/* Code conversion functions.
Copyright (C) 1991, 1995 Free Software Foundation, Inc.
Copyright (C) 1995 Sun Microsystems, Inc.
- Copyright (C) 1999,2000 MORIOKA Tomohiko
+ Copyright (C) 1999,2000,2001,2002 MORIOKA Tomohiko
This file is part of XEmacs.
/* Synched up with: Mule 2.3. Not in FSF. */
/* Rewritten by Ben Wing <ben@xemacs.org>. */
+/* Rewritten by MORIOKA Tomohiko <tomo@m17n.org> for XEmacs UTF-2000. */
#include <config.h>
#include "lisp.h"
Lisp_Object Vcoding_system_for_write;
Lisp_Object Vfile_name_coding_system;
+Lisp_Object Vcoded_charset_entity_reference_alist;
+
/* Table of symbols identifying each coding category. */
-Lisp_Object coding_category_symbol[CODING_CATEGORY_LAST + 1];
+Lisp_Object coding_category_symbol[CODING_CATEGORY_LAST];
struct file_coding_dump {
/* Coding system currently associated with each coding category. */
- Lisp_Object coding_category_system[CODING_CATEGORY_LAST + 1];
+ Lisp_Object coding_category_system[CODING_CATEGORY_LAST];
/* Table of all coding categories in decreasing order of priority.
This describes a permutation of the possible coding categories. */
- int coding_category_by_priority[CODING_CATEGORY_LAST + 1];
+ int coding_category_by_priority[CODING_CATEGORY_LAST];
#if defined(MULE) && !defined(UTF2000)
Lisp_Object ucs_to_mule_table[65536];
} *fcd;
static const struct lrecord_description fcd_description_1[] = {
- { XD_LISP_OBJECT_ARRAY, offsetof (struct file_coding_dump, coding_category_system), CODING_CATEGORY_LAST + 1 },
+ { XD_LISP_OBJECT_ARRAY, offsetof (struct file_coding_dump, coding_category_system), CODING_CATEGORY_LAST },
#if defined(MULE) && !defined(UTF2000)
{ XD_LISP_OBJECT_ARRAY, offsetof (struct file_coding_dump, ucs_to_mule_table), countof (fcd->ucs_to_mule_table) },
#endif
Lisp_Object Qpre_write_conversion;
#ifdef MULE
-Lisp_Object Qucs4, Qutf8;
+Lisp_Object Qucs4, Qutf16, Qutf8;
Lisp_Object Qbig5, Qshift_jis;
Lisp_Object Qcharset_g0, Qcharset_g1, Qcharset_g2, Qcharset_g3;
Lisp_Object Qforce_g0_on_output, Qforce_g1_on_output;
Lisp_Object Qshort, Qno_ascii_eol, Qno_ascii_cntl, Qseven, Qlock_shift;
#endif
#ifdef UTF2000
+Lisp_Object Qutf_8_mcs;
Lisp_Object Qdisable_composition;
+Lisp_Object Quse_entity_reference;
+Lisp_Object Qd, Qx, QX;
#endif
Lisp_Object Qencode, Qdecode;
static void
text_encode_generic (Lstream *encoding, const Bufbyte *src,
- unsigned_char_dynarr *dst, size_t n);
+ unsigned_char_dynarr *dst, Lstream_data_count n);
static int detect_coding_sjis (struct detection_state *st,
- const Extbyte *src, size_t n);
+ const Extbyte *src, Lstream_data_count n);
static void decode_coding_sjis (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n);
+ unsigned_char_dynarr *dst, Lstream_data_count n);
void char_encode_shift_jis (struct encoding_stream *str, Emchar c,
unsigned_char_dynarr *dst, unsigned int *flags);
void char_finish_shift_jis (struct encoding_stream *str,
unsigned_char_dynarr *dst, unsigned int *flags);
static int detect_coding_big5 (struct detection_state *st,
- const Extbyte *src, size_t n);
+ const Extbyte *src, Lstream_data_count n);
static void decode_coding_big5 (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n);
-static void encode_coding_big5 (Lstream *encoding, const Bufbyte *src,
- unsigned_char_dynarr *dst, size_t n);
+ unsigned_char_dynarr *dst, Lstream_data_count n);
+void char_encode_big5 (struct encoding_stream *str, Emchar c,
+ unsigned_char_dynarr *dst, unsigned int *flags);
+void char_finish_big5 (struct encoding_stream *str,
+ unsigned_char_dynarr *dst, unsigned int *flags);
static int detect_coding_ucs4 (struct detection_state *st,
- const Extbyte *src, size_t n);
+ const Extbyte *src, Lstream_data_count n);
static void decode_coding_ucs4 (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n);
+ unsigned_char_dynarr *dst, Lstream_data_count n);
void char_encode_ucs4 (struct encoding_stream *str, Emchar c,
unsigned_char_dynarr *dst, unsigned int *flags);
void char_finish_ucs4 (struct encoding_stream *str,
unsigned_char_dynarr *dst, unsigned int *flags);
+static int detect_coding_utf16 (struct detection_state *st,
+ const Extbyte *src, Lstream_data_count n);
+static void decode_coding_utf16 (Lstream *decoding, const Extbyte *src,
+ unsigned_char_dynarr *dst, Lstream_data_count n);
+void char_encode_utf16 (struct encoding_stream *str, Emchar c,
+ unsigned_char_dynarr *dst, unsigned int *flags);
+void char_finish_utf16 (struct encoding_stream *str,
+ unsigned_char_dynarr *dst, unsigned int *flags);
+
static int detect_coding_utf8 (struct detection_state *st,
- const Extbyte *src, size_t n);
+ const Extbyte *src, Lstream_data_count n);
static void decode_coding_utf8 (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n);
+ unsigned_char_dynarr *dst, Lstream_data_count n);
void char_encode_utf8 (struct encoding_stream *str, Emchar c,
unsigned_char_dynarr *dst, unsigned int *flags);
void char_finish_utf8 (struct encoding_stream *str,
static void reset_iso2022 (Lisp_Object coding_system,
struct iso2022_decoder *iso);
static int detect_coding_iso2022 (struct detection_state *st,
- const Extbyte *src, size_t n);
+ const Extbyte *src, Lstream_data_count n);
static void decode_coding_iso2022 (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n);
+ unsigned_char_dynarr *dst, Lstream_data_count n);
void char_encode_iso2022 (struct encoding_stream *str, Emchar c,
unsigned_char_dynarr *dst, unsigned int *flags);
void char_finish_iso2022 (struct encoding_stream *str,
unsigned_char_dynarr *dst, unsigned int *flags);
#endif /* MULE */
static void decode_coding_no_conversion (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n);
+ unsigned_char_dynarr *dst, Lstream_data_count n);
static void encode_coding_no_conversion (Lstream *encoding, const Bufbyte *src,
- unsigned_char_dynarr *dst, size_t n);
+ unsigned_char_dynarr *dst, Lstream_data_count n);
static void mule_decode (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n);
+ unsigned_char_dynarr *dst, Lstream_data_count n);
static void mule_encode (Lstream *encoding, const Bufbyte *src,
- unsigned_char_dynarr *dst, size_t n);
+ unsigned_char_dynarr *dst, Lstream_data_count n);
typedef struct codesys_prop codesys_prop;
struct codesys_prop
{ XD_STRUCT_PTR, offsetof (Lisp_Coding_System, iso2022.output_conv), 1, &ccsd_description },
{ XD_LISP_OBJECT, offsetof (Lisp_Coding_System, ccl.decode) },
{ XD_LISP_OBJECT, offsetof (Lisp_Coding_System, ccl.encode) },
+#ifdef UTF2000
+ { XD_LISP_OBJECT, offsetof (Lisp_Coding_System, ccs_priority_list) },
+#endif
#endif
{ XD_END }
};
}
}
break;
+#ifdef UTF2000
+
+ case CODESYS_BIG5:
+ mark_object (CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 0));
+ mark_object (CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 1));
+ break;
+#endif
case CODESYS_CCL:
mark_object (CODING_SYSTEM_CCL_DECODE (codesys));
}
mark_object (CODING_SYSTEM_PRE_WRITE_CONVERSION (codesys));
+#ifdef UTF2000
+ mark_object (CODING_SYSTEM_CCS_PRIORITY_LIST (codesys));
+#endif
return CODING_SYSTEM_POST_READ_CONVERSION (codesys);
}
CODING_SYSTEM_TYPE (codesys) = type;
CODING_SYSTEM_MNEMONIC (codesys) = Qnil;
#ifdef MULE
+#ifdef UTF2000
+ CODING_SYSTEM_CCS_PRIORITY_LIST (codesys) = Qnil;
+#endif
if (type == CODESYS_ISO2022)
{
int i;
for (i = 0; i < 4; i++)
CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i) = Qnil;
}
+#ifdef UTF2000
+ if (type == CODESYS_UTF8)
+ {
+ CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 0)
+ = Vcharset_ucs;
+ CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 1)
+ = Qnil;
+ CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 2)
+ = Qnil;
+ CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 3)
+ = Qnil;
+ }
+ else if (type == CODESYS_BIG5)
+ {
+ CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 0)
+ = Vcharset_ascii;
+ CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 1)
+ = Vcharset_chinese_big5;
+ CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 2)
+ = Qnil;
+ CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 3)
+ = Qnil;
+ }
+#endif
else if (type == CODESYS_CCL)
{
CODING_SYSTEM_CCL_DECODE (codesys) = Qnil;
If non-nil, composition/decomposition for combining characters
are disabled.
+'use-entity-reference
+ If non-nil, SGML style entity-reference is used for non-system-characters.
+
'post-read-conversion
Function called after a file has been read in, to perform the
decoding. Called with two arguments, START and END, denoting
else if (EQ (type, Qiso2022)) { ty = CODESYS_ISO2022; }
else if (EQ (type, Qbig5)) { ty = CODESYS_BIG5; }
else if (EQ (type, Qucs4)) { ty = CODESYS_UCS4; }
+ else if (EQ (type, Qutf16)) { ty = CODESYS_UTF16; }
else if (EQ (type, Qutf8)) { ty = CODESYS_UTF8; }
else if (EQ (type, Qccl)) { ty = CODESYS_CCL; }
#endif
#ifdef UTF2000
else if (EQ (key, Qdisable_composition))
CODING_SYSTEM_DISABLE_COMPOSITION (codesys) = !NILP (value);
+ else if (EQ (key, Quse_entity_reference))
+ CODING_SYSTEM_USE_ENTITY_REFERENCE (codesys) = !NILP (value);
#endif
#ifdef MULE
else if (ty == CODESYS_ISO2022)
else
signal_simple_error ("Unrecognized property", key);
}
+#ifdef UTF2000
+ else if (ty == CODESYS_UTF8)
+ {
+ if (EQ (key, Qcharset_g0)) FROB_INITIAL_CHARSET (0);
+ else if (EQ (key, Qcharset_g1))
+ CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 1) = value;
+ else if (EQ (key, Qcharset_g2))
+ CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, 2) = value;
+ else
+ signal_simple_error ("Unrecognized property", key);
+ }
+ else if (ty == CODESYS_BIG5)
+ {
+ if (EQ (key, Qcharset_g0)) FROB_INITIAL_CHARSET (0);
+ else if (EQ (key, Qcharset_g1)) FROB_INITIAL_CHARSET (1);
+ else
+ signal_simple_error ("Unrecognized property", key);
+ }
+#endif
else if (EQ (type, Qccl))
{
Lisp_Object sym;
case CODESYS_ISO2022: return Qiso2022;
case CODESYS_BIG5: return Qbig5;
case CODESYS_UCS4: return Qucs4;
+ case CODESYS_UTF16: return Qutf16;
case CODESYS_UTF8: return Qutf8;
case CODESYS_CCL: return Qccl;
#endif
else if (EQ (prop, Qpre_write_conversion))
return XCODING_SYSTEM_PRE_WRITE_CONVERSION (coding_system);
#ifdef MULE
+#ifdef UTF2000
+ else if (EQ (prop, Qdisable_composition))
+ return XCODING_SYSTEM_DISABLE_COMPOSITION (coding_system) ? Qt : Qnil;
+ else if (EQ (prop, Quse_entity_reference))
+ return XCODING_SYSTEM_USE_ENTITY_REFERENCE (coding_system) ? Qt : Qnil;
+#endif
else if (type == CODESYS_ISO2022)
{
if (EQ (prop, Qcharset_g0))
int i;
CHECK_SYMBOL (symbol);
- for (i = 0; i <= CODING_CATEGORY_LAST; i++)
+ for (i = 0; i < CODING_CATEGORY_LAST; i++)
if (EQ (coding_category_symbol[i], symbol))
return i;
int i;
Lisp_Object list = Qnil;
- for (i = CODING_CATEGORY_LAST; i >= 0; i--)
+ for (i = CODING_CATEGORY_LAST - 1; i >= 0; i--)
list = Fcons (coding_category_symbol[i], list);
return list;
}
*/
(list))
{
- int category_to_priority[CODING_CATEGORY_LAST + 1];
+ int category_to_priority[CODING_CATEGORY_LAST];
int i, j;
Lisp_Object rest;
/* First generate a list that maps coding categories to priorities. */
- for (i = 0; i <= CODING_CATEGORY_LAST; i++)
+ for (i = 0; i < CODING_CATEGORY_LAST; i++)
category_to_priority[i] = -1;
/* Highest priority comes from the specified list. */
/* Now go through the existing categories by priority to retrieve
the categories not yet specified and preserve their priority
order. */
- for (j = 0; j <= CODING_CATEGORY_LAST; j++)
+ for (j = 0; j < CODING_CATEGORY_LAST; j++)
{
int cat = fcd->coding_category_by_priority[j];
if (category_to_priority[cat] < 0)
/* Now we need to construct the inverse of the mapping we just
constructed. */
- for (i = 0; i <= CODING_CATEGORY_LAST; i++)
+ for (i = 0; i < CODING_CATEGORY_LAST; i++)
fcd->coding_category_by_priority[category_to_priority[i]] = i;
/* Phew! That was confusing. */
int i;
Lisp_Object list = Qnil;
- for (i = CODING_CATEGORY_LAST; i >= 0; i--)
+ for (i = CODING_CATEGORY_LAST - 1; i >= 0; i--)
list = Fcons (coding_category_symbol[fcd->coding_category_by_priority[i]],
list);
return list;
int mask;
int in_byte;
}
+ utf16;
+
+ struct
+ {
+ int mask;
+ int in_byte;
+ }
utf8;
struct
static eol_type_t
detect_eol_type (struct detection_state *st, const Extbyte *src,
- size_t n)
+ Lstream_data_count n)
{
while (n--)
{
static int
detect_coding_type (struct detection_state *st, const Extbyte *src,
- size_t n, int just_do_eol)
+ Lstream_data_count n, int just_do_eol)
{
if (st->eol_type == EOL_AUTODETECT)
st->eol_type = detect_eol_type (st, src, n);
st->shift_jis.mask = ~0;
st->big5.mask = ~0;
st->ucs4.mask = ~0;
+ st->utf16.mask = ~0;
st->utf8.mask = ~0;
st->iso2022.mask = ~0;
#endif
st->big5.mask = detect_coding_big5 (st, src, n);
if (!mask_has_at_most_one_bit_p (st->utf8.mask))
st->utf8.mask = detect_coding_utf8 (st, src, n);
+ if (!mask_has_at_most_one_bit_p (st->utf16.mask))
+ st->utf16.mask = detect_coding_utf16 (st, src, n);
if (!mask_has_at_most_one_bit_p (st->ucs4.mask))
st->ucs4.mask = detect_coding_ucs4 (st, src, n);
#endif
/* Look through the coding categories by priority and find
the first one that is allowed. */
- for (i = 0; i <= CODING_CATEGORY_LAST; i++)
+ for (i = 0; i < CODING_CATEGORY_LAST; i++)
{
cat = fcd->coding_category_by_priority[i];
if ((mask & (1 << cat)) &&
that should be unnecessary with the explicit eol-type argument. */
#define LENGTH(string_constant) (sizeof (string_constant) - 1)
+/* number of leading lines to check for a coding cookie */
+#define LINES_TO_CHECK 2
void
determine_real_coding_system (Lstream *stream, Lisp_Object *codesys_in_out,
Extbyte buf[4096];
Lisp_Object coding_system = Qnil;
Extbyte *p;
- ssize_t nread = Lstream_read (stream, buf, sizeof (buf));
+ Lstream_data_count nread = Lstream_read (stream, buf, sizeof (buf));
Extbyte *scan_end;
+ int lines_checked = 0;
/* Look for initial "-*-"; mode line prefix */
for (p = buf,
scan_end = buf + nread - LENGTH ("-*-coding:?-*-");
p <= scan_end
- && *p != '\n'
- && *p != '\r';
+ && lines_checked < LINES_TO_CHECK;
p++)
if (*p == '-' && *(p+1) == '*' && *(p+2) == '-')
{
for (p = local_vars_beg,
scan_end = buf + nread - LENGTH ("-*-");
p <= scan_end
- && *p != '\n'
- && *p != '\r';
+ && lines_checked < LINES_TO_CHECK;
p++)
if (*p == '-' && *(p+1) == '*' && *(p+2) == '-')
{
}
break;
}
+ /* #### file must use standard EOLs or we miss 2d line */
+ /* #### not to mention this is broken for UTF-16 DOS files */
+ else if (*p == '\n' || *p == '\r')
+ {
+ lines_checked++;
+ /* skip past multibyte (DOS) newline */
+ if (*p == '\r' && *(p+1) == '\n') p++;
+ }
break;
}
+ /* #### file must use standard EOLs or we miss 2d line */
+ /* #### not to mention this is broken for UTF-16 DOS files */
+ else if (*p == '\n' || *p == '\r')
+ {
+ lines_checked++;
+ /* skip past multibyte (DOS) newline */
+ if (*p == '\r' && *(p+1) == '\n') p++;
+ }
if (NILP (coding_system))
do
while (1)
{
Extbyte random_buffer[4096];
- ssize_t nread = Lstream_read (istr, random_buffer, sizeof (random_buffer));
+ Lstream_data_count nread = Lstream_read (istr, random_buffer, sizeof (random_buffer));
if (!nread)
break;
#ifdef MULE
decst.mask = postprocess_iso2022_mask (decst.mask);
#endif
- for (i = CODING_CATEGORY_LAST; i >= 0; i--)
+ for (i = CODING_CATEGORY_LAST - 1; i >= 0; i--)
{
int sys = fcd->coding_category_by_priority[i];
if (decst.mask & (1 << sys))
unsigned char counter;
#endif
#ifdef UTF2000
+ unsigned char er_counter;
+ unsigned char er_buf[16];
+
unsigned combined_char_count;
Emchar combined_chars[16];
Lisp_Object combining_table;
struct detection_state decst;
};
-#ifdef UTF2000
-extern Lisp_Object Vcharacter_composition_table;
-
-INLINE_HEADER void
-COMPOSE_FLUSH_CHARS (struct decoding_stream *str, unsigned_char_dynarr* dst);
-INLINE_HEADER void
-COMPOSE_FLUSH_CHARS (struct decoding_stream *str, unsigned_char_dynarr* dst)
-{
- unsigned i;
-
- for (i = 0; i < str->combined_char_count; i++)
- DECODE_ADD_UCS_CHAR (str->combined_chars[i], dst);
- str->combined_char_count = 0;
- str->combining_table = Qnil;
-}
-
-void COMPOSE_ADD_CHAR(struct decoding_stream *str, Emchar character,
- unsigned_char_dynarr* dst);
-void
-COMPOSE_ADD_CHAR(struct decoding_stream *str,
- Emchar character, unsigned_char_dynarr* dst)
-{
- if (CODING_SYSTEM_DISABLE_COMPOSITION (str->codesys))
- DECODE_ADD_UCS_CHAR (character, dst);
- else if (!CHAR_ID_TABLE_P (str->combining_table))
- {
- Lisp_Object ret
- = get_char_id_table (character, Vcharacter_composition_table);
-
- if (NILP (ret))
- DECODE_ADD_UCS_CHAR (character, dst);
- else
- {
- str->combined_chars[0] = character;
- str->combined_char_count = 1;
- str->combining_table = ret;
- }
- }
- else
- {
- Lisp_Object ret
- = get_char_id_table (character, str->combining_table);
-
- if (CHARP (ret))
- {
- Emchar char2 = XCHARVAL (ret);
- ret = get_char_id_table (char2, Vcharacter_composition_table);
- if (NILP (ret))
- {
- DECODE_ADD_UCS_CHAR (char2, dst);
- str->combined_char_count = 0;
- str->combining_table = Qnil;
- }
- else
- {
- str->combined_chars[0] = char2;
- str->combined_char_count = 1;
- str->combining_table = ret;
- }
- }
- else if (CHAR_ID_TABLE_P (ret))
- {
- str->combined_chars[str->combined_char_count++] = character;
- str->combining_table = ret;
- }
- else
- {
- COMPOSE_FLUSH_CHARS (str, dst);
- DECODE_ADD_UCS_CHAR (character, dst);
- }
- }
-}
-#else /* not UTF2000 */
-#define COMPOSE_FLUSH_CHARS(str, dst)
-#define COMPOSE_ADD_CHAR(str, ch, dst) DECODE_ADD_UCS_CHAR (ch, dst)
-#endif /* UTF2000 */
-
-static ssize_t decoding_reader (Lstream *stream,
- unsigned char *data, size_t size);
-static ssize_t decoding_writer (Lstream *stream,
- const unsigned char *data, size_t size);
+static Lstream_data_count decoding_reader (Lstream *stream,
+ unsigned char *data, Lstream_data_count size);
+static Lstream_data_count decoding_writer (Lstream *stream,
+ const unsigned char *data, Lstream_data_count size);
static int decoding_rewinder (Lstream *stream);
static int decoding_seekable_p (Lstream *stream);
static int decoding_flusher (Lstream *stream);
/* Read SIZE bytes of data and store it into DATA. We are a decoding stream
so we read data from the other end, decode it, and store it into DATA. */
-static ssize_t
-decoding_reader (Lstream *stream, unsigned char *data, size_t size)
+static Lstream_data_count
+decoding_reader (Lstream *stream, unsigned char *data, Lstream_data_count size)
{
struct decoding_stream *str = DECODING_STREAM_DATA (stream);
unsigned char *orig_data = data;
- ssize_t read_size;
+ Lstream_data_count read_size;
int error_occurred = 0;
/* We need to interface to mule_decode(), which expects to take some
most SIZE bytes, and delete the data from the runoff. */
if (Dynarr_length (str->runoff) > 0)
{
- size_t chunk = min (size, (size_t) Dynarr_length (str->runoff));
+ Lstream_data_count chunk = min (size, (Lstream_data_count) Dynarr_length (str->runoff));
memcpy (data, Dynarr_atp (str->runoff, 0), chunk);
Dynarr_delete_many (str->runoff, 0, chunk);
data += chunk;
return data - orig_data;
}
-static ssize_t
-decoding_writer (Lstream *stream, const unsigned char *data, size_t size)
+static Lstream_data_count
+decoding_writer (Lstream *stream, const unsigned char *data, Lstream_data_count size)
{
struct decoding_stream *str = DECODING_STREAM_DATA (stream);
- ssize_t retval;
+ Lstream_data_count retval;
/* Decode all our data into the runoff, and then attempt to write
it all out to the other end. Remove whatever chunk we succeeded
str->counter = 0;
#endif /* MULE */
#ifdef UTF2000
+ str->er_counter = 0;
str->combined_char_count = 0;
str->combining_table = Qnil;
#endif
+ if (CODING_SYSTEM_TYPE (str->codesys) == CODESYS_AUTODETECT
+ || CODING_SYSTEM_EOL_TYPE (str->codesys) == EOL_AUTODETECT)
+ {
+ xzero (str->decst);
+ str->decst.eol_type = EOL_AUTODETECT;
+ str->decst.mask = ~0;
+ }
str->flags = str->cpos = 0;
}
static void
mule_decode (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n)
+ unsigned_char_dynarr *dst, Lstream_data_count n)
{
struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
case CODESYS_UCS4:
decode_coding_ucs4 (decoding, src, dst, n);
break;
+ case CODESYS_UTF16:
+ decode_coding_utf16 (decoding, src, dst, n);
+ break;
case CODESYS_UTF8:
decode_coding_utf8 (decoding, src, dst, n);
break;
char tempbuf[1024]; /* some random amount */
Bufpos newpos, even_newer_pos;
Bufpos oldpos = lisp_buffer_stream_startpos (istr);
- ssize_t size_in_bytes = Lstream_read (istr, tempbuf, sizeof (tempbuf));
+ Lstream_data_count size_in_bytes = Lstream_read (istr, tempbuf, sizeof (tempbuf));
if (!size_in_bytes)
break;
#endif /* MULE */
};
-static ssize_t encoding_reader (Lstream *stream, unsigned char *data, size_t size);
-static ssize_t encoding_writer (Lstream *stream, const unsigned char *data,
- size_t size);
+static Lstream_data_count encoding_reader (Lstream *stream, unsigned char *data, Lstream_data_count size);
+static Lstream_data_count encoding_writer (Lstream *stream, const unsigned char *data,
+ Lstream_data_count size);
static int encoding_rewinder (Lstream *stream);
static int encoding_seekable_p (Lstream *stream);
static int encoding_flusher (Lstream *stream);
/* Read SIZE bytes of data and store it into DATA. We are a encoding stream
so we read data from the other end, encode it, and store it into DATA. */
-static ssize_t
-encoding_reader (Lstream *stream, unsigned char *data, size_t size)
+static Lstream_data_count
+encoding_reader (Lstream *stream, unsigned char *data, Lstream_data_count size)
{
struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
unsigned char *orig_data = data;
- ssize_t read_size;
+ Lstream_data_count read_size;
int error_occurred = 0;
/* We need to interface to mule_encode(), which expects to take some
return data - orig_data;
}
-static ssize_t
-encoding_writer (Lstream *stream, const unsigned char *data, size_t size)
+static Lstream_data_count
+encoding_writer (Lstream *stream, const unsigned char *data, Lstream_data_count size)
{
struct encoding_stream *str = ENCODING_STREAM_DATA (stream);
- ssize_t retval;
+ Lstream_data_count retval;
/* Encode all our data into the runoff, and then attempt to write
it all out to the other end. Remove whatever chunk we succeeded
str->encode_char = &char_encode_utf8;
str->finish = &char_finish_utf8;
break;
+ case CODESYS_UTF16:
+ str->encode_char = &char_encode_utf16;
+ str->finish = &char_finish_utf16;
+ break;
case CODESYS_UCS4:
str->encode_char = &char_encode_ucs4;
str->finish = &char_finish_ucs4;
str->encode_char = &char_encode_shift_jis;
str->finish = &char_finish_shift_jis;
break;
+ case CODESYS_BIG5:
+ str->encode_char = &char_encode_big5;
+ str->finish = &char_finish_big5;
+ break;
default:
break;
}
static void
mule_encode (Lstream *encoding, const Bufbyte *src,
- unsigned_char_dynarr *dst, size_t n)
+ unsigned_char_dynarr *dst, Lstream_data_count n)
{
struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
encode_coding_no_conversion (encoding, src, dst, n);
break;
#ifdef MULE
- case CODESYS_BIG5:
- encode_coding_big5 (encoding, src, dst, n);
- break;
case CODESYS_CCL:
str->ccl.last_block = str->flags & CODING_STATE_END;
/* When applying ccl program to stream, MUST NOT set NULL
char tempbuf[1024]; /* some random amount */
Bufpos newpos, even_newer_pos;
Bufpos oldpos = lisp_buffer_stream_startpos (istr);
- ssize_t size_in_bytes = Lstream_read (istr, tempbuf, sizeof (tempbuf));
+ Lstream_data_count size_in_bytes = Lstream_read (istr, tempbuf, sizeof (tempbuf));
if (!size_in_bytes)
break;
\f
static void
text_encode_generic (Lstream *encoding, const Bufbyte *src,
- unsigned_char_dynarr *dst, size_t n)
+ unsigned_char_dynarr *dst, Lstream_data_count n)
{
unsigned char c;
unsigned char char_boundary;
}
\f
+#ifdef UTF2000
+/************************************************************************/
+/* entity reference */
+/************************************************************************/
+
+INLINE_HEADER void
+decode_flush_er_chars (struct decoding_stream *str, unsigned_char_dynarr* dst);
+INLINE_HEADER void
+decode_flush_er_chars (struct decoding_stream *str, unsigned_char_dynarr* dst)
+{
+ if ( str->er_counter > 0)
+ {
+ Dynarr_add_many (dst, str->er_buf, str->er_counter);
+ str->er_counter = 0;
+ }
+}
+
+void decode_add_er_char (struct decoding_stream *str, Emchar character,
+ unsigned_char_dynarr* dst);
+void
+decode_add_er_char (struct decoding_stream *str, Emchar c,
+ unsigned_char_dynarr* dst)
+{
+ if (str->er_counter == 0)
+ {
+ if (CODING_SYSTEM_USE_ENTITY_REFERENCE (str->codesys)
+ && (c == '&') )
+ {
+ str->er_buf[0] = '&';
+ str->er_counter++;
+ }
+ else
+ DECODE_ADD_UCS_CHAR (c, dst);
+ }
+ else if (c == ';')
+ {
+ Lisp_Object string = make_string (str->er_buf,
+ str->er_counter);
+ Lisp_Object rest = Vcoded_charset_entity_reference_alist;
+ Lisp_Object cell;
+ Lisp_Object ret;
+ Lisp_Object pat;
+ Lisp_Object ccs;
+ Lisp_Object char_type;
+ int base;
+
+ while (!NILP (rest))
+ {
+ cell = Fcar (rest);
+ ccs = Fcar (cell);
+ if (CONSP (ccs))
+ {
+ char_type = XCDR (ccs);
+ ccs = XCAR (ccs);
+ }
+ else
+ char_type = Qnil;
+ if (NILP (ccs = Ffind_charset (ccs)))
+ continue;
+
+ cell = Fcdr (cell);
+ ret = Fcar (cell);
+ if (STRINGP (ret))
+ pat = ret;
+ else
+ continue;
+
+ cell = Fcdr (cell);
+ cell = Fcdr (cell);
+ ret = Fcar (cell);
+ if (EQ (ret, Qd))
+ {
+ pat = concat3 (build_string ("^&"),
+ pat, build_string ("\\([0-9]+\\)$"));
+ base = 10;
+ }
+ else if (EQ (ret, Qx))
+ {
+ pat = concat3 (build_string ("^&"),
+ pat, build_string ("\\([0-9a-f]+\\)$"));
+ base = 16;
+ }
+ else if (EQ (ret, QX))
+ {
+ pat = concat3 (build_string ("^&"),
+ pat, build_string ("\\([0-9A-F]+\\)$"));
+ base = 16;
+ }
+ else
+ continue;
+
+ if (!NILP (Fstring_match (pat, string, Qnil, Qnil)))
+ {
+ int code
+ = XINT (Fstring_to_number
+ (Fsubstring (string,
+ Fmatch_beginning (make_int (1)),
+ Fmatch_end (make_int (1))),
+ make_int (base)));
+ Emchar chr
+ = NILP (char_type)
+ ? DECODE_CHAR (ccs, code)
+ : decode_builtin_char (ccs, code);
+
+ DECODE_ADD_UCS_CHAR (chr, dst);
+ goto decoded;
+ }
+ rest = Fcdr (rest);
+ }
+ if (!NILP (Fstring_match (build_string ("^&MCS-\\([0-9A-F]+\\)$"),
+ string, Qnil, Qnil)))
+ {
+ int code
+ = XINT (Fstring_to_number
+ (Fsubstring (string,
+ Fmatch_beginning (make_int (1)),
+ Fmatch_end (make_int (1))),
+ make_int (16)));
+
+ DECODE_ADD_UCS_CHAR (code, dst);
+ }
+ else
+ {
+ Dynarr_add_many (dst, str->er_buf, str->er_counter);
+ Dynarr_add (dst, ';');
+ }
+ decoded:
+ str->er_counter = 0;
+ }
+ else if ( (str->er_counter >= 16) || (c >= 0x7F) )
+ {
+ Dynarr_add_many (dst, str->er_buf, str->er_counter);
+ str->er_counter = 0;
+ DECODE_ADD_UCS_CHAR (c, dst);
+ }
+ else
+ str->er_buf[str->er_counter++] = c;
+}
+
+void char_encode_as_entity_reference (Emchar ch, char* buf);
+void
+char_encode_as_entity_reference (Emchar ch, char* buf)
+{
+ Lisp_Object rest = Vcoded_charset_entity_reference_alist;
+ Lisp_Object cell;
+ Lisp_Object ccs;
+ Lisp_Object char_type;
+ int format_columns, idx;
+ char format[18];
+
+ while (!NILP (rest))
+ {
+ cell = Fcar (rest);
+ ccs = Fcar (cell);
+ if (CONSP (ccs))
+ {
+ char_type = XCDR (ccs);
+ ccs = XCAR (ccs);
+ }
+ else
+ char_type = Qnil;
+ if (!NILP (ccs = Ffind_charset (ccs)))
+ {
+ int code_point = charset_code_point (ccs, ch, 0);
+
+ if ( (code_point >= 0)
+ && (NILP (char_type)
+ || DECODE_CHAR (ccs, code_point) != ch) )
+ {
+ Lisp_Object ret;
+
+ cell = Fcdr (cell);
+ ret = Fcar (cell);
+ if (STRINGP (ret) && ((idx = XSTRING_LENGTH (ret)) <= 6))
+ {
+ format[0] = '&';
+ strncpy (&format[1], XSTRING_DATA (ret), idx);
+ idx++;
+ }
+ else
+ continue;
+
+ cell = Fcdr (cell);
+ ret = Fcar (cell);
+ if (INTP (ret))
+ {
+ format[idx++] = '%';
+ format_columns = XINT (ret);
+ if ( (2 <= format_columns) && (format_columns <= 8) )
+ {
+ format [idx++] = '0';
+ format [idx++] = '0' + format_columns;
+ }
+ }
+
+ cell = Fcdr (cell);
+ ret = Fcar (cell);
+ if (EQ (ret, Qd))
+ format [idx++] = 'd';
+ else if (EQ (ret, Qx))
+ format [idx++] = 'x';
+ else if (EQ (ret, QX))
+ format [idx++] = 'X';
+ else
+ continue;
+ format [idx++] = ';';
+ format [idx++] = 0;
+
+ sprintf (buf, format, code_point);
+ return;
+ }
+ }
+ rest = Fcdr (rest);
+ }
+ sprintf (buf, "&MCS-%08X;", ch);
+}
+
+\f
+/************************************************************************/
+/* character composition */
+/************************************************************************/
+extern Lisp_Object Qcomposition;
+
+INLINE_HEADER void
+COMPOSE_FLUSH_CHARS (struct decoding_stream *str, unsigned_char_dynarr* dst);
+INLINE_HEADER void
+COMPOSE_FLUSH_CHARS (struct decoding_stream *str, unsigned_char_dynarr* dst)
+{
+ unsigned i;
+
+ for (i = 0; i < str->combined_char_count; i++)
+ decode_add_er_char (str, str->combined_chars[i], dst);
+ str->combined_char_count = 0;
+ str->combining_table = Qnil;
+}
+
+void COMPOSE_ADD_CHAR (struct decoding_stream *str, Emchar character,
+ unsigned_char_dynarr* dst);
+void
+COMPOSE_ADD_CHAR (struct decoding_stream *str,
+ Emchar character, unsigned_char_dynarr* dst)
+{
+ if (CODING_SYSTEM_DISABLE_COMPOSITION (str->codesys))
+ decode_add_er_char (str, character, dst);
+ else if (!CONSP (str->combining_table))
+ {
+ Lisp_Object ret
+ = Fget_char_attribute (make_char (character), Qcomposition, Qnil);
+
+ if (NILP (ret))
+ decode_add_er_char (str, character, dst);
+ else
+ {
+ str->combined_chars[0] = character;
+ str->combined_char_count = 1;
+ str->combining_table = ret;
+ }
+ }
+ else
+ {
+ Lisp_Object ret
+ = Fcdr (Fassq (make_char (character), str->combining_table));
+
+ if (CHARP (ret))
+ {
+ Emchar char2 = XCHARVAL (ret);
+ ret = Fget_char_attribute (make_char (character), Qcomposition,
+ Qnil);
+ if (NILP (ret))
+ {
+ decode_add_er_char (str, character, dst);
+ str->combined_char_count = 0;
+ str->combining_table = Qnil;
+ }
+ else
+ {
+ str->combined_chars[0] = char2;
+ str->combined_char_count = 1;
+ str->combining_table = ret;
+ }
+ }
+ else
+ {
+ COMPOSE_FLUSH_CHARS (str, dst);
+ decode_add_er_char (str, character, dst);
+ }
+ }
+}
+#else /* not UTF2000 */
+#define COMPOSE_FLUSH_CHARS(str, dst)
+#define COMPOSE_ADD_CHAR(str, ch, dst) DECODE_ADD_UCS_CHAR (ch, dst)
+#endif /* UTF2000 */
+
+\f
/************************************************************************/
/* Shift-JIS methods */
/************************************************************************/
((c) >= 0xA1 && (c) <= 0xDF)
static int
-detect_coding_sjis (struct detection_state *st, const Extbyte *src, size_t n)
+detect_coding_sjis (struct detection_state *st, const Extbyte *src, Lstream_data_count n)
{
while (n--)
{
static void
decode_coding_sjis (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n)
+ unsigned_char_dynarr *dst, Lstream_data_count n)
{
struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
unsigned int flags = str->flags;
{
unsigned int s1, s2;
#ifdef UTF2000
- int code_point = charset_code_point (Vcharset_latin_jisx0201, ch);
+ int code_point = charset_code_point (Vcharset_latin_jisx0201, ch, 0);
if (code_point >= 0)
Dynarr_add (dst, code_point);
else if ((code_point
- = charset_code_point (Vcharset_japanese_jisx0208_1990, ch))
+ = charset_code_point (Vcharset_japanese_jisx0208_1990, ch, 0))
>= 0)
{
ENCODE_SJIS ((code_point >> 8) | 0x80,
Dynarr_add (dst, s2);
}
else if ((code_point
- = charset_code_point (Vcharset_katakana_jisx0201, ch))
+ = charset_code_point (Vcharset_katakana_jisx0201, ch, 0))
>= 0)
Dynarr_add (dst, code_point | 0x80);
else if ((code_point
- = charset_code_point (Vcharset_japanese_jisx0208, ch))
+ = charset_code_point (Vcharset_japanese_jisx0208, ch, 0))
>= 0)
{
ENCODE_SJIS ((code_point >> 8) | 0x80,
Dynarr_add (dst, s1);
Dynarr_add (dst, s2);
}
- else if ((code_point = charset_code_point (Vcharset_ascii, ch))
+ else if ((code_point = charset_code_point (Vcharset_ascii, ch, 0))
>= 0)
Dynarr_add (dst, code_point);
else
contains frequently used characters and the latter contains less
frequently used characters. */
+#ifdef UTF2000
+#define BYTE_BIG5_TWO_BYTE_1_P(c) \
+ ((c) >= 0x81 && (c) <= 0xFE)
+#else
#define BYTE_BIG5_TWO_BYTE_1_P(c) \
((c) >= 0xA1 && (c) <= 0xFE)
+#endif
/* Is this the second byte of a Shift-JIS two-byte char? */
} while (0)
static int
-detect_coding_big5 (struct detection_state *st, const Extbyte *src, size_t n)
+detect_coding_big5 (struct detection_state *st, const Extbyte *src, Lstream_data_count n)
{
while (n--)
{
unsigned char c = *(unsigned char *)src++;
- if (c == ISO_CODE_ESC || c == ISO_CODE_SI || c == ISO_CODE_SO ||
- (c >= 0x80 && c <= 0xA0))
+ if (c == ISO_CODE_ESC || c == ISO_CODE_SI || c == ISO_CODE_SO
+#ifndef UTF2000
+ || (c >= 0x80 && c <= 0xA0)
+#endif
+ )
return 0;
if (st->big5.in_second_byte)
{
if (c < 0x40 || (c >= 0x80 && c <= 0xA0))
return 0;
}
- else if (c >= 0xA1)
+ else if (
+#ifdef UTF2000
+ c >= 0x81
+#else
+ c >= 0xA1
+#endif
+ )
st->big5.in_second_byte = 1;
}
return CODING_CATEGORY_BIG5_MASK;
static void
decode_coding_big5 (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n)
+ unsigned_char_dynarr *dst, Lstream_data_count n)
{
struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
unsigned int flags = str->flags;
unsigned int cpos = str->cpos;
eol_type_t eol_type = str->eol_type;
+#ifdef UTF2000
+ Lisp_Object ccs
+ = CODING_SYSTEM_ISO2022_INITIAL_CHARSET (DECODING_STREAM_DATA
+ (decoding)->codesys, 1);
+#endif
while (n--)
{
if (BYTE_BIG5_TWO_BYTE_2_P (c))
{
#ifdef UTF2000
- DECODE_ADD_UCS_CHAR
- (DECODE_CHAR (Vcharset_chinese_big5, (cpos << 8) | c),
- dst);
+ int code_point = (cpos << 8) | c;
+ Emchar char_id = decode_defined_char (ccs, code_point);
+
+ if (char_id < 0)
+ char_id = DECODE_CHAR (Vcharset_chinese_big5, code_point);
+ DECODE_ADD_UCS_CHAR (char_id, dst);
#else
unsigned char b1, b2, b3;
DECODE_BIG5 (cpos, c, b1, b2, b3);
{
DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
if (BYTE_BIG5_TWO_BYTE_1_P (c))
- cpos = c;
+ {
+ decode_flush_er_chars (str, dst);
+ cpos = c;
+ }
+ else if ( c < ' ' )
+ {
+ decode_flush_er_chars (str, dst);
+ DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
+ DECODE_ADD_BINARY_CHAR (c, dst);
+ }
else
- DECODE_ADD_BINARY_CHAR (c, dst);
+ {
+ /* DECODE_ADD_BINARY_CHAR (c, dst); */
+ decode_add_er_char (str, c, dst);
+ }
}
label_continue_loop:;
}
- DECODE_HANDLE_END_OF_CONVERSION (flags, cpos, dst);
+ /* DECODE_HANDLE_END_OF_CONVERSION (flags, cpos, dst); */
+ if (flags & CODING_STATE_END)
+ {
+ decode_flush_er_chars (str, dst);
+ DECODE_OUTPUT_PARTIAL_CHAR (cpos);
+ if (flags & CODING_STATE_CR)
+ Dynarr_add (dst, '\r');
+ }
str->flags = flags;
str->cpos = cpos;
/* Convert internally-formatted data to Big5. */
-static void
-encode_coding_big5 (Lstream *encoding, const Bufbyte *src,
- unsigned_char_dynarr *dst, size_t n)
+void
+char_encode_big5 (struct encoding_stream *str, Emchar ch,
+ unsigned_char_dynarr *dst, unsigned int *flags)
{
-#ifndef UTF2000
- unsigned char c;
- struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
- unsigned int flags = str->flags;
- unsigned int ch = str->ch;
eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys);
- while (n--)
+ if (ch == '\n')
{
- c = *src++;
- if (c == '\n')
- {
- if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
- Dynarr_add (dst, '\r');
- if (eol_type != EOL_CR)
- Dynarr_add (dst, '\n');
- }
- else if (BYTE_ASCII_P (c))
+ if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT)
+ Dynarr_add (dst, '\r');
+ if (eol_type != EOL_CR)
+ Dynarr_add (dst, ch);
+ }
+ else
+ {
+#ifdef UTF2000
+ int code_point;
+ Lisp_Object ccs
+ = CODING_SYSTEM_ISO2022_INITIAL_CHARSET (str->codesys, 1);
+
+ if ((code_point = charset_code_point (Vcharset_ascii, ch, 0)) >= 0)
+ Dynarr_add (dst, code_point);
+ else if ((code_point = charset_code_point (ccs, ch, 0)) >= 0)
{
- /* ASCII. */
- Dynarr_add (dst, c);
+ Dynarr_add (dst, code_point >> 8);
+ Dynarr_add (dst, code_point & 0xFF);
}
- else if (BUFBYTE_LEADING_BYTE_P (c))
+ else if ((code_point
+ = charset_code_point (Vcharset_chinese_big5, ch, 0)) >= 0)
{
- if (c == LEADING_BYTE_CHINESE_BIG5_1 ||
- c == LEADING_BYTE_CHINESE_BIG5_2)
- {
- /* A recognized leading byte. */
- ch = c;
- continue; /* not done with this character. */
- }
- /* otherwise just ignore this character. */
+ Dynarr_add (dst, code_point >> 8);
+ Dynarr_add (dst, code_point & 0xFF);
}
- else if (ch == LEADING_BYTE_CHINESE_BIG5_1 ||
- ch == LEADING_BYTE_CHINESE_BIG5_2)
+ else if ((code_point
+ = charset_code_point (Vcharset_chinese_big5_1, ch, 0)) >= 0)
{
- /* Previous char was a recognized leading byte. */
- ch = (ch << 8) | c;
- continue; /* not done with this character. */
+ unsigned int I
+ = ((code_point >> 8) - 33) * (0xFF - 0xA1)
+ + ((code_point & 0xFF) - 33);
+ unsigned char b1 = I / BIG5_SAME_ROW + 0xA1;
+ unsigned char b2 = I % BIG5_SAME_ROW;
+
+ b2 += b2 < 0x3F ? 0x40 : 0x62;
+ Dynarr_add (dst, b1);
+ Dynarr_add (dst, b2);
}
- else if (ch)
+ else if ((code_point
+ = charset_code_point (Vcharset_chinese_big5_2, ch, 0)) >= 0)
{
- /* Encountering second byte of a Big5 character. */
+ unsigned int I
+ = ((code_point >> 8) - 33) * (0xFF - 0xA1)
+ + ((code_point & 0xFF) - 33);
unsigned char b1, b2;
- ENCODE_BIG5 (ch >> 8, ch & 0xFF, c, b1, b2);
+ I += BIG5_SAME_ROW * (0xC9 - 0xA1);
+ b1 = I / BIG5_SAME_ROW + 0xA1;
+ b2 = I % BIG5_SAME_ROW;
+ b2 += b2 < 0x3F ? 0x40 : 0x62;
Dynarr_add (dst, b1);
Dynarr_add (dst, b2);
}
+ else if (CODING_SYSTEM_USE_ENTITY_REFERENCE (str->codesys))
+ {
+ char buf[18];
- ch = 0;
+ char_encode_as_entity_reference (ch, buf);
+ Dynarr_add_many (dst, buf, strlen (buf));
+ }
+ else
+ Dynarr_add (dst, '?');
+#else
+#endif
}
+}
- str->flags = flags;
- str->ch = ch;
-#endif
+void
+char_finish_big5 (struct encoding_stream *str, unsigned_char_dynarr *dst,
+ unsigned int *flags)
+{
}
/************************************************************************/
static int
-detect_coding_ucs4 (struct detection_state *st, const Extbyte *src, size_t n)
+detect_coding_ucs4 (struct detection_state *st, const Extbyte *src, Lstream_data_count n)
{
while (n--)
{
static void
decode_coding_ucs4 (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n)
+ unsigned_char_dynarr *dst, Lstream_data_count n)
{
struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
unsigned int flags = str->flags;
\f
/************************************************************************/
+/* UTF-16 methods */
+/************************************************************************/
+
+static int
+detect_coding_utf16 (struct detection_state *st, const Extbyte *src, Lstream_data_count n)
+{
+ return CODING_CATEGORY_UTF16_MASK;
+}
+
+static void
+decode_coding_utf16 (Lstream *decoding, const Extbyte *src,
+ unsigned_char_dynarr *dst, Lstream_data_count n)
+{
+ struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
+ unsigned int flags = str->flags;
+ unsigned int cpos = str->cpos;
+ unsigned char counter = str->counter & 3;
+ unsigned char byte_order = str->counter >> 2;
+ eol_type_t eol_type = str->eol_type;
+
+ while (n--)
+ {
+ unsigned char c = *(unsigned char *)src++;
+ if (counter == 0)
+ {
+ cpos = c;
+ counter = 1;
+ }
+ else if (counter == 1)
+ {
+ int code;
+
+ if (byte_order == 0)
+ code = (c << 8) | cpos;
+ else
+ code = (cpos << 8) | c;
+ if (code == 0xFFFE)
+ {
+ code = ((code & 0xFF) << 8) | (code >> 8);
+ if ( byte_order == 0 )
+ byte_order = 1;
+ else
+ byte_order = 0;
+ }
+ if ( (0xD800 <= code) && (code <= 0xDBFF) )
+ {
+ counter = 2;
+ cpos = code;
+ }
+ else
+ {
+ counter = 0;
+ cpos = 0;
+ if (code != 0xFEFF)
+ {
+ DECODE_HANDLE_EOL_TYPE (eol_type, code, flags, dst);
+ DECODE_ADD_UCS_CHAR (code, dst);
+ }
+ }
+ }
+ else if (counter == 2)
+ {
+ cpos = (cpos << 8) | c;
+ counter++;
+ }
+ else
+ {
+ int x = cpos >> 8;
+ int y
+ = (byte_order == 0)
+ ? (c << 8) | (cpos & 0xFF)
+ : ((cpos & 0xFF) << 8) | c;
+
+ DECODE_ADD_UCS_CHAR ((x - 0xD800) * 0x400 + (y - 0xDC00)
+ + 0x10000, dst);
+ counter = 0;
+ cpos = 0;
+ }
+ label_continue_loop:;
+ }
+ if (counter & CODING_STATE_END)
+ DECODE_OUTPUT_PARTIAL_CHAR (cpos);
+
+ str->flags = flags;
+ str->cpos = cpos;
+ str->counter = (byte_order << 2) | counter;
+}
+
+void
+char_encode_utf16 (struct encoding_stream *str, Emchar ch,
+ unsigned_char_dynarr *dst, unsigned int *flags)
+{
+ if (ch <= 0xFFFF)
+ {
+ Dynarr_add (dst, ch);
+ Dynarr_add (dst, ch >> 8);
+ }
+ else
+ {
+ int y = ((ch - 0x10000) / 0x400) + 0xD800;
+ int z = ((ch - 0x10000) % 0x400) + 0xDC00;
+
+ Dynarr_add (dst, y);
+ Dynarr_add (dst, y >> 8);
+ Dynarr_add (dst, z);
+ Dynarr_add (dst, z >> 8);
+ }
+}
+
+void
+char_finish_utf16 (struct encoding_stream *str, unsigned_char_dynarr *dst,
+ unsigned int *flags)
+{
+}
+
+\f
+/************************************************************************/
/* UTF-8 methods */
/************************************************************************/
static int
-detect_coding_utf8 (struct detection_state *st, const Extbyte *src, size_t n)
+detect_coding_utf8 (struct detection_state *st, const Extbyte *src, Lstream_data_count n)
{
while (n--)
{
static void
decode_coding_utf8 (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n)
+ unsigned_char_dynarr *dst, Lstream_data_count n)
{
struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
unsigned int flags = str->flags;
unsigned int cpos = str->cpos;
eol_type_t eol_type = str->eol_type;
unsigned char counter = str->counter;
+#ifdef UTF2000
+ Lisp_Object ccs
+ = CODING_SYSTEM_ISO2022_INITIAL_CHARSET (DECODING_STREAM_DATA
+ (decoding)->codesys, 0);
+#endif
while (n--)
{
unsigned char c = *(unsigned char *)src++;
if (counter == 0)
{
- if ( c < 0xC0 )
+ if ( c < ' ' )
{
+ COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
DECODE_HANDLE_EOL_TYPE (eol_type, c, flags, dst);
DECODE_ADD_UCS_CHAR (c, dst);
}
- else if ( c < 0xE0 )
- {
- cpos = c & 0x1f;
- counter = 1;
- }
- else if ( c < 0xF0 )
- {
- cpos = c & 0x0f;
- counter = 2;
- }
- else if ( c < 0xF8 )
- {
- cpos = c & 0x07;
- counter = 3;
- }
- else if ( c < 0xFC )
- {
- cpos = c & 0x03;
- counter = 4;
- }
+ else if ( c < 0xC0 )
+ /* decode_add_er_char (str, c, dst); */
+ COMPOSE_ADD_CHAR (str, c, dst);
else
{
- cpos = c & 0x01;
- counter = 5;
+ /* decode_flush_er_chars (str, dst); */
+ if ( c < 0xE0 )
+ {
+ cpos = c & 0x1f;
+ counter = 1;
+ }
+ else if ( c < 0xF0 )
+ {
+ cpos = c & 0x0f;
+ counter = 2;
+ }
+ else if ( c < 0xF8 )
+ {
+ cpos = c & 0x07;
+ counter = 3;
+ }
+ else if ( c < 0xFC )
+ {
+ cpos = c & 0x03;
+ counter = 4;
+ }
+ else
+ {
+ cpos = c & 0x01;
+ counter = 5;
+ }
}
}
else if ( (c & 0xC0) == 0x80 )
cpos = ( cpos << 6 ) | ( c & 0x3f );
if (counter == 1)
{
- DECODE_ADD_UCS_CHAR (cpos, dst);
+ Emchar char_id;
+
+ if (!NILP (ccs))
+ {
+ char_id = decode_defined_char (ccs, cpos);
+
+ if (char_id < 0)
+ char_id = cpos;
+ }
+ else
+ ccs = char_id;
+ COMPOSE_ADD_CHAR (str, char_id, dst);
cpos = 0;
counter = 0;
}
}
else
{
+ COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
decode_output_utf8_partial_char (counter, cpos, dst);
DECODE_ADD_BINARY_CHAR (c, dst);
cpos = 0;
}
if (flags & CODING_STATE_END)
- if (counter > 0)
- {
- decode_output_utf8_partial_char (counter, cpos, dst);
- cpos = 0;
- counter = 0;
- }
+ {
+ COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
+ if (counter > 0)
+ {
+ decode_output_utf8_partial_char (counter, cpos, dst);
+ cpos = 0;
+ counter = 0;
+ }
+ }
str->flags = flags;
str->cpos = cpos;
str->counter = counter;
{
Dynarr_add (dst, ch);
}
- else if (ch <= 0x7ff)
- {
- Dynarr_add (dst, (ch >> 6) | 0xc0);
- Dynarr_add (dst, (ch & 0x3f) | 0x80);
- }
- else if (ch <= 0xffff)
- {
- Dynarr_add (dst, (ch >> 12) | 0xe0);
- Dynarr_add (dst, ((ch >> 6) & 0x3f) | 0x80);
- Dynarr_add (dst, (ch & 0x3f) | 0x80);
- }
- else if (ch <= 0x1fffff)
- {
- Dynarr_add (dst, (ch >> 18) | 0xf0);
- Dynarr_add (dst, ((ch >> 12) & 0x3f) | 0x80);
- Dynarr_add (dst, ((ch >> 6) & 0x3f) | 0x80);
- Dynarr_add (dst, (ch & 0x3f) | 0x80);
- }
- else if (ch <= 0x3ffffff)
- {
- Dynarr_add (dst, (ch >> 24) | 0xf8);
- Dynarr_add (dst, ((ch >> 18) & 0x3f) | 0x80);
- Dynarr_add (dst, ((ch >> 12) & 0x3f) | 0x80);
- Dynarr_add (dst, ((ch >> 6) & 0x3f) | 0x80);
- Dynarr_add (dst, (ch & 0x3f) | 0x80);
- }
else
{
- Dynarr_add (dst, (ch >> 30) | 0xfc);
- Dynarr_add (dst, ((ch >> 24) & 0x3f) | 0x80);
- Dynarr_add (dst, ((ch >> 18) & 0x3f) | 0x80);
- Dynarr_add (dst, ((ch >> 12) & 0x3f) | 0x80);
- Dynarr_add (dst, ((ch >> 6) & 0x3f) | 0x80);
- Dynarr_add (dst, (ch & 0x3f) | 0x80);
+ Lisp_Object ucs_ccs
+ = CODING_SYSTEM_ISO2022_INITIAL_CHARSET (str->codesys, 0);
+ int code_point = charset_code_point (ucs_ccs, ch, 0);
+
+ if ( (code_point < 0) || (code_point > 0x10FFFF) )
+ {
+ Lisp_Object map
+ = CODING_SYSTEM_ISO2022_INITIAL_CHARSET (str->codesys, 1);
+ Lisp_Object ret;
+
+ if ( !NILP (map)
+ && INTP (ret = Fget_char_attribute (make_char (ch),
+ map, Qnil)) )
+ code_point = XINT (ret);
+ else if ( !NILP (map =
+ CODING_SYSTEM_ISO2022_INITIAL_CHARSET
+ (str->codesys, 2))
+ && INTP (ret = Fget_char_attribute (make_char (ch),
+ map, Qnil)) )
+ code_point = XINT (ret);
+ else if (CODING_SYSTEM_USE_ENTITY_REFERENCE (str->codesys))
+ {
+ char buf[18];
+
+ char_encode_as_entity_reference (ch, buf);
+ Dynarr_add_many (dst, buf, strlen (buf));
+ return;
+ }
+ else
+ code_point = ch;
+ }
+ if (code_point <= 0x7ff)
+ {
+ Dynarr_add (dst, (code_point >> 6) | 0xc0);
+ Dynarr_add (dst, (code_point & 0x3f) | 0x80);
+ }
+ else if (code_point <= 0xffff)
+ {
+ Dynarr_add (dst, (code_point >> 12) | 0xe0);
+ Dynarr_add (dst, ((code_point >> 6) & 0x3f) | 0x80);
+ Dynarr_add (dst, (code_point & 0x3f) | 0x80);
+ }
+ else if (code_point <= 0x1fffff)
+ {
+ Dynarr_add (dst, (code_point >> 18) | 0xf0);
+ Dynarr_add (dst, ((code_point >> 12) & 0x3f) | 0x80);
+ Dynarr_add (dst, ((code_point >> 6) & 0x3f) | 0x80);
+ Dynarr_add (dst, (code_point & 0x3f) | 0x80);
+ }
+ else if (code_point <= 0x3ffffff)
+ {
+ Dynarr_add (dst, (code_point >> 24) | 0xf8);
+ Dynarr_add (dst, ((code_point >> 18) & 0x3f) | 0x80);
+ Dynarr_add (dst, ((code_point >> 12) & 0x3f) | 0x80);
+ Dynarr_add (dst, ((code_point >> 6) & 0x3f) | 0x80);
+ Dynarr_add (dst, (code_point & 0x3f) | 0x80);
+ }
+ else
+ {
+ Dynarr_add (dst, (code_point >> 30) | 0xfc);
+ Dynarr_add (dst, ((code_point >> 24) & 0x3f) | 0x80);
+ Dynarr_add (dst, ((code_point >> 18) & 0x3f) | 0x80);
+ Dynarr_add (dst, ((code_point >> 12) & 0x3f) | 0x80);
+ Dynarr_add (dst, ((code_point >> 6) & 0x3f) | 0x80);
+ Dynarr_add (dst, (code_point & 0x3f) | 0x80);
+ }
}
}
}
static int
-detect_coding_iso2022 (struct detection_state *st, const Extbyte *src, size_t n)
+detect_coding_iso2022 (struct detection_state *st, const Extbyte *src, Lstream_data_count n)
{
int mask;
static void
decode_coding_iso2022 (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n)
+ unsigned_char_dynarr *dst, Lstream_data_count n)
{
struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
unsigned int flags = str->flags;
case ISO_ESC_LITERAL:
COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
DECODE_ADD_BINARY_CHAR (c, dst);
break;
/* Output the (possibly invalid) sequence */
int i;
COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
for (i = 0; i < str->iso2022.esc_bytes_index; i++)
DECODE_ADD_BINARY_CHAR (str->iso2022.esc_bytes[i], dst);
flags &= CODING_STATE_ISO2022_LOCK;
escape sequence; it could mess things up anyway.
Just add it now. */
COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
DECODE_ADD_BINARY_CHAR (c, dst);
}
}
if (counter)
{
COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
while (counter > 0)
{
counter--;
if (flags & CODING_STATE_SS2)
{
COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
DECODE_ADD_BINARY_CHAR (ISO_CODE_SS2, dst);
flags &= ~CODING_STATE_SS2;
}
if (flags & CODING_STATE_SS3)
{
COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
DECODE_ADD_BINARY_CHAR (ISO_CODE_SS3, dst);
flags &= ~CODING_STATE_SS3;
}
if (c == '\r')
{
COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
if (eol_type == EOL_CR)
Dynarr_add (dst, '\n');
else if (eol_type != EOL_CRLF || flags & CODING_STATE_CR)
if (!parse_iso2022_esc (coding_system, &str->iso2022, c, &flags, 1))
{
COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
DECODE_ADD_BINARY_CHAR (c, dst);
}
}
if (c == '\r')
{
COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
if (eol_type == EOL_CR)
Dynarr_add (dst, '\n');
else if (eol_type != EOL_CRLF || flags & CODING_STATE_CR)
to preserve it for the output. */
{
COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
while (counter > 0)
{
counter--;
if (flags & CODING_STATE_END)
{
COMPOSE_FLUSH_CHARS (str, dst);
+ decode_flush_er_chars (str, dst);
DECODE_OUTPUT_PARTIAL_CHAR (cpos);
}
str->flags = flags;
for (i = 0; i < 4; i++)
{
if ((CHARSETP (charset = str->iso2022.charset[i])
- && ((code_point = charset_code_point (charset, ch)) >= 0))
+ && ((code_point = charset_code_point (charset, ch, 0)) >= 0))
||
(CHARSETP
(charset
= CODING_SYSTEM_ISO2022_INITIAL_CHARSET (codesys, i))
- && ((code_point = charset_code_point (charset, ch)) >= 0)))
+ && ((code_point = charset_code_point (charset, ch, 0)) >= 0)))
{
reg = i;
break;
interpreted as being in any particular decoding. */
static void
decode_coding_no_conversion (Lstream *decoding, const Extbyte *src,
- unsigned_char_dynarr *dst, size_t n)
+ unsigned_char_dynarr *dst, Lstream_data_count n)
{
struct decoding_stream *str = DECODING_STREAM_DATA (decoding);
unsigned int flags = str->flags;
static void
encode_coding_no_conversion (Lstream *encoding, const Bufbyte *src,
- unsigned_char_dynarr *dst, size_t n)
+ unsigned_char_dynarr *dst, Lstream_data_count n)
{
unsigned char c;
struct encoding_stream *str = ENCODING_STREAM_DATA (encoding);
{
INIT_LRECORD_IMPLEMENTATION (coding_system);
- deferror (&Qcoding_system_error, "coding-system-error",
- "Coding-system error", Qio_error);
+ DEFERROR_STANDARD (Qcoding_system_error, Qio_error);
DEFSUBR (Fcoding_system_p);
DEFSUBR (Ffind_coding_system);
defsymbol (&Qshift_jis, "shift-jis");
defsymbol (&Qucs4, "ucs-4");
defsymbol (&Qutf8, "utf-8");
+ defsymbol (&Qutf16, "utf-16");
defsymbol (&Qccl, "ccl");
defsymbol (&Qiso2022, "iso2022");
#endif /* MULE */
defsymbol (&Qescape_quoted, "escape-quoted");
#endif /* MULE */
#ifdef UTF2000
+ defsymbol (&Qutf_8_mcs, "utf-8-mcs");
defsymbol (&Qdisable_composition, "disable-composition");
+ defsymbol (&Quse_entity_reference, "use-entity-reference");
+ defsymbol (&Qd, "d");
+ defsymbol (&Qx, "x");
+ defsymbol (&QX, "X");
#endif
defsymbol (&Qencode, "encode");
defsymbol (&Qdecode, "decode");
"big5");
defsymbol (&coding_category_symbol[CODING_CATEGORY_UCS4],
"ucs-4");
+ defsymbol (&coding_category_symbol[CODING_CATEGORY_UTF16],
+ "utf-16");
defsymbol (&coding_category_symbol[CODING_CATEGORY_UTF8],
"utf-8");
defsymbol (&coding_category_symbol[CODING_CATEGORY_ISO_7],
int i;
fcd = xnew (struct file_coding_dump);
- dumpstruct (&fcd, &fcd_description);
+ dump_add_root_struct_ptr (&fcd, &fcd_description);
/* Initialize to something reasonable ... */
- for (i = 0; i <= CODING_CATEGORY_LAST; i++)
+ for (i = 0; i < CODING_CATEGORY_LAST; i++)
{
fcd->coding_category_system[i] = Qnil;
fcd->coding_category_by_priority[i] = i;
*/ );
Vfile_name_coding_system = Qnil;
+ DEFVAR_LISP ("coded-charset-entity-reference-alist",
+ &Vcoded_charset_entity_reference_alist /*
+Alist of coded-charset vs corresponding entity-reference.
+Each element looks like (CCS PREFIX CODE-COLUMNS CODE-TYPE).
+CCS is coded-charset.
+CODE-COLUMNS is columns of code-point of entity-reference.
+CODE-TYPE is format type of code-point of entity-reference.
+`d' means decimal value and `x' means hexadecimal value.
+*/ );
+ Vcoded_charset_entity_reference_alist = Qnil;
+
DEFVAR_BOOL ("enable-multibyte-characters", &enable_multibyte_characters /*
Non-nil means the buffer contents are regarded as multi-byte form
of characters, not a binary code. This affects the display, file I/O,
make_lisp_hash_table (50, HASH_TABLE_NON_WEAK, HASH_TABLE_EQ);
the_codesys_prop_dynarr = Dynarr_new (codesys_prop);
- dumpstruct (&the_codesys_prop_dynarr, &codesys_prop_dynarr_description);
+ dump_add_root_struct_ptr (&the_codesys_prop_dynarr, &codesys_prop_dynarr_description);
#define DEFINE_CODESYS_PROP(Prop_Type, Sym) do \
{ \
DEFINE_CODESYS_PROP (CODESYS_PROP_CCL, Qencode);
DEFINE_CODESYS_PROP (CODESYS_PROP_CCL, Qdecode);
+#ifdef UTF2000
+ DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Qdisable_composition);
+ DEFINE_CODESYS_PROP (CODESYS_PROP_ALL_OK, Quse_entity_reference);
+#endif
#endif /* MULE */
/* Need to create this here or we're really screwed. */
Fmake_coding_system
#ifdef UTF2000
Fmake_coding_system
- (Qutf8, Qutf8,
- build_string ("Coding-system of ISO/IEC 10646 UTF-8."),
- list2 (Qmnemonic, build_string ("UTF8")));
+ (Qutf_8_mcs, Qutf8,
+ build_string
+ ("Coding-system of UTF-8 with Multiple Coded-character-Sets extension."),
+ list2 (Qmnemonic, build_string ("MTF8")));
#endif
Fdefine_coding_system_alias (Qno_conversion, Qraw_text);
#ifdef UTF2000
fcd->coding_category_system[CODING_CATEGORY_UTF8]
- = Fget_coding_system (Qutf8);
+ = Fget_coding_system (Qutf_8_mcs);
#endif
#if defined(MULE) && !defined(UTF2000)
{
- unsigned int i;
+ size_t i;
for (i = 0; i < countof (fcd->ucs_to_mule_table); i++)
fcd->ucs_to_mule_table[i] = Qnil;