- /* list of two char unicode sequence that, when combined, are
- equivalent to single unicode chars that can be represented in
- ISO-8859-1/Latin-1.
- Regular iconv on Linux at least does not seem to convert these,
- but since MARC-8 to UTF-8 generates these composed sequence
- we get a better chance of a successful MARC-8 -> ISO-8859-1
- conversion */
- static struct {
- unsigned long x1, x2;
- unsigned y;
- } comb[] = {
- { 'A', 0x0300, 0xc0}, /* LATIN CAPITAL LETTER A WITH GRAVE */
- { 'A', 0x0301, 0xc1}, /* LATIN CAPITAL LETTER A WITH ACUTE */
- { 'A', 0x0302, 0xc2}, /* LATIN CAPITAL LETTER A WITH CIRCUMFLEX */
- { 'A', 0x0303, 0xc3}, /* LATIN CAPITAL LETTER A WITH TILDE */
- { 'A', 0x0308, 0xc4}, /* LATIN CAPITAL LETTER A WITH DIAERESIS */
- { 'A', 0x030a, 0xc5}, /* LATIN CAPITAL LETTER A WITH RING ABOVE */
- /* no need for 0xc6 LATIN CAPITAL LETTER AE */
- { 'C', 0x0327, 0xc7}, /* LATIN CAPITAL LETTER C WITH CEDILLA */
- { 'E', 0x0300, 0xc8}, /* LATIN CAPITAL LETTER E WITH GRAVE */
- { 'E', 0x0301, 0xc9}, /* LATIN CAPITAL LETTER E WITH ACUTE */
- { 'E', 0x0302, 0xca}, /* LATIN CAPITAL LETTER E WITH CIRCUMFLEX */
- { 'E', 0x0308, 0xcb}, /* LATIN CAPITAL LETTER E WITH DIAERESIS */
- { 'I', 0x0300, 0xcc}, /* LATIN CAPITAL LETTER I WITH GRAVE */
- { 'I', 0x0301, 0xcd}, /* LATIN CAPITAL LETTER I WITH ACUTE */
- { 'I', 0x0302, 0xce}, /* LATIN CAPITAL LETTER I WITH CIRCUMFLEX */
- { 'I', 0x0308, 0xcf}, /* LATIN CAPITAL LETTER I WITH DIAERESIS */
- { 'N', 0x0303, 0xd1}, /* LATIN CAPITAL LETTER N WITH TILDE */
- { 'O', 0x0300, 0xd2}, /* LATIN CAPITAL LETTER O WITH GRAVE */
- { 'O', 0x0301, 0xd3}, /* LATIN CAPITAL LETTER O WITH ACUTE */
- { 'O', 0x0302, 0xd4}, /* LATIN CAPITAL LETTER O WITH CIRCUMFLEX */
- { 'O', 0x0303, 0xd5}, /* LATIN CAPITAL LETTER O WITH TILDE */
- { 'O', 0x0308, 0xd6}, /* LATIN CAPITAL LETTER O WITH DIAERESIS */
- /* omitted: 0xd7 MULTIPLICATION SIGN */
- /* omitted: 0xd8 LATIN CAPITAL LETTER O WITH STROKE */
- { 'U', 0x0300, 0xd9}, /* LATIN CAPITAL LETTER U WITH GRAVE */
- { 'U', 0x0301, 0xda}, /* LATIN CAPITAL LETTER U WITH ACUTE */
- { 'U', 0x0302, 0xdb}, /* LATIN CAPITAL LETTER U WITH CIRCUMFLEX */
- { 'U', 0x0308, 0xdc}, /* LATIN CAPITAL LETTER U WITH DIAERESIS */
- { 'Y', 0x0301, 0xdd}, /* LATIN CAPITAL LETTER Y WITH ACUTE */
- /* omitted: 0xde LATIN CAPITAL LETTER THORN */
- /* omitted: 0xdf LATIN SMALL LETTER SHARP S */
- { 'a', 0x0300, 0xe0}, /* LATIN SMALL LETTER A WITH GRAVE */
- { 'a', 0x0301, 0xe1}, /* LATIN SMALL LETTER A WITH ACUTE */
- { 'a', 0x0302, 0xe2}, /* LATIN SMALL LETTER A WITH CIRCUMFLEX */
- { 'a', 0x0303, 0xe3}, /* LATIN SMALL LETTER A WITH TILDE */
- { 'a', 0x0308, 0xe4}, /* LATIN SMALL LETTER A WITH DIAERESIS */
- { 'a', 0x030a, 0xe5}, /* LATIN SMALL LETTER A WITH RING ABOVE */
- /* omitted: 0xe6 LATIN SMALL LETTER AE */
- { 'c', 0x0327, 0xe7}, /* LATIN SMALL LETTER C WITH CEDILLA */
- { 'e', 0x0300, 0xe8}, /* LATIN SMALL LETTER E WITH GRAVE */
- { 'e', 0x0301, 0xe9}, /* LATIN SMALL LETTER E WITH ACUTE */
- { 'e', 0x0302, 0xea}, /* LATIN SMALL LETTER E WITH CIRCUMFLEX */
- { 'e', 0x0308, 0xeb}, /* LATIN SMALL LETTER E WITH DIAERESIS */
- { 'i', 0x0300, 0xec}, /* LATIN SMALL LETTER I WITH GRAVE */
- { 'i', 0x0301, 0xed}, /* LATIN SMALL LETTER I WITH ACUTE */
- { 'i', 0x0302, 0xee}, /* LATIN SMALL LETTER I WITH CIRCUMFLEX */
- { 'i', 0x0308, 0xef}, /* LATIN SMALL LETTER I WITH DIAERESIS */
- /* omitted: 0xf0 LATIN SMALL LETTER ETH */
- { 'n', 0x0303, 0xf1}, /* LATIN SMALL LETTER N WITH TILDE */
- { 'o', 0x0300, 0xf2}, /* LATIN SMALL LETTER O WITH GRAVE */
- { 'o', 0x0301, 0xf3}, /* LATIN SMALL LETTER O WITH ACUTE */
- { 'o', 0x0302, 0xf4}, /* LATIN SMALL LETTER O WITH CIRCUMFLEX */
- { 'o', 0x0303, 0xf5}, /* LATIN SMALL LETTER O WITH TILDE */
- { 'o', 0x0308, 0xf6}, /* LATIN SMALL LETTER O WITH DIAERESIS */
- /* omitted: 0xf7 DIVISION SIGN */
- /* omitted: 0xf8 LATIN SMALL LETTER O WITH STROKE */
- { 'u', 0x0300, 0xf9}, /* LATIN SMALL LETTER U WITH GRAVE */
- { 'u', 0x0301, 0xfa}, /* LATIN SMALL LETTER U WITH ACUTE */
- { 'u', 0x0302, 0xfb}, /* LATIN SMALL LETTER U WITH CIRCUMFLEX */
- { 'u', 0x0308, 0xfc}, /* LATIN SMALL LETTER U WITH DIAERESIS */
- { 'y', 0x0301, 0xfd}, /* LATIN SMALL LETTER Y WITH ACUTE */
- /* omitted: 0xfe LATIN SMALL LETTER THORN */
- { 'y', 0x0308, 0xff}, /* LATIN SMALL LETTER Y WITH DIAERESIS */
-
- { 0, 0, 0}
- };
- unsigned char *outp = (unsigned char *) *outbuf;
-
- if (cd->compose_char)
- {
- int i;
- for (i = 0; comb[i].x1; i++)
- if (cd->compose_char == comb[i].x1 && x == comb[i].x2)
- {
- x = comb[i].y;
- break;
- }
- if (*outbytesleft < 1)
- { /* no room. Retain compose_char and bail out */
- cd->my_errno = YAZ_ICONV_E2BIG;
- return (size_t)(-1);
- }
- if (!comb[i].x1)
- { /* not found. Just write compose_char */
- *outp++ = (unsigned char) cd->compose_char;
- (*outbytesleft)--;
- *outbuf = (char *) outp;
- }
- /* compose_char used so reset it. x now holds current char */
- cd->compose_char = 0;
- }
-
- if (!last && x > 32 && x < 127 && cd->compose_char == 0)
- {
- cd->compose_char = x;
- return 0;
- }
- else if (x > 255 || x < 1)
- {
- cd->my_errno = YAZ_ICONV_EILSEQ;
- return (size_t) -1;
- }
- else if (*outbytesleft < 1)
- {
- cd->my_errno = YAZ_ICONV_E2BIG;
- return (size_t)(-1);
- }
- *outp++ = (unsigned char) x;
- (*outbytesleft)--;
- *outbuf = (char *) outp;