INPUT; \
|
|
break; \
|
|
} \
|
|
\
|
|
inptr += 2; \
|
|
u2 = bswap_16 (get16 (inptr)); \
|
|
if (u2 < 0xdc00 || u2 >= 0xdfff) \
|
|
{ \
|
|
/* This is no valid second word for a surrogate. */ \
|
|
result = __GCONV_ILLEGAL_INPUT; \
|
|
inptr -= 2; \
|
|
break; \
|
|
} \
|
|
\
|
|
put32 (outptr, ((u1 - 0xd7c0) << 10) + (u2 - 0xdc00)); \
|
|
inptr += 2; \
|
|
} \
|
|
} \
|
|
else \
|
|
{ \
|
|
if (u1 < 0xd800 || u1 > 0xdfff) \
|
|
{ \
|
|
/* No surrogate. */ \
|
|
put32 (outptr, u1); \
|
|
inptr += 2; \
|
|
} \
|
|
else \
|
|
{ \
|
|
uint16_t u2; \
|
|
\
|
|
/* It's a surrogate character. At least the first word says \
|
|
it is. */ \
|
|
if (NEED_LENGTH_TEST && inptr + 4 > inend) \
|
|
{ \
|
|
/* We don't have enough input for another complete input \
|
|
character. */ \
|
|
result = __GCONV_INCOMPLETE_INPUT; \
|
|
break; \
|
|
} \
|
|
\
|
|
inptr += 2; \
|
|
u2 = get16 (inptr); \
|
|
if (u2 < 0xdc00 || u2 >= 0xdfff) \
|
|
{ \
|
|
/* This is no valid second word for a surrogate. */ \
|
|
result = __GCONV_ILLEGAL_INPUT; \
|
|
inptr -= 2; \
|
|
break; \
|
|
} \
|
|
\
|
|
put32 (outptr, ((u1 - 0xd7c0) << 10) + (u2 - 0xdc00)); \
|
|
inptr += 2; \
|
|
} \
|
|
} \
|
|
outptr += 4; \
|
|
}
|
|
#define EXTRA_LOOP_DECLS \
|
|
, enum variant var, struct __gconv_step_data *step_data, int swap
|
|
#include <iconv/loop.c>
|
|
|
|
|
|
/* Now define the toplevel functions. */
|
|
#include <iconv/skeleton.c>
|