* Copyright (C) 1995-2005, Index Data ApS
* See the file LICENSE for details.
*
- * $Id: siconv.c,v 1.14 2005-10-28 18:36:58 adam Exp $
+ * $Id: siconv.c,v 1.15 2005-11-06 01:28:09 adam Exp $
*/
/**
* \file siconv.c
unsigned long yaz_marc8_9_conv (unsigned char *inp, size_t inbytesleft,
size_t *no_read, int *combining);
-#define NEW_COMB 1
-
struct yaz_iconv_struct {
int my_errno;
int init_flag;
char **outbuf, size_t *outbytesleft,
int last);
int marc8_esc_mode;
-#if NEW_COMB
+
int comb_offset;
int comb_size;
unsigned long comb_x[8];
size_t comb_no_read[8];
-#else
- int marc8_comb_x;
- int marc8_comb_no_read;
-#endif
size_t no_read_x;
unsigned long unget_x;
#if HAVE_ICONV_H
#endif
-#if NEW_COMB
static unsigned long yaz_read_marc8_comb (yaz_iconv_t cd, unsigned char *inp,
size_t inbytesleft, size_t *no_read,
int *comb);
We'll increment the no_read counter by 1, since we want to skip over
the processing of the closing ligature character
*/
+#if 0
if (x == 0x0361 || x == 0x0360)
*no_read += 1;
+#endif
cd->comb_offset++;
return x;
}
return x;
}
}
-#else
-static unsigned long yaz_read_marc8 (yaz_iconv_t cd, unsigned char *inp,
- size_t inbytesleft, size_t *no_read)
-{
- if (cd->marc8_comb_x)
- {
- unsigned long x = cd->marc8_comb_x;
- *no_read = cd->marc8_comb_no_read;
- cd->marc8_comb_x = 0;
- return x;
- }
- *no_read = 0;
- while(inbytesleft >= 1 && inp[0] == 27)
- {
- size_t inbytesleft0 = inbytesleft;
- inp++;
- inbytesleft--;
- while(inbytesleft > 0 && strchr("(,$!", *inp))
- {
- inbytesleft--;
- inp++;
- }
- if (inbytesleft <= 0)
- {
- *no_read = 0;
- cd->my_errno = YAZ_ICONV_EINVAL;
- return 0;
- }
- cd->marc8_esc_mode = *inp++;
- inbytesleft--;
- (*no_read) += inbytesleft0 - inbytesleft;
- }
- if (inbytesleft <= 0)
- return 0;
- else
- {
- unsigned long x;
- int comb = 0;
- size_t no_read_sub = 0;
-
- switch(cd->marc8_esc_mode)
- {
- case 'B': /* Basic ASCII */
- case 'E': /* ANSEL */
- case 's': /* ASCII */
- x = yaz_marc8_1_conv(inp, inbytesleft, &no_read_sub, &comb);
- break;
- case 'g': /* Greek */
- x = yaz_marc8_2_conv(inp, inbytesleft, &no_read_sub, &comb);
- break;
- case 'b': /* Subscripts */
- x = yaz_marc8_3_conv(inp, inbytesleft, &no_read_sub, &comb);
- break;
- case 'p': /* Superscripts */
- x = yaz_marc8_4_conv(inp, inbytesleft, &no_read_sub, &comb);
- break;
- case '2': /* Basic Hebrew */
- x = yaz_marc8_5_conv(inp, inbytesleft, &no_read_sub, &comb);
- break;
- case 'N': /* Basic Cyrillic */
- case 'Q': /* Extended Cyrillic */
- x = yaz_marc8_6_conv(inp, inbytesleft, &no_read_sub, &comb);
- break;
- case '3': /* Basic Arabic */
- case '4': /* Extended Arabic */
- x = yaz_marc8_7_conv(inp, inbytesleft, &no_read_sub, &comb);
- break;
- case 'S': /* Greek */
- x = yaz_marc8_8_conv(inp, inbytesleft, &no_read_sub, &comb);
- break;
- case '1': /* Chinese, Japanese, Korean (EACC) */
- x = yaz_marc8_9_conv(inp, inbytesleft, &no_read_sub, &comb);
- break;
- default:
- *no_read = 0;
- cd->my_errno = YAZ_ICONV_EILSEQ;
- return 0;
- }
-#if 0
- printf ("esc mode=%c x=%04lX comb=%d\n", cd->marc8_esc_mode, x, comb);
-#endif
- *no_read += no_read_sub;
-
- if (comb && cd->marc8_comb_x == 0)
- {
- size_t tmp_read = 0;
- unsigned long next_x;
-
- /* read next char .. */
- next_x = yaz_read_marc8(cd, inp + *no_read,
- inbytesleft - *no_read, &tmp_read);
- /* save this x for later .. */
- cd->marc8_comb_x = x;
- /* save next read for later .. */
- cd->marc8_comb_no_read = tmp_read;
- /* return next x - thereby swap */
- x = next_x;
- }
- return x;
- }
-}
-#endif
static size_t yaz_write_UTF8 (yaz_iconv_t cd, unsigned long x,
char **outbuf, size_t *outbytesleft,
cd->init_handle = 0;
cd->my_errno = YAZ_ICONV_UNKNOWN;
cd->marc8_esc_mode = 'B';
-#if NEW_COMB
cd->comb_offset = cd->comb_size = 0;
-#else
- cd->marc8_comb_x = 0;
-#endif
cd->compose_char = 0;
/* a useful hack: if fromcode has leading @,