From fa00744e514a99087f5fe70cac9334b29a04c93a Mon Sep 17 00:00:00 2001 From: Ulrich Drepper Date: Tue, 24 Sep 2002 04:19:03 +0000 Subject: Update. * iconv/iconv_prog.c (main): Provide an error message that identifies the wrong encoding. 2002-09-22 Bruno Haible * iconvdata/tscii.c: New file. * iconvdata/testdata/TSCII: New file. * iconvdata/testdata/TSCII..UTF8: New file. * iconvdata/TSCII.precomposed: New file. * iconvdata/TSCII.irreversible: New file. * iconvdata/gconv-modules (TSCII): New module. * iconvdata/Makefile (modules): Add TSCII. (distribute): Add tscii.c. * iconvdata/tst-table-from.c (try, utf8_decode, main): Double output buffer size. * iconvdata/tst-tables.sh: Add TSCII. * iconvdata/TESTS: Add TSCII. 2002-09-22 Bruno Haible Revert 2002-04-18 patch. * iconvdata/euc-jisx0213.c (EMIT_SHIFT_TO_INIT, BODY for FROM_DIRECTION): Make the FROM direction stateful again. * iconvdata/shift_jisx0213.c (EMIT_SHIFT_TO_INIT, BODY for FROM_DIRECTION): Likewise. 2002-09-22 Bruno Haible --- iconvdata/shift_jisx0213.c | 189 +++++++++++++++++++++++++-------------------- 1 file changed, 105 insertions(+), 84 deletions(-) (limited to 'iconvdata/shift_jisx0213.c') diff --git a/iconvdata/shift_jisx0213.c b/iconvdata/shift_jisx0213.c index d48d67ef0f..119e0017b9 100644 --- a/iconvdata/shift_jisx0213.c +++ b/iconvdata/shift_jisx0213.c @@ -67,7 +67,9 @@ *statep = saved_state -/* During UCS-4 to Shift_JISX0213 conversion, the COUNT element of the state +/* During Shift_JISX0213 to UCS-4 conversion, the COUNT element of the state + contains the last UCS-4 character, shifted by 3 bits. + During UCS-4 to Shift_JISX0213 conversion, the COUNT element of the state contains the last two bytes to be output, shifted by 3 bits. */ /* Since this is a stateful encoding we have to provide code which resets @@ -77,8 +79,17 @@ if (data->__statep->__count != 0) \ { \ if (FROM_DIRECTION) \ - /* We don't use shift states in the FROM_DIRECTION. */ \ - data->__statep->__count = 0; \ + { \ + if (__builtin_expect (outbuf + 4 <= outend, 1)) \ + { \ + /* Write out the last character. */ \ + *((uint32_t *) outbuf)++ = data->__statep->__count >> 3; \ + data->__statep->__count = 0; \ + } \ + else \ + /* We don't have enough room in the output buffer. */ \ + status = __GCONV_FULL_OUTPUT; \ + } \ else \ { \ if (__builtin_expect (outbuf + 2 <= outend, 1)) \ @@ -104,106 +115,116 @@ #define LOOPFCT FROM_LOOP #define BODY \ { \ - uint32_t ch = *inptr; \ + uint32_t ch; \ \ - if (ch < 0x80) \ - { \ - /* Plain ISO646-JP character. */ \ - if (__builtin_expect (ch == 0x5c, 0)) \ - ch = 0xa5; \ - else if (__builtin_expect (ch == 0x7e, 0)) \ - ch = 0x203e; \ - ++inptr; \ - } \ - else if (ch >= 0xa1 && ch <= 0xdf) \ + /* Determine whether there is a buffered character pending. */ \ + ch = *statep >> 3; \ + if (__builtin_expect (ch == 0, 1)) \ { \ - /* Half-width katakana. */ \ - ch += 0xfec0; \ - ++inptr; \ - } \ - else if ((ch >= 0x81 && ch <= 0x9f) || (ch >= 0xe0 && ch <= 0xfc)) \ - { \ - /* Two byte character. */ \ - uint32_t ch2; \ + /* No - so look at the next input byte. */ \ + ch = *inptr; \ \ - if (__builtin_expect (inptr + 1 >= inend, 0)) \ + if (ch < 0x80) \ { \ - /* The second byte is not available. */ \ - result = __GCONV_INCOMPLETE_INPUT; \ - break; \ + /* Plain ISO646-JP character. */ \ + if (__builtin_expect (ch == 0x5c, 0)) \ + ch = 0xa5; \ + else if (__builtin_expect (ch == 0x7e, 0)) \ + ch = 0x203e; \ + ++inptr; \ } \ - \ - ch2 = inptr[1]; \ - \ - /* The second byte must be in the range 0x{40..7E,80..FC}. */ \ - if (__builtin_expect (ch2 < 0x40 || ch2 == 0x7f || ch2 > 0xfc, 0)) \ + else if (ch >= 0xa1 && ch <= 0xdf) \ { \ - /* This is an illegal character. */ \ - STANDARD_FROM_LOOP_ERR_HANDLER (1); \ + /* Half-width katakana. */ \ + ch += 0xfec0; \ + ++inptr; \ } \ - \ - /* Convert to row and column. */ \ - if (ch < 0xe0) \ - ch -= 0x81; \ - else \ - ch -= 0xc1; \ - if (ch2 < 0x80) \ - ch2 -= 0x40; \ - else \ - ch2 -= 0x41; \ - /* Now 0 <= ch <= 0x3b, 0 <= ch2 <= 0xbb. */ \ - ch = 2 * ch; \ - if (ch2 >= 0x5e) \ - ch2 -= 0x5e, ch++; \ - ch2 += 0x21; \ - if (ch >= 0x5e) \ + else if ((ch >= 0x81 && ch <= 0x9f) || (ch >= 0xe0 && ch <= 0xfc)) \ { \ - /* Handling of JISX 0213 plane 2 rows. */ \ - if (ch >= 0x67) \ - ch += 230; \ - else if (ch >= 0x63 || ch == 0x5f) \ - ch += 168; \ - else \ - ch += 162; \ - } \ - \ - ch = jisx0213_to_ucs4 (0x121 + ch, ch2); \ + /* Two byte character. */ \ + uint32_t ch2; \ \ - if (ch == 0) \ - { \ - /* This is an illegal character. */ \ - STANDARD_FROM_LOOP_ERR_HANDLER (1); \ - } \ + if (__builtin_expect (inptr + 1 >= inend, 0)) \ + { \ + /* The second byte is not available. */ \ + result = __GCONV_INCOMPLETE_INPUT; \ + break; \ + } \ \ - if (ch < 0x80) \ - { \ - /* It's a combining character. */ \ - uint32_t u1 = __jisx0213_to_ucs_combining[ch - 1][0]; \ - uint32_t u2 = __jisx0213_to_ucs_combining[ch - 1][1]; \ + ch2 = inptr[1]; \ \ - /* See whether we have room for two characters. */ \ - if (outptr + 8 <= outend) \ + /* The second byte must be in the range 0x{40..7E,80..FC}. */ \ + if (__builtin_expect (ch2 < 0x40 || ch2 == 0x7f || ch2 > 0xfc, 0))\ { \ - inptr += 2; \ - put32 (outptr, u1); \ - outptr += 4; \ - put32 (outptr, u2); \ - outptr += 4; \ - continue; \ + /* This is an illegal character. */ \ + STANDARD_FROM_LOOP_ERR_HANDLER (1); \ } \ + \ + /* Convert to row and column. */ \ + if (ch < 0xe0) \ + ch -= 0x81; \ + else \ + ch -= 0xc1; \ + if (ch2 < 0x80) \ + ch2 -= 0x40; \ else \ + ch2 -= 0x41; \ + /* Now 0 <= ch <= 0x3b, 0 <= ch2 <= 0xbb. */ \ + ch = 2 * ch; \ + if (ch2 >= 0x5e) \ + ch2 -= 0x5e, ch++; \ + ch2 += 0x21; \ + if (ch >= 0x5e) \ { \ + /* Handling of JISX 0213 plane 2 rows. */ \ + if (ch >= 0x67) \ + ch += 230; \ + else if (ch >= 0x63 || ch == 0x5f) \ + ch += 168; \ + else \ + ch += 162; \ + } \ + \ + ch = jisx0213_to_ucs4 (0x121 + ch, ch2); \ + \ + if (ch == 0) \ + { \ + /* This is an illegal character. */ \ + STANDARD_FROM_LOOP_ERR_HANDLER (1); \ + } \ + \ + inptr += 2; \ + \ + if (ch < 0x80) \ + { \ + /* It's a combining character. */ \ + uint32_t u1 = __jisx0213_to_ucs_combining[ch - 1][0]; \ + uint32_t u2 = __jisx0213_to_ucs_combining[ch - 1][1]; \ + \ + put32 (outptr, u1); \ + outptr += 4; \ + \ + /* See whether we have room for two characters. */ \ + if (outptr + 4 <= outend) \ + { \ + put32 (outptr, u2); \ + outptr += 4; \ + continue; \ + } \ + \ + /* Otherwise store only the first character now, and \ + put the second one into the queue. */ \ + *statep = u2 << 3; \ + /* Tell the caller why we terminate the loop. */ \ result = __GCONV_FULL_OUTPUT; \ break; \ } \ } \ - \ - inptr += 2; \ - } \ - else \ - { \ - /* This is illegal. */ \ - STANDARD_FROM_LOOP_ERR_HANDLER (1); \ + else \ + { \ + /* This is illegal. */ \ + STANDARD_FROM_LOOP_ERR_HANDLER (1); \ + } \ } \ \ put32 (outptr, ch); \ -- cgit v1.2.3