mirror of
https://github.com/pdf2htmlEX/pdf2htmlEX.git
synced 2024-10-06 03:51:40 +00:00
204 lines
6.9 KiB
C++
204 lines
6.9 KiB
C++
// pdftohtmlEX.cc
|
|
//
|
|
// Copyright (C) 2012 Lu Wang coolwanglu<at>gmail.com
|
|
|
|
#include <cstdio>
|
|
#include <cstdlib>
|
|
#include <cstddef>
|
|
#include <cstring>
|
|
#include <ctime>
|
|
#include <string>
|
|
#include <limits>
|
|
#include <iostream>
|
|
|
|
#include <boost/program_options.hpp>
|
|
#include <boost/filesystem.hpp>
|
|
#include <boost/algorithm/string/predicate.hpp>
|
|
|
|
#include <goo/GooString.h>
|
|
|
|
#include <Object.h>
|
|
#include <PDFDoc.h>
|
|
#include <PDFDocFactory.h>
|
|
#include <GlobalParams.h>
|
|
|
|
#include "HTMLRenderer.h"
|
|
#include "Param.h"
|
|
#include "config.h"
|
|
|
|
namespace po = boost::program_options;
|
|
using namespace std;
|
|
using namespace boost::filesystem;
|
|
|
|
Param param;
|
|
|
|
// variables
|
|
PDFDoc *doc = nullptr;
|
|
GooString *fileName = nullptr;
|
|
GooString *ownerPW, *userPW;
|
|
|
|
HTMLRenderer *htmlOut = nullptr;
|
|
|
|
int finished = -1;
|
|
|
|
po::options_description opt_visible("Options"), opt_hidden, opt_all;
|
|
po::positional_options_description opt_positional;
|
|
|
|
void show_usage(void)
|
|
{
|
|
cerr << "pdftohtmlEX version " << PDF2HTMLEX_VERSION << endl;
|
|
cerr << endl;
|
|
cerr << "Copyright 2012 Lu Wang (coolwanglu<at>gmail.com)" << endl;
|
|
cerr << endl;
|
|
cerr << "Usage: pdf2htmlEX [Options] <PDF-file>" << endl;
|
|
cerr << endl;
|
|
cerr << opt_visible << endl;
|
|
}
|
|
|
|
po::variables_map parse_options (int argc, char **argv)
|
|
{
|
|
opt_visible.add_options()
|
|
("help", "show all options")
|
|
("version,v", "show copyright and version info")
|
|
|
|
("owner-password,o", po::value<string>(¶m.owner_password)->default_value(""), "owner password (for encrypted files)")
|
|
("user-password,u", po::value<string>(¶m.user_password)->default_value(""), "user password (for encrypted files)")
|
|
|
|
("dest-dir", po::value<string>(¶m.dest_dir)->default_value("."), "destination directory")
|
|
("tmp-dir", po::value<string>(¶m.tmp_dir)->default_value("/tmp/pdf2htmlEX"), "temporary directory")
|
|
|
|
("first-page,f", po::value<int>(¶m.first_page)->default_value(1), "first page to process")
|
|
("last-page,l", po::value<int>(¶m.last_page)->default_value(numeric_limits<int>::max()), "last page to process")
|
|
|
|
("zoom", po::value<double>(¶m.zoom)->default_value(1.0), "zoom ratio")
|
|
("hdpi", po::value<double>(¶m.h_dpi)->default_value(144.0), "horizontal DPI for non-text")
|
|
("vdpi", po::value<double>(¶m.v_dpi)->default_value(144.0), "vertical DPI for non-text")
|
|
|
|
("process-nontext", po::value<int>(¶m.process_nontext)->default_value(1), "process nontext objects")
|
|
("single-html", po::value<int>(¶m.single_html)->default_value(1), "combine everything into one single HTML file")
|
|
("embed-base-font", po::value<int>(¶m.embed_base_font)->default_value(1), "embed local matched font for base 14 fonts in the PDF file")
|
|
("embed-external-font", po::value<int>(¶m.embed_external_font)->default_value(0), "embed local matched font for external fonts in the PDF file")
|
|
|
|
("heps", po::value<double>(¶m.h_eps)->default_value(1.0), "max tolerated horizontal offset (in pixels)")
|
|
("veps", po::value<double>(¶m.v_eps)->default_value(1.0), "max tolerated vertical offset (in pixels)")
|
|
("space-threshold", po::value<double>(¶m.space_threshold)->default_value(1.0/6), "distance no thiner than (threshold * em) will be considered as a space character")
|
|
("font-size-multiplier", po::value<double>(¶m.font_size_multiplier)->default_value(10.0), "setting a value greater than 1 would increase the rendering accuracy")
|
|
("always-apply-tounicode", po::value<int>(¶m.always_apply_tounicode)->default_value(0), "ToUnicode map is ignore for non-TTF fonts unless this switch is on")
|
|
|
|
("font-suffix", po::value<string>(¶m.font_suffix)->default_value(".ttf"), "suffix for extracted font files")
|
|
("font-format", po::value<string>(¶m.font_format)->default_value("truetype"), "format for extracted font files")
|
|
|
|
("debug", po::value<int>(¶m.debug)->default_value(0), "output debug information")
|
|
("clean-tmp", po::value<int>(¶m.clean_tmp)->default_value(1), "clean temporary files after processing")
|
|
;
|
|
|
|
opt_hidden.add_options()
|
|
("inputfilename", po::value<string>(¶m.input_filename)->default_value(""), "")
|
|
("outputfilename", po::value<string>(¶m.output_filename)->default_value(""), "")
|
|
;
|
|
|
|
opt_positional.add("inputfilename", 1).add("outputfilename",1);
|
|
|
|
opt_all.add(opt_visible).add(opt_hidden);
|
|
|
|
try {
|
|
po::variables_map opt_vm;
|
|
po::store(po::command_line_parser(argc, argv).options(opt_all).positional(opt_positional).run()
|
|
, opt_vm);
|
|
po::notify(opt_vm);
|
|
return opt_vm;
|
|
}
|
|
catch(...) {
|
|
show_usage();
|
|
exit(-1);
|
|
}
|
|
}
|
|
|
|
int main(int argc, char **argv)
|
|
{
|
|
auto opt_map = parse_options(argc, argv);
|
|
if (opt_map.count("version") || opt_map.count("help") || (param.input_filename == ""))
|
|
{
|
|
show_usage();
|
|
return -1;
|
|
}
|
|
|
|
//prepare the directories
|
|
for(const auto & p : {param.dest_dir, param.tmp_dir})
|
|
{
|
|
if(equivalent(PDF2HTMLEX_DATA_PATH, p))
|
|
{
|
|
cerr << "The specified directory \"" << p << "\" is the library path for pdf2htmlEX. Please use another one." << endl;
|
|
return -1;
|
|
}
|
|
}
|
|
|
|
try
|
|
{
|
|
create_directories(param.dest_dir);
|
|
create_directories(param.tmp_dir);
|
|
}
|
|
catch (const filesystem_error& err)
|
|
{
|
|
cerr << err.what() << endl;
|
|
return -1;
|
|
}
|
|
|
|
// read config file
|
|
globalParams = new GlobalParams();
|
|
|
|
// open PDF file
|
|
ownerPW = (param.owner_password == "") ? (nullptr) : (new GooString(param.owner_password.c_str()));
|
|
userPW = (param.user_password == "") ? (nullptr) : (new GooString(param.user_password.c_str()));
|
|
fileName = new GooString(param.input_filename.c_str());
|
|
|
|
doc = PDFDocFactory().createPDFDoc(*fileName, ownerPW, userPW);
|
|
|
|
delete userPW;
|
|
delete ownerPW;
|
|
|
|
if (!doc->isOk()) {
|
|
goto error;
|
|
}
|
|
|
|
// check for copy permission
|
|
if (!doc->okToCopy()) {
|
|
error(errNotAllowed, -1, "Copying of text from this document is not allowed.");
|
|
goto error;
|
|
}
|
|
|
|
param.first_page = min(max(param.first_page, 1), doc->getNumPages());
|
|
param.last_page = min(max(param.last_page, param.first_page), doc->getNumPages());
|
|
|
|
if(param.output_filename == "")
|
|
{
|
|
const string s = path(param.input_filename).filename().string();
|
|
if(boost::algorithm::ends_with(s, ".pdf"))
|
|
{
|
|
param.output_filename = s.substr(0, s.size() - 4) + ".html";
|
|
}
|
|
else
|
|
{
|
|
param.output_filename = s + ".html";
|
|
}
|
|
}
|
|
|
|
htmlOut = new HTMLRenderer(¶m);
|
|
htmlOut->process(doc);
|
|
delete htmlOut;
|
|
|
|
finished = 0;
|
|
|
|
// clean up
|
|
error:
|
|
if(doc) delete doc;
|
|
delete fileName;
|
|
if(globalParams) delete globalParams;
|
|
|
|
// check for memory leaks
|
|
Object::memCheck(stderr);
|
|
gMemReport(stderr);
|
|
|
|
return finished;
|
|
}
|