htplain.c

用于linux和其他unix下面的

		    } else if ((*p & 0xf8) == 0xf0) {
			me->utf_count = 3;
			me->utf_char = (c & 0x07);
		    } else if ((*p & 0xfc) == 0xf8) {
			me->utf_count = 4;
			me->utf_char = (c & 0x03);
		    } else if ((*p & 0xfe) == 0xfc) {
			me->utf_count = 5;
			me->utf_char = (c & 0x01);
		    } else {
			/*
			 *  We got garbage, so ignore it. - FM
			 */
			me->utf_count = 0;
			me->utf_buf_p[0] = '\0';
			me->utf_buf_p = me->utf_buf;
		    }
		    /*
		    **	Get the next byte. - FM
		    */
		    continue;
		}
	    } else if (me->utf_count > 0) {
		/*
		**  Got an ASCII character when expecting
		**  UTF-8 multibytes, so ignore the buffered
		**  multibye characters and fall through with
		**  the current ASCII character. - FM
		*/
		me->utf_count = 0;
		me->utf_buf[0] = '\0';
		me->utf_buf_p = me->utf_buf;
		code = (UCode_t)c_unsign;
	    } else {
		/*
		**  Got a valid ASCII character, so fall
		**  through with it. - FM
		*/
		code = (UCode_t)c_unsign;
	    }
	}
	/*
	**  Convert characters from non-UTF-8 charsets
	**  to Unicode (if appropriate). - FM
	*/
	if (!(me->T.decode_utf8 &&
	      UCH(*p) > 127)) {
#ifdef NOTDEFINED
	    if (me->T.strip_raw_char_in)
		saved_char_in = c;
#endif /* NOTDEFINED */
	if (me->T.trans_to_uni &&
	    (TOASCII(code) >= LYlowest_eightbit[me->inUCLYhndl] ||  /* S/390 -- gil -- 0389 */
	     (code < ' ' && code != 0 &&
	     me->T.trans_C0_to_uni))) {
		/*
		**  Convert the octet to Unicode. - FM
		*/
	    code = (UCode_t)UCTransToUni(c, me->inUCLYhndl);
	    if (code > 0) {
		    saved_char_in = c;
		if (code < 256) {
			c = FROMASCII((char)code);
			c_unsign = UCH(c);
		}
	    }
	    } else if (code < 32 && code != 0 &&
		       me->T.trans_C0_to_uni) {
		/*
		**  Quote from SGML.c:
		**  	"This else if may be too ugly to keep. - KW"
		*/
		if (me->T.trans_from_uni &&
		    (((code = UCTransToUni(c, me->inUCLYhndl)) >= 32) ||
		     (me->T.transp &&
		      (code = UCTransToUni(c, me->inUCLYhndl)) > 0))) {
		    saved_char_in = c;
		    if (code < 256) {
			c = FROMASCII((char)code);
			c_unsign = UCH(c);
		    }
		} else {
		    uck = -1;
		    if (me->T.transp) {
			uck = UCTransCharStr(replace_buf, 60, c,
					     me->inUCLYhndl,
					     me->inUCLYhndl, NO);
		    }
		    if (!me->T.transp || uck < 0) {
			uck = UCTransCharStr(replace_buf, 60, c,
					     me->inUCLYhndl,
					     me->outUCLYhndl, YES);
		    }
		    if (uck == 0) {
			continue;
		    } else if (uck < 0) {
			me->utf_buf[0] = '\0';
			code = UCH(c);
		    } else {
			c = replace_buf[0];
			if (c && replace_buf[1]) {
			    HText_appendText(me->text, replace_buf);
			    continue;
			}
		    }
		    me->utf_buf[0] = '\0';
		    code = UCH(c);
		} /*  Next line end of ugly stuff for C0. - KW */
	    } else {
		me->utf_buf[0] = '\0';
		code = UCH(c);
	    }
	}
	/*
	**  At this point we have either code in Unicode
	**  (and c in latin1 if code is in the latin1 range),
	**  or code and c will have to be passed raw.
	*/

	/*
	**  If CJK mode is on, we'll assume the document matches
	**  the user's display character set, and if not, the
	**  user should toggle off raw/CJK mode to reload. - FM
	*/
	if (HTCJK != NOCJK) {
	    HText_appendCharacter(me->text, c);

#define PASSHICTRL (me->T.transp || \
		    code >= LYlowest_eightbit[me->inUCLYhndl])
#define PASS8859SPECL me->T.pass_160_173_raw
#define PASSHI8BIT (HTPassEightBitRaw || \
		    (me->T.do_8bitraw && !me->T.trans_from_uni))
	/*
	**  If HTPassHighCtrlRaw is set (e.g., for KOI8-R) assume the
	**  document matches and pass 127-160 8-bit characters.  If it
	**  doesn't match, the user should toggle raw/CJK mode off. - FM
	*/
	} else if (TOASCII(code) >= 127 && TOASCII(code) < 161 &&  /* S/390 -- gil -- 0427 */
		   PASSHICTRL && PASS8859SPECL) {
	    HText_appendCharacter(me->text, c);
	} else if (code == CH_SHY && PASS8859SPECL) {
	    HText_appendCharacter(me->text, c);
	/*
	**  If neither HTPassHighCtrlRaw nor CJK is set, play it safe
	**  and treat 160 (nbsp) as an ASCII space (32). - FM
	*/
	} else if (code == CH_NBSP) {
	    HText_appendCharacter(me->text, ' ');
	/*
	**  If neither HTPassHighCtrlRaw nor CJK is set, play it safe
	**  and ignore 173 (shy). - FM
	**  Now only ignore it for color style, which doesn't handle it anyway.
	**  Otherwise pass it on as LY_SOFT_HYPHEN and let HText deal with it.
	**  It should be either ignored, or displayed as a hyphen if it was
	**  indeed at the end of a line.  Well it should. - kw
	*/
	} else if (code == CH_SHY) {
#ifndef USE_COLOR_STYLE
	    HText_appendCharacter(me->text, LY_SOFT_HYPHEN);
#endif
	    continue;
	/*
	**  If we get to here, pass the displayable ASCII characters. - FM
	*/
	} else if ((code >= ' ' && TOASCII(code) < 127) ||
		   (PASSHI8BIT &&
		    c >= LYlowest_eightbit[me->outUCLYhndl]) ||
		   *p == '\n' || *p == '\t') {
	    HText_appendCharacter(me->text, c);
	/*
	**  Use an ASCII space (32) for ensp, emsp or thinsp. - FM
	*/
	} else if (code == 8194 || code == 8195 || code == 8201) {
	    HText_appendCharacter(me->text, ' ');
	/*
	**  If we want the raw character, pass it now. - FM
	*/
	} else if (me->T.use_raw_char_in && saved_char_in) {
	    HText_appendCharacter(me->text, saved_char_in);
/******************************************************************
 *   I. LATIN-1 OR UCS2  TO  DISPLAY CHARSET
 ******************************************************************/
	} else if ((chk = (BOOL) (me->T.trans_from_uni && code >= 160)) &&
		   (uck = UCTransUniChar(code,
					 me->outUCLYhndl)) >= ' ' &&  /* S/390 -- gil -- 0464 */
		   uck < 256) {
	    CTRACE((tfp, "UCTransUniChar returned 0x%.2lX:'%c'.\n",
			uck, FROMASCII((char)uck)));
	    HText_appendCharacter(me->text, ((char)(uck & 0xff)));
	} else if (chk &&
		   (uck == -4 ||
		    (me->T.repl_translated_C0 && uck > 0 && uck < ' ')) &&  /* S/390 -- gil -- 0481 */
		   /*
		   **  Not found; look for replacement string.
		   */
		   (uck = UCTransUniCharStr(replace_buf, 60, code,
					    me->outUCLYhndl, 0) >= 0)) {
	    /*
	    **	No further tests for valididy - assume that whoever
	    **	defined replacement strings knew what she was doing.
	    */
	    HText_appendText(me->text, replace_buf);
	/*
	**  If we get to here, and should have translated,
	**  translation has failed so far.
	*/
	} else if (chk && TOASCII(code) > 127 && me->T.output_utf8) {  /* S/390 -- gil -- 0498 */
	    /*
	    **	We want UTF-8 output, so do it now. - FM
	    */
	    if (*me->utf_buf) {
		HText_appendText(me->text, me->utf_buf);
		me->utf_buf[0] = '\0';
		me->utf_buf_p = me->utf_buf;
	    } else if (UCConvertUniToUtf8(code, replace_buf)) {
		HText_appendText(me->text, replace_buf);
	    } else {
		/*
		**  Out of luck, so use the UHHH notation (ugh). - gil
		*/  /* S/390 -- gil -- 0517 */
		sprintf(replace_buf, "U%.2lX", TOASCII(code));
		HText_appendText(me->text, replace_buf);
	    }
#ifdef NOTDEFINED
	} else if (me->T.strip_raw_char_in &&
		   UCH(*p) >= 192 &&
		   UCH(*p) < 255) {
	    /*
	    **	KOI special: strip high bit, gives
	    **	(somewhat) readable ASCII.
	    */
	    HText_appendCharacter(me->text, (char)(*p & 0x7f));
#endif /* NOTDEFINED */
	    /*
	    **  If we don't actually want the character,
	    **  make it safe and output that now. - FM
	    */
	} else if ((c_unsign > 0 &&
		      (int) c_unsign < LYlowest_eightbit[me->outUCLYhndl]) ||
		      (me->T.trans_from_uni && !HTPassEightBitRaw)) {
	    /*
	    **	If we do not have the "7-bit approximations" as our
	    **	output character set (in which case we did it already)
	    **	seek a translation for that.  Otherwise, or if the
	    **	translation fails, use UHHH notation. - FM
	    */
	    if ((chk = (BOOL) (me->outUCLYhndl !=
			UCGetLYhndl_byMIME("us-ascii"))) &&
		   (uck = UCTransUniChar(code,
					 UCGetLYhndl_byMIME("us-ascii")))
				      >= ' ' && TOASCII(uck) < 127) {  /* S/390 -- gil -- 0535 */
		/*
		**  Got an ASCII character (yippey). - FM
		*/
		c = FROMASCII((char)uck);
		HText_appendCharacter(me->text, c);
	    } else if ((chk && uck == -4) &&
		       (uck = UCTransUniCharStr(replace_buf,
						60, code,
						UCGetLYhndl_byMIME("us-ascii"),
						0) >= 0)) {
		/*
		**  Got a repacement string (yippey). - FM
		*/
		HText_appendText(me->text, replace_buf);
	    } else if (code == 8204 || code == 8205) {
		/*
		**	Ignore 8204 (zwnj) or 8205 (zwj), if we get to here. - FM
		*/
		CTRACE((tfp, "HTPlain_write: Ignoring '%ld'.\n", code));
	    } else if (code == 8206 || code == 8207) {
		/*
		**	Ignore 8206 (lrm) or 8207 (rlm), if we get to here. - FM
		*/
		CTRACE((tfp, "HTPlain_write: Ignoring '%ld'.\n", code));
	    } else {
		/*
		**  Out of luck, so use the UHHH notation (ugh). - FM
		*/
			/* do not print UHHH for now
			sprintf(replace_buf, "U%.2lX", code);
			HText_appendText(me->text, replace_buf);
			*/
		}
		/*
		**  If we get to here and have a monobyte character,
		**  pass it. - FM
		*/
	} else if (c_unsign != 0 && c_unsign < 256) {
	    HText_appendCharacter(me->text, c);
	}
#endif /* REMOVE_CR_ONLY */
    }
}

/*	Free an HTML object
**	-------------------
**
**	Note that the SGML parsing context is freed, but the created object is
**	not, as it takes on an existence of its own unless explicitly freed.
*/
PRIVATE void HTPlain_free ARGS1(
	HTStream *,	me)
{
    if (HTPlain_bs_pending >= 2)
	HText_appendCharacter(me->text, '_');
    FREE(me);
}

/*	End writing
*/
PRIVATE void HTPlain_abort ARGS2(
	HTStream *,	me,
	HTError,	e GCC_UNUSED)
{
    HTPlain_free(me);
}

/*		Structured Object Class
**		-----------------------
*/
PUBLIC CONST HTStreamClass HTPlain =
{
	"PlainPresenter",
	HTPlain_free,
	HTPlain_abort,
	HTPlain_put_character,	HTPlain_put_string, HTPlain_write,
};

/*		New object
**		----------
*/
PUBLIC HTStream* HTPlainPresent ARGS3(
	HTPresentation *,	pres GCC_UNUSED,
	HTParentAnchor *,	anchor,
	HTStream *,		sink GCC_UNUSED)
{

    HTStream* me = (HTStream*)malloc(sizeof(*me));
    if (me == NULL)
	outofmem(__FILE__, "HTPlain_new");
    me->isa = &HTPlain;

    HTPlain_lastraw = -1;

    me->utf_count = 0;
    me->utf_char = 0;
    me->utf_buf[0] = me->utf_buf[6] =me->utf_buf[7] = '\0';
    me->utf_buf_p = me->utf_buf;
    me->outUCLYhndl = HTAnchor_getUCLYhndl(anchor,UCT_STAGE_HTEXT);
    me->inUCLYhndl = HTAnchor_getUCLYhndl(anchor,UCT_STAGE_PARSER);
    HTPlain_getChartransInfo(me, anchor);
    UCSetTransParams(&me->T,
		     me->inUCLYhndl, me->inUCI,
		     me->outUCLYhndl,
		     HTAnchor_getUCInfoStage(anchor,UCT_STAGE_HTEXT));

    me->text = HText_new(anchor);
    HText_setStyle(me->text, LYstyles(HTML_XMP) );
    HText_beginAppend(me->text);

    return (HTStream*) me;
}