X-Git-Url: http://git.chise.org/gitweb/?a=blobdiff_plain;f=src%2Ffile-coding.c;h=3068c896823dac45e334b378d4adf0bf92d709c6;hb=f8c889244f022ae695ea8dc67267499286921903;hp=808d34374b1209a77c9dd6c6a48d522f92b73dca;hpb=2db0c70abf3c54214bbd8c1d8df0e4e79de1b405;p=chise%2Fxemacs-chise.git- diff --git a/src/file-coding.c b/src/file-coding.c index 808d343..3068c89 100644 --- a/src/file-coding.c +++ b/src/file-coding.c @@ -25,6 +25,7 @@ Boston, MA 02111-1307, USA. */ #include #include "lisp.h" + #include "buffer.h" #include "elhash.h" #include "insdel.h" @@ -244,10 +245,46 @@ static Lisp_Object mark_coding_system (Lisp_Object, void (*) (Lisp_Object)); static void print_coding_system (Lisp_Object, Lisp_Object, int); static void finalize_coding_system (void *header, int for_disksave); +#ifdef MULE +static const struct lrecord_description ccs_description_1[] = { + { XD_LISP_OBJECT, offsetof(charset_conversion_spec, from_charset), 2 }, + { XD_END } +}; + +static const struct struct_description ccs_description = { + sizeof(charset_conversion_spec), + ccs_description_1 +}; + +static const struct lrecord_description ccsd_description_1[] = { + XD_DYNARR_DESC(charset_conversion_spec_dynarr, &ccs_description), + { XD_END } +}; + +static const struct struct_description ccsd_description = { + sizeof(charset_conversion_spec_dynarr), + ccsd_description_1 +}; +#endif + +static const struct lrecord_description coding_system_description[] = { + { XD_LISP_OBJECT, offsetof(struct Lisp_Coding_System, name), 2 }, + { XD_LISP_OBJECT, offsetof(struct Lisp_Coding_System, mnemonic), 3 }, + { XD_LISP_OBJECT, offsetof(struct Lisp_Coding_System, eol_lf), 3 }, +#ifdef MULE + { XD_LISP_OBJECT, offsetof(struct Lisp_Coding_System, iso2022.initial_charset), 4 }, + { XD_STRUCT_PTR, offsetof(struct Lisp_Coding_System, iso2022.input_conv), 1, &ccsd_description }, + { XD_STRUCT_PTR, offsetof(struct Lisp_Coding_System, iso2022.output_conv), 1, &ccsd_description }, + { XD_LISP_OBJECT, offsetof(struct Lisp_Coding_System, ccl.decode), 2 }, +#endif + { XD_END } +}; + DEFINE_LRECORD_IMPLEMENTATION ("coding-system", coding_system, mark_coding_system, print_coding_system, finalize_coding_system, - 0, 0, struct Lisp_Coding_System); + 0, 0, coding_system_description, + struct Lisp_Coding_System); static Lisp_Object mark_coding_system (Lisp_Object obj, void (*markobj) (Lisp_Object)) @@ -821,10 +858,6 @@ if TYPE is 'ccl: CHECK_STRING (doc_string); CODING_SYSTEM_DOC_STRING (codesys) = doc_string; -#ifdef UTF2000 - if (ty == CODESYS_NO_CONVERSION) - codesys->fixed.size = 1; -#endif EXTERNAL_PROPERTY_LIST_LOOP (rest, key, value, props) { if (EQ (key, Qmnemonic)) @@ -1587,26 +1620,65 @@ determine_real_coding_system (Lstream *stream, Lisp_Object *codesys_in_out, if (XCODING_SYSTEM_TYPE (*codesys_in_out) == CODESYS_AUTODETECT || *eol_type_in_out == EOL_AUTODETECT) { + unsigned char random_buffer[4096]; + int nread; + Lisp_Object coding_system = Qnil; - while (1) + nread = Lstream_read (stream, random_buffer, sizeof (random_buffer)); + if (nread) { - unsigned char random_buffer[4096]; - int nread; + unsigned char *cp = random_buffer; - nread = Lstream_read (stream, random_buffer, sizeof (random_buffer)); - if (!nread) - break; - if (detect_coding_type (&decst, random_buffer, nread, - XCODING_SYSTEM_TYPE (*codesys_in_out) != - CODESYS_AUTODETECT)) - break; - } + while (cp < random_buffer + nread) + { + if ((*cp++ == 'c') && (cp < random_buffer + nread) && + (*cp++ == 'o') && (cp < random_buffer + nread) && + (*cp++ == 'd') && (cp < random_buffer + nread) && + (*cp++ == 'i') && (cp < random_buffer + nread) && + (*cp++ == 'n') && (cp < random_buffer + nread) && + (*cp++ == 'g') && (cp < random_buffer + nread) && + (*cp++ == ':') && (cp < random_buffer + nread)) + { + unsigned char coding_system_name[4096 - 6]; + unsigned char *np = coding_system_name; + while ( (cp < random_buffer + nread) + && ((*cp == ' ') || (*cp == '\t')) ) + { + cp++; + } + while ( (cp < random_buffer + nread) && + (*cp != ' ') && (*cp != '\t') && (*cp != ';') ) + { + *np++ = *cp++; + } + *np = 0; + coding_system + = Ffind_coding_system (intern (coding_system_name)); + break; + } + } + if (EQ(coding_system, Qnil)) + do{ + if (detect_coding_type (&decst, random_buffer, nread, + XCODING_SYSTEM_TYPE (*codesys_in_out) + != CODESYS_AUTODETECT)) + break; + nread = Lstream_read (stream, + random_buffer, sizeof (random_buffer)); + if (!nread) + break; + } while(1); + } *eol_type_in_out = decst.eol_type; if (XCODING_SYSTEM_TYPE (*codesys_in_out) == CODESYS_AUTODETECT) - *codesys_in_out = coding_system_from_mask (decst.mask); + { + if (EQ(coding_system, Qnil)) + *codesys_in_out = coding_system_from_mask (decst.mask); + else + *codesys_in_out = coding_system; + } } - /* If we absolutely can't determine the EOL type, just assume LF. */ if (*eol_type_in_out == EOL_AUTODETECT) *eol_type_in_out = EOL_LF; @@ -1860,6 +1932,9 @@ struct decoding_stream /* Additional information (the state of the running CCL program) used by the CCL decoder. */ struct ccl_program ccl; + + /* counter for UTF-8 or UCS-4 */ + unsigned char counter; #endif struct detection_state decst; }; @@ -1994,6 +2069,7 @@ reset_decoding_stream (struct decoding_stream *str) { setup_ccl_program (&str->ccl, CODING_SYSTEM_CCL_DECODE (str->codesys)); } + str->counter = 0; #endif /* MULE */ str->flags = str->ch = 0; } @@ -2177,7 +2253,8 @@ mule_decode (Lstream *decoding, CONST unsigned char *src, decode_coding_utf8 (decoding, src, dst, n); break; case CODESYS_CCL: - ccl_driver (&str->ccl, src, dst, n, 0); + str->ccl.last_block = str->flags & CODING_STATE_END; + ccl_driver (&str->ccl, src, dst, n, 0, CCL_MODE_DECODING); break; case CODESYS_ISO2022: decode_coding_iso2022 (decoding, src, dst, n); @@ -2593,7 +2670,8 @@ mule_encode (Lstream *encoding, CONST unsigned char *src, encode_coding_utf8 (encoding, src, dst, n); break; case CODESYS_CCL: - ccl_driver (&str->ccl, src, dst, n, 0); + str->ccl.last_block = str->flags & CODING_STATE_END; + ccl_driver (&str->ccl, src, dst, n, 0, CCL_MODE_ENCODING); break; case CODESYS_ISO2022: encode_coding_iso2022 (encoding, src, dst, n); @@ -2752,10 +2830,16 @@ decode_coding_sjis (Lstream *decoding, CONST unsigned char *src, { unsigned char e1, e2; - Dynarr_add (dst, LEADING_BYTE_JAPANESE_JISX0208); DECODE_SJIS (ch, c, e1, e2); +#ifdef UTF2000 + DECODE_ADD_UCS_CHAR(MAKE_CHAR(Vcharset_japanese_jisx0208, + e1 & 0x7F, + e2 & 0x7F), dst); +#else + Dynarr_add (dst, LEADING_BYTE_JAPANESE_JISX0208); Dynarr_add (dst, e1); Dynarr_add (dst, e2); +#endif } else { @@ -2771,8 +2855,13 @@ decode_coding_sjis (Lstream *decoding, CONST unsigned char *src, ch = c; else if (BYTE_SJIS_KATAKANA_P (c)) { +#ifdef UTF2000 + DECODE_ADD_UCS_CHAR(MAKE_CHAR(Vcharset_katakana_jisx0201, + c & 0x7F, 0), dst); +#else Dynarr_add (dst, LEADING_BYTE_KATAKANA_JISX0201); Dynarr_add (dst, c); +#endif } else DECODE_ADD_BINARY_CHAR (c, dst); @@ -2797,10 +2886,82 @@ encode_coding_sjis (Lstream *encoding, CONST unsigned char *src, unsigned int flags = str->flags; unsigned int ch = str->ch; eol_type_t eol_type = CODING_SYSTEM_EOL_TYPE (str->codesys); +#ifdef UTF2000 + unsigned char char_boundary = str->iso2022.current_char_boundary; +#endif while (n--) { c = *src++; +#ifdef UTF2000 + switch (char_boundary) + { + case 0: + if ( c >= 0xfc ) + { + ch = c & 0x01; + char_boundary = 5; + } + else if ( c >= 0xf8 ) + { + ch = c & 0x03; + char_boundary = 4; + } + else if ( c >= 0xf0 ) + { + ch = c & 0x07; + char_boundary = 3; + } + else if ( c >= 0xe0 ) + { + ch = c & 0x0f; + char_boundary = 2; + } + else if ( c >= 0xc0 ) + { + ch = c & 0x1f; + char_boundary = 1; + } + else + { + ch = 0; + if (c == '\n') + { + if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT) + Dynarr_add (dst, '\r'); + if (eol_type != EOL_CR) + Dynarr_add (dst, c); + } + else + Dynarr_add (dst, c); + char_boundary = 0; + } + break; + case 1: + ch = ( ch << 6 ) | ( c & 0x3f ); + { + Lisp_Object charset; + unsigned int c1, c2, s1, s2; + + BREAKUP_CHAR (ch, charset, c1, c2); + if (EQ(charset, Vcharset_katakana_jisx0201)) + { + Dynarr_add (dst, c1 | 0x80); + } + else if (EQ(charset, Vcharset_japanese_jisx0208)) + { + ENCODE_SJIS (c1 | 0x80, c2 | 0x80, s1, s2); + Dynarr_add (dst, s1); + Dynarr_add (dst, s2); + } + } + char_boundary = 0; + break; + default: + ch = ( ch << 6 ) | ( c & 0x3f ); + char_boundary--; + } +#else if (c == '\n') { if (eol_type != EOL_LF && eol_type != EOL_AUTODETECT) @@ -2837,10 +2998,14 @@ encode_coding_sjis (Lstream *encoding, CONST unsigned char *src, ch = 0; } } +#endif } str->flags = flags; str->ch = ch; +#ifdef UTF2000 + str->iso2022.current_char_boundary = char_boundary; +#endif } DEFUN ("decode-shift-jis-char", Fdecode_shift_jis_char, 1, 1, 0, /* @@ -3072,6 +3237,7 @@ static void encode_coding_big5 (Lstream *encoding, CONST unsigned char *src, unsigned_char_dynarr *dst, unsigned int n) { +#ifndef UTF2000 unsigned char c; struct encoding_stream *str = ENCODING_STREAM_DATA (encoding); unsigned int flags = str->flags; @@ -3126,6 +3292,7 @@ encode_coding_big5 (Lstream *encoding, CONST unsigned char *src, str->flags = flags; str->ch = ch; +#endif } @@ -3146,7 +3313,7 @@ Return the corresponding character. if (BYTE_BIG5_TWO_BYTE_1_P (b1) && BYTE_BIG5_TWO_BYTE_2_P (b2)) { - int leading_byte; + Charset_ID leading_byte; Lisp_Object charset; DECODE_BIG5 (b1, b2, leading_byte, c1, c2); charset = CHARSET_BY_LEADING_BYTE (leading_byte); @@ -3373,37 +3540,40 @@ decode_coding_ucs4 (Lstream *decoding, CONST unsigned char *src, struct decoding_stream *str = DECODING_STREAM_DATA (decoding); unsigned int flags = str->flags; unsigned int ch = str->ch; + unsigned char counter = str->counter; while (n--) { unsigned char c = *src++; - switch (flags) + switch (counter) { case 0: ch = c; - flags = 3; + counter = 3; break; case 1: decode_ucs4 ( ( ch << 8 ) | c, dst); ch = 0; - flags = 0; + counter = 0; break; default: ch = ( ch << 8 ) | c; - flags--; + counter--; } } - if (flags & CODING_STATE_END) + if (counter & CODING_STATE_END) DECODE_OUTPUT_PARTIAL_CHAR (ch); str->flags = flags; str->ch = ch; + str->counter = counter; } static void encode_coding_ucs4 (Lstream *encoding, CONST unsigned char *src, unsigned_char_dynarr *dst, unsigned int n) { +#ifndef UTF2000 struct encoding_stream *str = ENCODING_STREAM_DATA (encoding); unsigned int flags = str->flags; unsigned int ch = str->ch; @@ -3532,6 +3702,7 @@ encode_coding_ucs4 (Lstream *encoding, CONST unsigned char *src, str->iso2022.current_charset = charset; /* Verbum caro factum est! */ +#endif } @@ -3582,37 +3753,38 @@ decode_coding_utf8 (Lstream *decoding, CONST unsigned char *src, unsigned int flags = str->flags; unsigned int ch = str->ch; eol_type_t eol_type = str->eol_type; + unsigned char counter = str->counter; while (n--) { unsigned char c = *src++; - switch (flags) + switch (counter) { case 0: if ( c >= 0xfc ) { ch = c & 0x01; - flags = 5; + counter = 5; } else if ( c >= 0xf8 ) { ch = c & 0x03; - flags = 4; + counter = 4; } else if ( c >= 0xf0 ) { ch = c & 0x07; - flags = 3; + counter = 3; } else if ( c >= 0xe0 ) { ch = c & 0x0f; - flags = 2; + counter = 2; } else if ( c >= 0xc0 ) { ch = c & 0x1f; - flags = 1; + counter = 1; } else { @@ -3624,11 +3796,11 @@ decode_coding_utf8 (Lstream *decoding, CONST unsigned char *src, ch = ( ch << 6 ) | ( c & 0x3f ); decode_ucs4 (ch, dst); ch = 0; - flags = 0; + counter = 0; break; default: ch = ( ch << 6 ) | ( c & 0x3f ); - flags--; + counter--; } label_continue_loop:; } @@ -3638,6 +3810,7 @@ decode_coding_utf8 (Lstream *decoding, CONST unsigned char *src, str->flags = flags; str->ch = ch; + str->counter = counter; } #ifndef UTF2000 @@ -4784,7 +4957,7 @@ decode_coding_iso2022 (Lstream *decoding, CONST unsigned char *src, { /* Graphic characters */ Lisp_Object charset; #ifndef UTF2000 - int lb; + Charset_ID lb; #endif int reg; @@ -4828,9 +5001,23 @@ decode_coding_iso2022 (Lstream *decoding, CONST unsigned char *src, charset = new_charset; } -#ifndef UTF2000 +#ifdef UTF2000 + if (XCHARSET_DIMENSION (charset) == 1) + { + DECODE_OUTPUT_PARTIAL_CHAR (ch); + DECODE_ADD_UCS_CHAR + (MAKE_CHAR (charset, c & 0x7F, 0), dst); + } + else if (ch) + { + DECODE_ADD_UCS_CHAR + (MAKE_CHAR (charset, ch & 0x7F, c & 0x7F), dst); + ch = 0; + } + else + ch = c; +#else lb = XCHARSET_LEADING_BYTE (charset); -#endif switch (XCHARSET_REP_BYTES (charset)) { case 1: /* ASCII */ @@ -4840,40 +5027,25 @@ decode_coding_iso2022 (Lstream *decoding, CONST unsigned char *src, case 2: /* one-byte official */ DECODE_OUTPUT_PARTIAL_CHAR (ch); -#ifdef UTF2000 - DECODE_ADD_UCS_CHAR(MAKE_CHAR(charset, c & 0x7F, 0), dst); -#else Dynarr_add (dst, lb); Dynarr_add (dst, c | 0x80); -#endif break; case 3: /* one-byte private or two-byte official */ if (XCHARSET_PRIVATE_P (charset)) { DECODE_OUTPUT_PARTIAL_CHAR (ch); -#ifdef UTF2000 - DECODE_ADD_UCS_CHAR(MAKE_CHAR(charset, c & 0x7F, 0), - dst); -#else Dynarr_add (dst, PRE_LEADING_BYTE_PRIVATE_1); Dynarr_add (dst, lb); Dynarr_add (dst, c | 0x80); -#endif } else { if (ch) { -#ifdef UTF2000 - DECODE_ADD_UCS_CHAR(MAKE_CHAR(charset, - ch & 0x7F, - c & 0x7F), dst); -#else Dynarr_add (dst, lb); Dynarr_add (dst, ch | 0x80); Dynarr_add (dst, c | 0x80); -#endif ch = 0; } else @@ -4884,21 +5056,16 @@ decode_coding_iso2022 (Lstream *decoding, CONST unsigned char *src, default: /* two-byte private */ if (ch) { -#ifdef UTF2000 - DECODE_ADD_UCS_CHAR(MAKE_CHAR(charset, - ch & 0x7F, - c & 0x7F), dst); -#else Dynarr_add (dst, PRE_LEADING_BYTE_PRIVATE_2); Dynarr_add (dst, lb); Dynarr_add (dst, ch | 0x80); Dynarr_add (dst, c | 0x80); -#endif ch = 0; } else ch = c; } +#endif } if (!ch) @@ -5623,31 +5790,7 @@ encode_coding_no_conversion (Lstream *encoding, CONST unsigned char *src, break; case 1: ch = ( ch << 6 ) | ( c & 0x3f ); - switch ( str->codesys->fixed.size ) - { - case 1: - Dynarr_add (dst, ch & 0xff); - break; - case 2: - Dynarr_add (dst, (ch >> 8) & 0xff); - Dynarr_add (dst, ch & 0xff); - break; - case 3: - Dynarr_add (dst, (ch >> 16) & 0xff); - Dynarr_add (dst, (ch >> 8) & 0xff); - Dynarr_add (dst, ch & 0xff); - break; - case 4: - Dynarr_add (dst, (ch >> 24) & 0xff); - Dynarr_add (dst, (ch >> 16) & 0xff); - Dynarr_add (dst, (ch >> 8) & 0xff); - Dynarr_add (dst, ch & 0xff); - break; - default: - fprintf(stderr, "It seems %d bytes stream.\n", - str->codesys->fixed.size); - abort (); - } + Dynarr_add (dst, ch & 0xff); char_boundary = 0; break; default: