X-Git-Url: https://git.distorted.org.uk/~mdw/sgt/putty/blobdiff_plain/126ce23485be0c770b2e0c866c17331e3de7a5cc..6b5cf8b49ddde467168034099c19447df4ba30b8:/unix/uxucs.c diff --git a/unix/uxucs.c b/unix/uxucs.c index 164d8890..8731239f 100644 --- a/unix/uxucs.c +++ b/unix/uxucs.c @@ -1,86 +1,20 @@ #include #include #include +#include +#include +#include #include + #include "putty.h" +#include "terminal.h" #include "misc.h" /* * Unix Unicode-handling routines. - * - * FIXME: currently trivial stub versions assuming all codepages - * are ISO8859-1. */ -void lpage_send(int codepage, char *buf, int len, int interactive) -{ - ldisc_send(buf, len, interactive); -} - -void luni_send(wchar_t * widebuf, int len, int interactive) -{ - static char *linebuffer = 0; - static int linesize = 0; - int ratio = (in_utf)?6:1; - int i; - char *p; - - if (len * ratio > linesize) { - sfree(linebuffer); - linebuffer = smalloc(len * ratio * 2 * sizeof(wchar_t)); - linesize = len * ratio * 2; - } - - if (in_utf) { - /* UTF is a simple algorithm */ - for (p = linebuffer, i = 0; i < len; i++) { - wchar_t ch = widebuf[i]; - - if ((ch&0xF800) == 0xD800) ch = '.'; - - if (ch < 0x80) { - *p++ = (char) (ch); - } else if (ch < 0x800) { - *p++ = (0xC0 | (ch >> 6)); - *p++ = (0x80 | (ch & 0x3F)); - } else if (ch < 0x10000) { - *p++ = (0xE0 | (ch >> 12)); - *p++ = (0x80 | ((ch >> 6) & 0x3F)); - *p++ = (0x80 | (ch & 0x3F)); - } else if (ch < 0x200000) { - *p++ = (0xF0 | (ch >> 18)); - *p++ = (0x80 | ((ch >> 12) & 0x3F)); - *p++ = (0x80 | ((ch >> 6) & 0x3F)); - *p++ = (0x80 | (ch & 0x3F)); - } else if (ch < 0x4000000) { - *p++ = (0xF8 | (ch >> 24)); - *p++ = (0x80 | ((ch >> 18) & 0x3F)); - *p++ = (0x80 | ((ch >> 12) & 0x3F)); - *p++ = (0x80 | ((ch >> 6) & 0x3F)); - *p++ = (0x80 | (ch & 0x3F)); - } else { - *p++ = (0xFC | (ch >> 30)); - *p++ = (0x80 | ((ch >> 24) & 0x3F)); - *p++ = (0x80 | ((ch >> 18) & 0x3F)); - *p++ = (0x80 | ((ch >> 12) & 0x3F)); - *p++ = (0x80 | ((ch >> 6) & 0x3F)); - *p++ = (0x80 | (ch & 0x3F)); - } - } - } else { - for (p = linebuffer, i = 0; i < len; i++) { - wchar_t ch = widebuf[i]; - if (ch < 0x100) - *p++ = (char) ch; - else - *p++ = '.'; - } - } - if (p > linebuffer) - ldisc_send(linebuffer, p - linebuffer, interactive); -} - int is_dbcs_leadbyte(int codepage, char byte) { return 0; /* we don't do DBCS */ @@ -89,48 +23,198 @@ int is_dbcs_leadbyte(int codepage, char byte) int mb_to_wc(int codepage, int flags, char *mbstr, int mblen, wchar_t *wcstr, int wclen) { - int ret = 0; - while (mblen > 0 && wclen > 0) { - *wcstr++ = (unsigned char) *mbstr++; - mblen--, wclen--, ret++; - } - return ret; /* FIXME: check error codes! */ + if (codepage == DEFAULT_CODEPAGE) { + int n = 0; + mbstate_t state = { 0 }; + + setlocale(LC_CTYPE, ""); + + while (mblen > 0) { + size_t i = mbrtowc(wcstr+n, mbstr, (size_t)mblen, &state); + if (i == (size_t)-1 || i == (size_t)-2) + break; + n++; + mbstr += i; + mblen -= i; + } + + setlocale(LC_CTYPE, "C"); + + return n; + } else if (codepage == CS_NONE) { + int n = 0; + + while (mblen > 0) { + wcstr[n] = 0xD800 | (mbstr[0] & 0xFF); + n++; + mbstr++; + mblen--; + } + + return n; + } else + return charset_to_unicode(&mbstr, &mblen, wcstr, wclen, codepage, + NULL, NULL, 0); } int wc_to_mb(int codepage, int flags, wchar_t *wcstr, int wclen, - char *mbstr, int mblen, char *defchr, int *defused) + char *mbstr, int mblen, char *defchr, int *defused, + struct unicode_data *ucsdata) { - int ret = 0; + /* FIXME: we should remove the defused param completely... */ if (defused) *defused = 0; - while (mblen > 0 && wclen > 0) { - if (*wcstr >= 0x100) { - if (defchr) - *mbstr++ = *defchr; - else - *mbstr++ = '\xBF'; - if (defused) - *defused = 1; - } else - *mbstr++ = (unsigned char) *wcstr; - wcstr++; - mblen--, wclen--, ret++; + + if (codepage == DEFAULT_CODEPAGE) { + char output[MB_LEN_MAX]; + mbstate_t state = { 0 }; + int n = 0; + + setlocale(LC_CTYPE, ""); + + while (wclen > 0) { + int i = wcrtomb(output, wcstr[0], &state); + if (i == (size_t)-1 || i > n - mblen) + break; + memcpy(mbstr+n, output, i); + n += i; + wcstr++; + wclen--; + } + + setlocale(LC_CTYPE, "C"); + + return n; + } else if (codepage == CS_NONE) { + int n = 0; + while (wclen > 0 && n < mblen) { + if (*wcstr >= 0xD800 && *wcstr < 0xD900) + mbstr[n++] = (*wcstr & 0xFF); + else if (defchr) + mbstr[n++] = *defchr; + wcstr++; + wclen--; + } + return n; + } else { + return charset_from_unicode(&wcstr, &wclen, mbstr, mblen, codepage, + NULL, defchr?defchr:NULL, defchr?1:0); } - return ret; /* FIXME: check error codes! */ } -void init_ucs(void) +/* + * Return value is TRUE if pterm is to run in direct-to-font mode. + */ +int init_ucs(struct unicode_data *ucsdata, + char *linecharset, int font_charset) { - int i; - /* Find the line control characters. FIXME: this is not right. */ - for (i = 0; i < 256; i++) - if (i < ' ' || (i >= 0x7F && i < 0xA0)) - unitab_ctrl[i] = i; + int i, ret = 0; + + /* + * In the platform-independent parts of the code, font_codepage + * is used only for system DBCS support - which we don't + * support at all. So we set this to something which will never + * be used. + */ + ucsdata->font_codepage = -1; + + /* + * line_codepage should be decoded from the specification in + * cfg. + */ + ucsdata->line_codepage = charset_from_mimeenc(linecharset); + if (ucsdata->line_codepage == CS_NONE) + ucsdata->line_codepage = charset_from_xenc(linecharset); + + /* + * If line_codepage is _still_ CS_NONE, we assume we're using + * the font's own encoding. This has been passed in to us, so + * we use that. If it's still CS_NONE after _that_ - i.e. the + * font we were given had an incomprehensible charset - then we + * fall back to using the D800 page. + */ + if (ucsdata->line_codepage == CS_NONE) + ucsdata->line_codepage = font_charset; + + if (ucsdata->line_codepage == CS_NONE) + ret = 1; + + /* + * Set up unitab_line, by translating each individual character + * in the line codepage into Unicode. + */ + for (i = 0; i < 256; i++) { + char c[1], *p; + wchar_t wc[1]; + int len; + c[0] = i; + p = c; + len = 1; + if (ucsdata->line_codepage == CS_NONE) + ucsdata->unitab_line[i] = 0xD800 | i; + else if (1 == charset_to_unicode(&p, &len, wc, 1, + ucsdata->line_codepage, + NULL, L"", 0)) + ucsdata->unitab_line[i] = wc[0]; + else + ucsdata->unitab_line[i] = 0xFFFD; + } + + /* + * Set up unitab_xterm. This is the same as unitab_line except + * in the line-drawing regions, where it follows the Unicode + * encoding. + * + * (Note that the strange X encoding of line-drawing characters + * in the bottom 32 glyphs of ISO8859-1 fonts is taken care of + * by the font encoding, which will spot such a font and act as + * if it were in a variant encoding of ISO8859-1.) + */ + for (i = 0; i < 256; i++) { + static const wchar_t unitab_xterm_std[32] = { + 0x2666, 0x2592, 0x2409, 0x240c, 0x240d, 0x240a, 0x00b0, 0x00b1, + 0x2424, 0x240b, 0x2518, 0x2510, 0x250c, 0x2514, 0x253c, 0x23ba, + 0x23bb, 0x2500, 0x23bc, 0x23bd, 0x251c, 0x2524, 0x2534, 0x252c, + 0x2502, 0x2264, 0x2265, 0x03c0, 0x2260, 0x00a3, 0x00b7, 0x0020 + }; + if (i >= 0x5F && i < 0x7F) + ucsdata->unitab_xterm[i] = unitab_xterm_std[i & 0x1F]; + else + ucsdata->unitab_xterm[i] = ucsdata->unitab_line[i]; + } + + /* + * Set up unitab_scoacs. The SCO Alternate Character Set is + * simply CP437. + */ + for (i = 0; i < 256; i++) { + char c[1], *p; + wchar_t wc[1]; + int len; + c[0] = i; + p = c; + len = 1; + if (1 == charset_to_unicode(&p, &len, wc, 1, CS_CP437, NULL, L"", 0)) + ucsdata->unitab_scoacs[i] = wc[0]; else - unitab_ctrl[i] = 0xFF; + ucsdata->unitab_scoacs[i] = 0xFFFD; + } + /* + * Find the control characters in the line codepage. For + * direct-to-font mode using the D800 hack, we assume 00-1F and + * 7F are controls, but allow 80-9F through. (It's as good a + * guess as anything; and my bet is that half the weird fonts + * used in this way will be IBM or MS code pages anyway.) + */ for (i = 0; i < 256; i++) { - unitab_line[i] = unitab_scoacs[i] = i; - unitab_xterm[i] = (i+1) & 0x1F; + int lineval = ucsdata->unitab_line[i]; + if (lineval < ' ' || (lineval >= 0x7F && lineval < 0xA0) || + (lineval >= 0xD800 && lineval < 0xD820) || (lineval == 0xD87F)) + ucsdata->unitab_ctrl[i] = i; + else + ucsdata->unitab_ctrl[i] = 0xFF; } + + return ret; }