void ucstring_append_bytes(de_ucstring *s, const de_byte *buf, de_int64 buflen,
unsigned int conv_flags, int encoding)
{
int ret;
de_int64 pos = 0;
de_int32 ch;
de_int64 code_len;
// Adjust buflen if necessary.
if(conv_flags & DE_CONVFLAG_STOP_AT_NUL) {
char *tmpp;
tmpp = de_memchr(buf, 0, (size_t)buflen);
if(tmpp) {
buflen = (const de_byte*)tmpp - buf;
}
}
while(pos<buflen) {
if(encoding==DE_ENCODING_UTF8) {
ret = de_utf8_to_uchar(&buf[pos], buflen-pos, &ch, &code_len);
if(!ret) {
ch = '_';
code_len = 1;
}
}
else if(encoding==DE_ENCODING_UTF16LE) {
ret = de_utf16x_to_uchar(&buf[pos], buflen-pos, &ch, &code_len, 1);
if(!ret) {
ch = '_';
code_len = 2;
}
}
else if(encoding==DE_ENCODING_UTF16BE) {
ret = de_utf16x_to_uchar(&buf[pos], buflen-pos, &ch, &code_len, 0);
if(!ret) {
ch = '_';
code_len = 2;
}
}
else {
ch = de_char_to_unicode(s->c, buf[pos], encoding);
if(ch==DE_INVALID_CODEPOINT) {
ch = '_';
}
code_len = 1;
}
ucstring_append_char(s, ch);
pos += code_len;
}
}
// create bitmap_font object
static void do_make_image(deark *c, lctx *d)
{
struct de_bitmap_font *font = NULL;
de_int64 i;
de_int64 pos;
font = de_create_bitmap_font(c);
font->has_nonunicode_codepoints = 1;
if(d->encoding!=DE_ENCODING_UNKNOWN)
font->has_unicode_codepoints = 1;
font->prefer_unicode = 0;
font->nominal_width = (int)d->nominal_char_width;
font->nominal_height = (int)d->char_height;
font->num_chars = d->num_chars_stored;
font->char_array = de_malloc(c, font->num_chars * sizeof(struct de_bitmap_font_char));
for(i=0; i<d->num_chars_stored; i++) {
de_int64 char_width;
de_int64 char_offset;
de_int32 char_index;
de_int64 num_tiles;
de_int64 tile;
de_int64 row;
pos = d->hdrsize + d->char_entry_size*i;
char_width = de_getui16le(pos);
if(d->char_entry_size==6)
char_offset = de_getui32le(pos+2);
else
char_offset = de_getui16le(pos+2);
de_dbg2(c, "char[%d] width=%d offset=%d\n", (int)(d->first_char + i), (int)char_width, (int)char_offset);
num_tiles = (char_width+7)/8;
if(i == d->num_chars_stored-1) {
// Arbitrarily put the "absolute space" char at codepoint 256,
// and U+2002 EN SPACE (best I can do).
font->char_array[i].codepoint_nonunicode = 256;
font->char_array[i].codepoint_unicode = 0x2002;
}
else {
char_index = (de_int32)d->first_char + (de_int32)i;
font->char_array[i].codepoint_nonunicode = char_index;
if(font->has_unicode_codepoints) {
if(char_index<32 && d->dfCharSet==0) {
// This kind of font usually doesn't have glyphs below 32.
// If it does, assume that they are VT100 line drawing characters.
font->char_array[i].codepoint_unicode =
de_char_to_unicode(c, 95+char_index, DE_ENCODING_DEC_SPECIAL_GRAPHICS);
}
else {
font->char_array[i].codepoint_unicode =
de_char_to_unicode(c, char_index, d->encoding);
}
}
}
font->char_array[i].width = (int)char_width;
font->char_array[i].height = (int)d->char_height;
font->char_array[i].rowspan = num_tiles;
font->char_array[i].bitmap = de_malloc(c, d->char_height * num_tiles);
for(row=0; row<d->char_height; row++) {
for(tile=0; tile<num_tiles; tile++) {
font->char_array[i].bitmap[row * font->char_array[i].rowspan + tile] =
de_getbyte(char_offset + tile*d->char_height + row);
}
}
}
de_font_bitmap_font_to_image(c, font, d->fi, 0);
if(font) {
if(font->char_array) {
for(i=0; i<font->num_chars; i++) {
de_free(c, font->char_array[i].bitmap);
}
de_free(c, font->char_array);
}
de_destroy_bitmap_font(c, font);
}
}
static void do_text_paragraph(deark *c, lctx *d, struct para_info *pinfo)
{
i64 i, k;
if(!d->html_outf) return;
if((pinfo->papflags & 0x06)!=0) {
// TODO: Decode headers and footers somehow.
do_emit_raw_sz(c, d, pinfo, "<p class=r>");
do_emit_raw_sz(c, d, pinfo, (pinfo->papflags&0x01)?"footer":"header");
do_emit_raw_sz(c, d, pinfo, " definition</p>\n");
return;
}
pinfo->in_para = 0;
pinfo->xpos = 0;
pinfo->space_count = 0;
pinfo->has_content = 0;
pinfo->in_span = 0;
default_text_styles(&pinfo->text_styles_wanted);
default_text_styles(&pinfo->text_styles_current);
for(i=0; i<pinfo->thisparalen; i++) {
u8 incp;
incp = de_getbyte(pinfo->thisparapos+i);
if(incp==0x0d && i<pinfo->thisparalen-1) {
if(de_getbyte(pinfo->thisparapos+i+1)==0x0a) {
// Found CR-LF combo
i++;
ensure_in_para(c, d, pinfo);
end_para(c, d, pinfo);
continue;
}
}
if(incp!=32 && pinfo->space_count>0) {
int nonbreaking_count, breaking_count;
if(!pinfo->in_para && pinfo->space_count==1) {
// If the paragraph starts with a single space, make it nonbreaking.
nonbreaking_count = 1;
breaking_count = 0;
}
else {
// Else make all spaces but the last one nonbreaking
nonbreaking_count = pinfo->space_count-1;
breaking_count = 1;
}
ensure_in_para(c, d, pinfo);
for(k=0; k<nonbreaking_count; k++) {
do_emit_codepoint(c, d, pinfo, 0xa0);
}
if(breaking_count>0) {
if(pinfo->xpos>70) {
// We don't do proper word wrapping of the HTML source, but
// maybe this is better than nothing.
do_emit_raw_sz(c, d, pinfo, "\n");
}
else {
do_emit_codepoint(c, d, pinfo, 32);
}
}
pinfo->space_count=0;
}
if(incp>=33) {
i32 outcp;
outcp = de_char_to_unicode(c, (i32)incp, d->input_encoding);
do_emit_codepoint(c, d, pinfo, outcp);
}
else {
switch(incp) {
case 9: // tab
pinfo->text_styles_wanted.tab_style = 1;
do_emit_codepoint(c, d, pinfo, 0x2192);
pinfo->text_styles_wanted.tab_style = 0;
break;
case 10:
case 11:
ensure_in_para(c, d, pinfo);
do_emit_raw_sz(c, d, pinfo, "<br>\n");
pinfo->has_content = 1;
break;
case 12: // page break
end_para(c, d, pinfo);
do_emit_raw_sz(c, d, pinfo, "<hr>\n");
break;
case 31:
break;
case 32:
pinfo->space_count++;
break;
default:
do_emit_codepoint(c, d, pinfo, 0xfffd);
}
}
}
end_para(c, d, pinfo);
}
