mirror of
https://github.com/pdf2htmlEX/pdf2htmlEX.git
synced 2024-09-18 21:40:06 +00:00
136 lines
3.2 KiB
C++
136 lines
3.2 KiB
C++
/*
|
|
* text.cc
|
|
*
|
|
* Handling text & font, and relative stuffs
|
|
*
|
|
* Copyright (C) 2012 Lu Wang <coolwanglu@gmail.com>
|
|
*/
|
|
|
|
|
|
#include <algorithm>
|
|
|
|
#include "HTMLRenderer.h"
|
|
#include "TextLineBuffer.h"
|
|
#include "util/namespace.h"
|
|
#include "util/unicode.h"
|
|
|
|
namespace pdf2htmlEX {
|
|
|
|
using std::all_of;
|
|
using std::cerr;
|
|
using std::endl;
|
|
|
|
void HTMLRenderer::drawString(GfxState * state, GooString * s)
|
|
{
|
|
if(s->getLength() == 0)
|
|
return;
|
|
|
|
auto font = state->getFont();
|
|
double cur_letter_space = state->getCharSpace();
|
|
double cur_word_space = state->getWordSpace();
|
|
|
|
// Writing mode fonts and Type 3 fonts are rendered as images
|
|
// I don't find a way to display writing mode fonts in HTML except for one div for each character, which is too costly
|
|
// For type 3 fonts, due to the font matrix, still it's hard to show it on HTML
|
|
if( (font == nullptr)
|
|
|| (font->getWMode())
|
|
|| (font->getType() == fontType3)
|
|
)
|
|
{
|
|
return;
|
|
}
|
|
|
|
//hidden
|
|
if((state->getRender() & 3) == 3)
|
|
{
|
|
return;
|
|
}
|
|
|
|
// see if the line has to be closed due to state change
|
|
check_state_change(state);
|
|
prepare_text_line(state);
|
|
|
|
// Now ready to output
|
|
// get the unicodes
|
|
char *p = s->getCString();
|
|
int len = s->getLength();
|
|
|
|
double dx = 0;
|
|
double dy = 0;
|
|
double dx1,dy1;
|
|
double ox, oy;
|
|
|
|
int nChars = 0;
|
|
int nSpaces = 0;
|
|
int uLen;
|
|
|
|
CharCode code;
|
|
Unicode *u = nullptr;
|
|
|
|
while (len > 0)
|
|
{
|
|
auto n = font->getNextChar(p, len, &code, &u, &uLen, &dx1, &dy1, &ox, &oy);
|
|
|
|
if(!(equal(ox, 0) && equal(oy, 0)))
|
|
{
|
|
cerr << "TODO: non-zero origins" << endl;
|
|
}
|
|
|
|
bool is_space = false;
|
|
if (n == 1 && *p == ' ')
|
|
{
|
|
++nSpaces;
|
|
is_space = true;
|
|
}
|
|
|
|
if(is_space && (param->space_as_offset))
|
|
{
|
|
// ignore horiz_scaling, as it's merged in CTM
|
|
text_line_buf->append_offset((dx1 * cur_font_size + cur_letter_space + cur_word_space) * draw_text_scale);
|
|
}
|
|
else
|
|
{
|
|
if((param->decompose_ligature) && (uLen > 1) && all_of(u, u+uLen, isLegalUnicode))
|
|
{
|
|
text_line_buf->append_unicodes(u, uLen);
|
|
}
|
|
else
|
|
{
|
|
if(cur_font_info->use_tounicode)
|
|
{
|
|
Unicode uu = check_unicode(u, uLen, code, font);
|
|
text_line_buf->append_unicodes(&uu, 1);
|
|
}
|
|
else
|
|
{
|
|
Unicode uu = unicode_from_font(code, font);
|
|
text_line_buf->append_unicodes(&uu, 1);
|
|
}
|
|
}
|
|
}
|
|
|
|
dx += dx1;
|
|
dy += dy1;
|
|
|
|
++nChars;
|
|
p += n;
|
|
len -= n;
|
|
}
|
|
|
|
double hs = state->getHorizScaling();
|
|
|
|
// horiz_scaling is merged into ctm now,
|
|
// so the coordinate system is ugly
|
|
dx = (dx * cur_font_size + nChars * cur_letter_space + nSpaces * cur_word_space) * hs;
|
|
|
|
dy *= cur_font_size;
|
|
|
|
cur_tx += dx;
|
|
cur_ty += dy;
|
|
|
|
draw_tx += dx;
|
|
draw_ty += dy;
|
|
}
|
|
|
|
} // namespace pdf2htmlEX
|