gridtext.c

基于rtos开发的浏览器！

	char,		ch)
{
    HTLine * line;
    HTStyle * style;
    int indent;

    /*
     *  Make sure we don't crash on NULLs.
     */
    if (!text)
	return;

    if (text->halted > 1) {
	/*
	 *  We should stop outputting more text, because low memory was
	 *  detected.  - kw
	 */
	if (text->halted == 2) {
	    /*
	     *  But if we haven't done so yet, first append a warning.
	     *  We should still have a few bytes left for that :).
	     *  We temporarily reset test->halted to 0 for this, since
	     *  this function will get called recursively. - kw
	     */
	    text->halted = 0;
	    text->kanji_buf = '\0';
	    HText_appendText(text, " *** MEMORY EXHAUSTED ***");
	}
	text->halted = 3;
	return;
    }
    /*
     *  Make sure we don't hang on escape sequences.
     */
    if (ch == '\033' && HTCJK == NOCJK)			/* decimal 27 */
	return;
#ifndef USE_SLANG
    /*
     *  Block 8-bit chars not allowed by the current display character
     *  set if they are below what LYlowest_eightbit indicates.
     *  Slang used its own replacements, so for USE_SLANG blocking here
     *  is not necessary to protect terminals from those characters.
     *  They should have been filtered out or translated by an earlier
     *  processing stage anyway. - kw
     */
    if ((unsigned char)ch >= 128 && HTCJK == NOCJK &&
	!text->T.transp && !text->T.output_utf8 &&
	(unsigned char)ch < LYlowest_eightbit[current_char_set])
	return;
#endif /* USE_SLANG */
    if ((unsigned char)ch == 155 && HTCJK == NOCJK) {	/* octal 233 */
	if (!HTPassHighCtrlRaw &&
	    !text->T.transp && !text->T.output_utf8 &&
	    (155 < LYlowest_eightbit[current_char_set]) &&
	    strncmp(LYchar_set_names[current_char_set],
		    "DosLatin1 (cp850)", 17) &&
	    strncmp(LYchar_set_names[current_char_set],
		    "DosLatinUS (cp437)", 18) &&
	    strncmp(LYchar_set_names[current_char_set],
		    "Macintosh (8 bit)", 17) &&
	    strncmp(LYchar_set_names[current_char_set],
		    "NeXT character set", 18)) {
	    return;
	}
    }

    line = text->last_line;
    style = text->style;

    indent = text->in_line_1 ? (int)style->indent1st : (int)style->leftIndent;

    if (HTCJK != NOCJK) {
	switch(text->state) {
	    case S_text:
		if (ch == '\033') {
		    /*
		    **  Setting up for CJK escape sequence handling (based on
		    **  Takuya ASADA's ([e-mail]) CJK Lynx). - FM
		    */
		    text->state = S_esc;
		    text->kanji_buf = '\0';
		    return;
		}
		break;

		case S_esc:
		/*
		 *  Expecting '$'or '(' following CJK ESC.
		 */
		if (ch == '$') {
		    text->state = S_dollar;
		    return;
		} else if (ch == '(') {
		    text->state = S_paren;
		    return;
		} else {
		    text->state = S_text;
		}

		case S_dollar:
		/*
		 *  Expecting '@', 'B', 'A' or '(' after CJK "ESC$".
		 */
		if (ch == '@' || ch == 'B' || ch=='A') {
		    text->state = S_nonascii_text;
		    return;
		} else if (ch == '(') {
		    text->state = S_dollar_paren;
		    return;
		} else {
		    text->state = S_text;
		}
		break;

		case S_dollar_paren:
		/*
		 * Expecting 'C' after CJK "ESC$(".
		 */
		if (ch == 'C') {
		    text->state = S_nonascii_text;
		    return;
		} else {
		    text->state = S_text;
		}
		break;

		case S_paren:
		/*
		 *  Expecting 'B', 'J', 'T' or 'I' after CJK "ESC(".
		 */
		if (ch == 'B' || ch == 'J' || ch == 'T')  {
		    /*
		     *  Can split here. - FM
		     */
		    text->permissible_split = (int)text->last_line->size;
		    text->state = S_text;
		    return;
		} else if (ch == 'I')  {
		    text->state = S_jisx0201_text;
		    /*
		     *  Can split here. - FM
		     */
		    text->permissible_split = (int)text->last_line->size;
		    return;
		} else {
		    text->state = S_text;
		}
		break;

		case S_nonascii_text:
		/*
		 *  Expecting CJK ESC after non-ASCII text.
		 */
		if (ch == '\033') {
		    text->state = S_esc;
		    text->kanji_buf = '\0';
		    return;
		} else {
		    ch |= 0200;
		}
		break;

		/*
		 *  JIS X0201 Kana in JIS support. - by ASATAKU
		 */
		case S_jisx0201_text:
		if (ch == '\033') {
		    text->state = S_esc;
		    text->kanji_buf = '\0';
		    return;
		} else {
		    text->kanji_buf = '\x8E';
		    ch |= 0200;
		}
		break;
	}

	if (!text->kanji_buf) {
	    if ((ch & 0200) != 0) {
		/*
		 *  JIS X0201 Kana in SJIS support. - by ASATAKU
		 */
	        if ((text->kcode == SJIS) &&
		    ((unsigned char)ch >= 0xA1) &&
		    ((unsigned char)ch <= 0xDF)) {
		    unsigned char c = (unsigned char)ch;
		    unsigned char kb = (unsigned char)text->kanji_buf;
		    JISx0201TO0208_SJIS(c,
					(unsigned char *)&kb,
					(unsigned char *)&c);
		    ch = (char)c;
		    text->kanji_buf = (char)kb;
	        } else {
		    text->kanji_buf = ch;
		    /*
		     *  Can split here. - FM
		     */
		    text->permissible_split = (int)text->last_line->size;
		    return;
	        }
	    }
	} else {
	    goto check_IgnoreExcess;
	}
    } else if (ch == '\033') {
	return;
    }

    if (IsSpecialAttrChar(ch)) {
#ifndef USE_COLOR_STYLE
	if (ch == LY_UNDERLINE_START_CHAR) {
	    line->data[line->size++] = LY_UNDERLINE_START_CHAR;
	    line->data[line->size] = '\0';
	    underline_on = ON;
	    if (!(dump_output_immediately && use_underscore))
		ctrl_chars_on_this_line++;
	    return;
	} else if (ch == LY_UNDERLINE_END_CHAR) {
	    line->data[line->size++] = LY_UNDERLINE_END_CHAR;
	    line->data[line->size] = '\0';
	    underline_on = OFF;
	    if (!(dump_output_immediately && use_underscore))
		ctrl_chars_on_this_line++;
	    return;
	} else if (ch == LY_BOLD_START_CHAR) {
	    line->data[line->size++] = LY_BOLD_START_CHAR;
	    line->data[line->size] = '\0';
	    bold_on = ON;
	    ctrl_chars_on_this_line++;
	    return;
	} else if (ch == LY_BOLD_END_CHAR) {
	    line->data[line->size++] = LY_BOLD_END_CHAR;
	    line->data[line->size] = '\0';
	    bold_on = OFF;
	    ctrl_chars_on_this_line++;
	    return;
	} else if (ch == LY_SOFT_HYPHEN) {
	    int i;

	    /*
	     *  Ignore the soft hyphen if it is the first character
	     *  on the line, or if it is preceded by a space or
	     *  hyphen. - FM
	     */
	    if (line->size < 1 || text->permissible_split >= (int)line->size)
		return;

	    for (i = (text->permissible_split + 1); line->data[i]; i++) {
		if (!IsSpecialAttrChar((unsigned char)line->data[i]) &&
		    !isspace((unsigned char)line->data[i]) &&
		    (unsigned char)line->data[i] != '-' &&
		    (unsigned char)line->data[i] != HT_NON_BREAK_SPACE &&
		    (unsigned char)line->data[i] != HT_EM_SPACE) {
		    break;
		}
	    }
	    if (line->data[i] == '\0') {
		return;
	    }
	}
#else
	return;
#endif
    }

    if (IS_UTF_EXTRA(ch)) {
	line->data[line->size++] = ch;
	line->data[line->size] = '\0';
	ctrl_chars_on_this_line++;
	return;
    }

    /*
     *  New Line.
     */
    if (ch == '\n') {
	    new_line(text);
	    text->in_line_1 = YES;	/* First line of new paragraph */
	    /*
	     *  There are some pages written in
	     *  different kanji codes. - TA & kw
	     */
	    if (HTCJK == JAPANESE)
		text->kcode = NOKANJI;
	    return;
    }

    /*
     *  Convert EM_SPACE to a space here so that it doesn't get collapsed.
     */
    if (ch == HT_EM_SPACE)
	ch = ' ';

    /*
     *  I'm going to cheat here in a BIG way.  Since I know that all
     *  \r's will be trapped by HTML_put_character I'm going to use
     *  \r to mean go down a line but don't start a new paragraph.
     *  i.e. use the second line indenting.
     */
    if (ch == '\r') {
	new_line(text);
	text->in_line_1 = NO;
	/*
	 *  There are some pages written in
	 *  different kanji codes. - TA & kw
	 */
	if (HTCJK == JAPANESE)
	    text->kcode = NOKANJI;
	return;
    }


    /*
     *  Tabs.
     */
    if (ch == '\t') {
	CONST HTTabStop * Tab;
	int target;	/* Where to tab to */
	int here;

	if (line->size > 0 && line->data[line->size-1] == LY_SOFT_HYPHEN) {
	    /*
	     *  A tab shouldn't follow a soft hyphen, so
	     *  if one does, we'll dump the soft hyphen. - FM
	     */
	    line->data[--line->size] = '\0';
	    ctrl_chars_on_this_line--;
	}
	here = (((int)line->size + (int)line->offset) + indent)
		- ctrl_chars_on_this_line; /* Consider special chars GAB */
	if (style->tabs) {	/* Use tab table */
	    for (Tab = style->tabs;
		Tab->position <= here;
		Tab++)
		if (!Tab->position) {
		    new_line(text);
		    return;
		}
	    target = Tab->position;
	} else if (text->in_line_1) {	/* Use 2nd indent */
	    if (here >= (int)style->leftIndent) {
	        new_line(text); /* wrap */
		return;
	    } else {
	        target = (int)style->leftIndent;
	    }
	} else {		/* Default tabs align with left indent mod 8 */
#ifdef DEFAULT_TABS_8
	    target = (((int)line->offset + (int)line->size + 8) & (-8))
			+ (int)style->leftIndent;
#else
	    new_line(text);
	    return;
#endif
	}

	if (target > (LYcols-1) - (int)style->rightIndent &&
	    HTOutputFormat != WWW_SOURCE) {
	    new_line(text);
	    return;
	} else {
	    /*
	     *  Can split here. - FM
	     */
	    text->permissible_split = (int)line->size;
	    if (line->size == 0) {
	        line->offset = line->offset + target - here;
	    } else {
	        for (; here<target; here++) {
		    /* Put character into line */
		    line->data[line->size++] = ' ';
		    line->data[line->size] = '\0';
	        }
	    }
	    return;
	}
	/*NOTREACHED*/
    } /* if tab */


    if (ch == ' ') {
	/*
	 *  Can split here. - FM
	 */
	text->permissible_split = (int)text->last_line->size;
	/*
	 *  There are some pages written in
	 *  different kanji codes. - TA
	 */
	if (HTCJK == JAPANESE)
	    text->kcode = NOKANJI;
    }

    /*
     *  Check if we should ignore characters at the wrap point.
     */
check_IgnoreExcess:
    if (text->IgnoreExcess &&
	((indent + (int)line->offset + (int)line->size) +
	(int)style->rightIndent - ctrl_chars_on_this_line) >= (LYcols-1))
	return;

    /*
     *  Check for end of line.
     */
    if (((indent + (int)line->offset + (int)line->size) +
	 (int)style->rightIndent - ctrl_chars_on_this_line +
	 ((line->size > 0) &&
	  (int)(line->data[line->size-1] ==
				LY_SOFT_HYPHEN ?
					     1 : 0))) >= (LYcols - 1)) {

	if (style->wordWrap && HTOutputFormat != WWW_SOURCE) {
	    split_line(text, text->permissible_split);
	    if (ch == ' ') return;	/* Ignore space causing split */

	}  else if (HTOutputFormat == WWW_SOURCE) {
		 /*
		  *  For source output we don't want to wrap this stuff
		  *  unless absolutely necessary. - LJM
		  *  !
		  *  If we don't wrap here we might get a segmentation fault.
		  *  but let's see what happens
		  */
		if ((int)line->size >= (int)(MAX_LINE-1))
		   new_line(text);  /* try not to linewrap */
	} else {
		/*
		 *  For normal stuff like pre let's go ahead and
		 *  wrap so the user can see all of the text.
		 */
		new_line(text);
	}
    } else if ((int)line->size >= (int)(MAX_LINE-1)) {
	/*
	 *  Never overrun memory if LYcols is set to a large value - KW
	 */
	new_line(text);
    }

    /*
     *  Insert normal characters.
     */
    if (ch == HT_NON_BREAK_SPACE) {
	ch = ' ';
    }

    if (ch & 0x80)
	text->have_8bit_chars = YES;

    {
	HTFont font = style->font;
	unsigned char hi, lo, tmp[2];

	line = text->last_line; /* May have changed */
	if (HTCJK != NOCJK && text->kanji_buf) {
	    hi = (unsigned char)text->kanji_buf, lo = (unsigned char)ch;
	    if (HTCJK == JAPANESE && text->kcode == NOKANJI) {
		if (IS_SJIS(hi, lo, text->in_sjis) && IS_EUC(hi, lo)) {
		    text->kcode = NOKANJI;
		} else if (IS_SJIS(hi, lo, text->in_sjis)) {
		    text->kcode = SJIS;
		} else if (IS_EUC(hi, lo)) {
		    text->kcode = EUC;
		}
	    }
	    if (HTCJK == JAPANESE &&
		(kanji_code == EUC) && (text->kcode == SJIS)) {
		SJIS_TO_EUC1(hi, lo, tmp);
		line->data[line->size++] = tmp[0];
		line->data[line->size++] = tmp[1];
	    } else if (HTCJK == JAPANESE &&
		       (kanji_code == EUC) && (text->kcode == EUC)) {
		JISx0201TO0208_EUC(hi, lo, &hi, &lo);
		line->data[line->size++] = hi;
		line->data[line->size++] = lo;
	    } else if (HTCJK == JAPANESE &&
		       (kanji_code == SJIS) && (text->kcode == EUC)) {
		EUC_TO_SJIS1(hi, lo, tmp);
		line->data[line->size++] = tmp[0];
		line->data[line->size++] = tmp[1];
	    } else {
		line->data[line->size++] = hi;
		line->data[line->size++] = lo;
	    }
	    text->kanji_buf = 0;
	} else if (HTCJK != NOCJK) {
	    line->data[line->size++] = (kanji_code != NOKANJI) ?
							    ch :
					  (font & HT_CAPITALS) ?
						   TOUPPER(ch) : ch;
	} else {
	    line->data[line->size++] =	/* Put character into line */
		font & HT_CAPITALS ? TOUPPER(ch) : ch;
	}
	line->data[line->size] = '\0';
	if (font & HT_DOUBLE)		/* Do again if doubled */
	    HText_appendCharacter(text, HT_NON_BREAK_SPACE);
	    /* NOT a permissible split */

	if (ch == LY_SOFT_HYPHEN) {
	    ctrl_char