summaryrefslogtreecommitdiff
path: root/iconvdata
diff options
context:
space:
mode:
authorUlrich Drepper <drepper@redhat.com>2000-03-21 20:18:34 +0000
committerUlrich Drepper <drepper@redhat.com>2000-03-21 20:18:34 +0000
commit8d617a716df0ed5fd9ea1c8e65dd8e59b168573e (patch)
tree9bae23f82c38958c166ca3d65bb308f52ab2d14c /iconvdata
parentbc4831b956f96efd9f4185b739b8ce8f3fa4dae6 (diff)
downloadglibc-8d617a716df0ed5fd9ea1c8e65dd8e59b168573e.tar.gz
Update.
* iconv/gconv_builtin.c: Include <endian.h>. * iconv/gconv_builtin.h: Add UCS-BE aliases. Add UCS-4LE transformation. Define UNICODEBIG and UNICODELITTLE according to current platform. * iconv/gconv_int.h: Declare __gconv_transform_ucs2reverse_internal, __gconv_transform_internal_ucs2reverse, and __gconv_transform_internal_ucs4le. * iconv/gconv_simple.c: Implement __gconv_transform_internal_ucs4le, __gconv_transform_ucs2reverse_internal and __gconv_transform_internal_ucs2reverse. * iconvdata/Makefile (modules): Add UNICODE. (distribute): Add unicode.c. * iconvdata/gconv-modules: Add definitions for UNICODE module. * iconvdata/unicode.c: New file. * iconvdata/utf-16.c: Rewrite code to emit BOM. Correct code to determine byte order of input and convert accordingly.
Diffstat (limited to 'iconvdata')
-rw-r--r--iconvdata/Makefile4
-rw-r--r--iconvdata/gconv-modules5
-rw-r--r--iconvdata/unicode.c190
-rw-r--r--iconvdata/utf-16.c51
4 files changed, 229 insertions, 21 deletions
diff --git a/iconvdata/Makefile b/iconvdata/Makefile
index dee418e6a9..05ae2ac62e 100644
--- a/iconvdata/Makefile
+++ b/iconvdata/Makefile
@@ -45,7 +45,7 @@ modules := ISO8859-1 ISO8859-2 ISO8859-3 ISO8859-4 ISO8859-5 \
INIS-CYRILLIC ISO_6937-2 ISO_2033 ISO_5427 ISO_5427-EXT \
ISO_5428 ISO_10367-BOX MAC-IS MAC-UK NATS-DANO NATS-SEFI \
SAMI-WS2 ISO-IR-197 TIS-620 KOI8-U GBK ISIRI-3342 GBGBK \
- ISO-2022-CN libISOIR165 UTF-16
+ ISO-2022-CN libISOIR165 UTF-16 UNICODE
modules.so := $(addsuffix .so, $(modules))
@@ -118,7 +118,7 @@ distribute := gconv-modules extra-module.mk gap.awk gaptab.awk \
macintosh.c mac-is.c mac-uk.c nats-dano.c nats-sefi.c sjis.c \
t.61.c uhc.c sami-ws2.c iso-ir-197.c tis-620.c koi8-u.c \
isiri-3342.c gbgbk.c iso-2022-cn.c cns11643l2.h iso8859-16.c \
- utf-16.c
+ utf-16.c unicode.c
# We build the transformation modules only when we build shared libs.
ifeq (yes,$(build-shared))
diff --git a/iconvdata/gconv-modules b/iconvdata/gconv-modules
index 6a3c521c3c..3ca028501b 100644
--- a/iconvdata/gconv-modules
+++ b/iconvdata/gconv-modules
@@ -1184,3 +1184,8 @@ module INTERNAL UTF-16LE// UTF-16 1
# from to module cost
module UTF-16BE// INTERNAL UTF-16 1
module INTERNAL UTF-16BE// UTF-16 1
+
+# from to module cost
+alias CSUNICODE// UNICODE//
+module UNICODE// INTERNAL UNICODE 1
+module INTERNAL UNICODE// UNICODE 1
diff --git a/iconvdata/unicode.c b/iconvdata/unicode.c
new file mode 100644
index 0000000000..b29976b0c8
--- /dev/null
+++ b/iconvdata/unicode.c
@@ -0,0 +1,190 @@
+/* Conversion module for Unicode
+ 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 endian format. */
+#define BOM_OE 0xfffe
+
+
+/* Definitions used in the body of the `gconv' function. */
+#define FROM_LOOP from_unicode_loop
+#define TO_LOOP to_unicode_loop
+#define DEFINE_INIT 0
+#define DEFINE_FINI 0
+#define MIN_NEEDED_FROM 2
+#define MIN_NEEDED_TO 4
+#define FROM_DIRECTION (dir == from_unicode)
+#define PREPARE_LOOP \
+ enum direction dir = ((struct unicode_data *) step->__data)->dir; \
+ int swap; \
+ if (FROM_DIRECTION) \
+ { \
+ if (data->__invocation_counter == 0) \
+ { \
+ /* We have to find out which byte order the file is encoded in. */ \
+ if (inptr + 2 > inbufend) \
+ return __GCONV_EMPTY_INPUT; \
+ \
+ if (*(uint16_t *) inptr == BOM) \
+ /* Simply ignore the BOM character. */ \
+ inptr += 2; \
+ else if (*(uint16_t *) inptr == BOM_OE) \
+ { \
+ ((struct unicode_data *) step->__data)->swap = 1; \
+ inptr += 2; \
+ } \
+ } \
+ } \
+ else if (!data->__internal_use && data->__invocation_counter == 0) \
+ { \
+ /* Emit the Byte Order Mark. */ \
+ if (outbuf + 2 > outend) \
+ return __GCONV_FULL_OUTPUT; \
+ \
+ *(uint16_t *) outbuf = BOM; \
+ outbuf += 2; \
+ } \
+ swap = ((struct unicode_data *) step->__data)->swap;
+#define EXTRA_LOOP_ARGS , data, swap
+
+
+/* Direction of the transformation. */
+enum direction
+{
+ illegal_dir,
+ to_unicode,
+ from_unicode
+};
+
+struct unicode_data
+{
+ enum direction dir;
+ int swap;
+};
+
+
+int
+gconv_init (struct __gconv_step *step)
+{
+ /* Determine which direction. */
+ struct unicode_data *new_data;
+ enum direction dir = illegal_dir;
+ int result;
+
+ if (__strcasecmp (step->__from_name, "UNICODE") == 0)
+ dir = from_unicode;
+ else
+ dir = to_unicode;
+
+ new_data = (struct unicode_data *) malloc (sizeof (struct unicode_data));
+
+ result = __GCONV_NOMEM;
+ if (new_data != NULL)
+ {
+ new_data->dir = dir;
+ new_data->swap = 0;
+ step->__data = new_data;
+
+ if (dir == from_unicode)
+ {
+ 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 UCS2. */
+#define MIN_NEEDED_INPUT MIN_NEEDED_TO
+#define MIN_NEEDED_OUTPUT MIN_NEEDED_FROM
+#define LOOPFCT TO_LOOP
+#define BODY \
+ { \
+ uint32_t c = *((uint32_t *) inptr); \
+ \
+ if (c >= 0x10000) \
+ { \
+ result = __GCONV_ILLEGAL_INPUT; \
+ break; \
+ } \
+ \
+ *((uint16_t *) outptr) = c; \
+ \
+ outptr += 2; \
+ inptr += 4; \
+ }
+#define EXTRA_LOOP_DECLS \
+ , struct __gconv_step_data *step_data, int swap
+#include <iconv/loop.c>
+
+
+/* Convert from UCS2 to the internal (UCS4-like) format. */
+#define MIN_NEEDED_INPUT MIN_NEEDED_FROM
+#define MIN_NEEDED_OUTPUT MIN_NEEDED_TO
+#define LOOPFCT FROM_LOOP
+#define BODY \
+ { \
+ uint16_t u1 = *(uint16_t *) inptr; \
+ \
+ if (swap) \
+ u1 = bswap_16 (u1); \
+ \
+ *((uint32_t *) outptr) = u1; \
+ \
+ inptr += 2; \
+ outptr += 4; \
+ }
+#define EXTRA_LOOP_DECLS \
+ , struct __gconv_step_data *step_data, int swap
+#include <iconv/loop.c>
+
+
+/* Now define the toplevel functions. */
+#include <iconv/skeleton.c>
diff --git a/iconvdata/utf-16.c b/iconvdata/utf-16.c
index 55540c9849..c7e677e376 100644
--- a/iconvdata/utf-16.c
+++ b/iconvdata/utf-16.c
@@ -27,6 +27,8 @@
/* 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. */
@@ -41,8 +43,27 @@
#define PREPARE_LOOP \
enum direction dir = ((struct utf16_data *) step->__data)->dir; \
enum variant var = ((struct utf16_data *) step->__data)->var; \
- if (!FROM_DIRECTION && var == UTF_16 && !data->__internal_use \
- && data->__invocation_counter == 0) \
+ 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 > inbufend) \
+ return __GCONV_EMPTY_INPUT; \
+ \
+ if (*(uint16_t *) inptr == BOM) \
+ /* Simply ignore the BOM character. */ \
+ inptr += 2; \
+ else if (*(uint16_t *) 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) \
@@ -51,7 +72,7 @@
*(uint16_t *) outbuf = BOM; \
outbuf += 2; \
}
-#define EXTRA_LOOP_ARGS , var, data
+#define EXTRA_LOOP_ARGS , var, data, swap
/* Direction of the transformation. */
@@ -74,6 +95,7 @@ struct utf16_data
{
enum direction dir;
enum variant var;
+ int swap;
};
@@ -127,6 +149,9 @@ gconv_init (struct __gconv_step *step)
{
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)
@@ -170,8 +195,7 @@ gconv_end (struct __gconv_step *data)
{ \
uint32_t c = *((uint32_t *) inptr); \
\
- if ((__BYTE_ORDER == __LITTLE_ENDIAN && var == UTF_16BE) \
- || (__BYTE_ORDER == __BIG_ENDIAN && var == UTF_16LE)) \
+ if (swap) \
{ \
if (c >= 0x10000) \
{ \
@@ -225,7 +249,7 @@ gconv_end (struct __gconv_step *data)
inptr += 4; \
}
#define EXTRA_LOOP_DECLS \
- , enum variant var, struct __gconv_step_data *step_data
+ , enum variant var, struct __gconv_step_data *step_data, int swap
#include <iconv/loop.c>
@@ -238,8 +262,7 @@ gconv_end (struct __gconv_step *data)
{ \
uint16_t u1 = *(uint16_t *) inptr; \
\
- if ((__BYTE_ORDER == __LITTLE_ENDIAN && var == UTF_16BE) \
- || (__BYTE_ORDER == __BIG_ENDIAN && var == UTF_16LE)) \
+ if (swap) \
{ \
u1 = bswap_16 (u1); \
\
@@ -277,16 +300,6 @@ gconv_end (struct __gconv_step *data)
} \
else \
{ \
- if (u1 == BOM && var == UTF_16 && !step_data->__internal_use \
- && step_data->__invocation_counter == 0 && inptr == *inptrp) \
- { \
- /* This is the first word in the file and it is the BOM and \
- we are converting a file without specified byte order. \
- Simply sack the BOM. */ \
- inptr += 2; \
- continue; \
- } \
- \
if (u1 < 0xd800 || u1 > 0xdfff) \
{ \
/* No surrogate. */ \
@@ -322,7 +335,7 @@ gconv_end (struct __gconv_step *data)
outptr += 4; \
}
#define EXTRA_LOOP_DECLS \
- , enum variant var, struct __gconv_step_data *step_data
+ , enum variant var, struct __gconv_step_data *step_data, int swap
#include <iconv/loop.c>